Power Supply Technical Guide - XP (072-152)

AC UPS Systems
• Software
Virtually all models of UPS are provided as standard with shutdown and monitoring software. This
software is supplied on a CD and typically covers all windows-based platforms. Connection is usually
via a supplied RS232 cable.
The key reason for utilizing shutdown software is to ensure that when a mains fail or battery low
signal is received from the UPS, the computer powered by the UPS is shut down in an orderly
fashion, saving all files and avoiding loss of data.
Control is via the console in the software and allows the user, via a serial connection, to schedule
battery tests, shut down the system and carry out general UPS housekeeping.
In combination with its flexible shutdown features, the software also lets the user know which
programs were shut down at last mains failure to ensure that no saved files are missed.
A typical software screen shot and CD are shown below.
66
Inverters, Frequency Converters & Static Switches
• DC/AC Inverters
DC/AC inverters are used to produce an AC supply from a DC power source and are available in
numerous shapes and sizes. Inverters are utilized within UPS systems to ensure a continuous source
of power in the event of a mains failure.
As a standalone item they can be used in a wide range of fixed and mobile applications, each with
their own particular requirements and limitations. Following are some of the features to consider
when specifying or evaluating an inverter for a particular application.
Output Waveform
There are 2 major waveform types of inverters, sine wave (or true sine wave), and modified sine wave
(actually a square wave).
Sine Wave
Sine wave inverters generate a waveform similar to that supplied by the utility company. These
inverters have the advantage that they will work with virtually all mains-powered equipment with the
same corresponding voltage. Some appliances, such as motors, will only produce full output with
sine wave power. A few appliances, such as light dimmers and some older battery chargers, require
a sine wave in order to work at all.
Modified Sine Wave (quasi-sine)
A modified sine wave inverter actually has a waveform more like a square wave, but with a dwell
period at zero cross-over. This type of inverter can readily be used with switch mode power supplies
as the load.
67
Input Voltage
The DC voltage available for a particular application will affect the rating and efficiency of an inverter
due to the higher currents being drawn from low voltage power sources. Typical input ranges are
detailed in the table below.
Nominal Battery Voltage (V) Voltage Ranges (V)

12 10.2 - 14.4
24 20.0 - 32.0
48 40.0 - 60.0
110 88.0 - 132.0
220 176.0 - 264.0
Surge Rating
Inverters typically have a continuous rating and a surge rating. The surge rating is specified in Watts
or VA for a fixed time period (normally several seconds). The surge capacity varies considerably
between inverters, even within the same brand, and may range from as little as 20% to as much as
300%. Generally, a three to fifteen second surge rating is enough to cover 99% of applications.
Pump motors, for example, have a surge of approximately half a second on start-up.
Synchronization Signal
If the inverter is to be used in conjunction with a static bypass switch, the inverter switching
frequency and phase relation of the output must be synchronized to that of the mains, or alternative
AC source, in order to reduce disturbance in load supply during transfer.
Reverse Input Protection
With mobile inverters, where the input supply is likely to be connected and disconnected multiple
times during the product’s lifetime, it is recommended that some form of reverse input protection is
fitted to ensure the unit is not damaged if the supply is accidentally reversed.
Alarms & Controls
Most inverters come with some form of monitoring, either front panel LEDs or Volt-free/ dry contacts.
Typical alarm types are:
• DC OK
• Low DC voltage
• Overload
• Output healthy
• Inverter ON
More complex alarms and control options can be added. However, it should be remembered that the
more complex the system, the less reliable it becomes.
68
• Frequency Converters
The advent of the universal input PSU has enabled the majority of electrical and electronic equipment
manufacturers to market their equipment almost anywhere in the world.
However, there are some applications incorporating motors/compressors/timers/avionics etc. where

the frequency can have a dramatic effect on the functionality of the product used.
Typical frequency conversions are:
50Hz to 60Hz or 60Hz to 50Hz

50Hz or 60Hz to 400Hz
There are two types of frequency converters available; rotary and static.
Rotary Converter
A rotary converter comprises an electric motor and alternator, the incoming supply being converted
to electro-mechanical energy; this is then converted back to the required frequency and voltage via
the alternator. These types of converters tend to be employed for high power avionics applications.
Static Converters
In a static converter, the frequency change is achieved by first converting the primary AC source into
DC via an AC/DC rectifier, before conversion back to AC at the required frequency/voltage via a
DC/AC inverter. In this way, the output characteristic of the converter is independent from the input
source, providing a clean and stable power source.
Some commercial UPS have the option of setting the output frequency independently from the input
by inhibiting the static switch enabling them to be used as frequency converters.
• Static Transfer Switches
Static Transfer Switches (STS) are solid-state switches designed to automatically or manually switch
between two synchronized AC power sources with minimal interruption of the power to the load.
This interruption is typically no longer than a quarter of a cycle in case of failure of one source or by
manual initiation for test or maintenance.
DC Source Inverter Critical

Load
Static
Transfer
Switch
Alternate Source(s)
Static transfer switch system
69
The switching action is designed to ensure that two power sources are not connected together, and
is based around the Silicon Controlled Rectifier (SCR).
The basic configuration is back-to-back SCRs on

the preferred and alternate sides of the switch. Preferred
Figure 1 (right) shows a single-phase, two-source Source
version. The three-phase and/or three-source STS

is merely an extension of the same fundamental
concept.
Critical
Load
An SCR allows conduction only in the forward
direction.
During normal operation, SCRs associated with the Alternate

Source
preferred source are in the ON-state, while those
associated with the alternate source are in the
OFF-state (in Figure 1, the bold lines denote
conduction). Figure 1
Current-sensing circuits constantly monitor the states of the preferred and alternate sources and
feed the information to the supervisory microprocessor controller.
Upon sensing the loss of the preferred source, the microprocessor control instructs the gate-driven
SCRs on the alternate side to turn ON (denoted by the bold lines in Figure 2). The transfer from the
preferred to the alternate source is so fast (less than a quarter electrical cycle), that even the most
sensitive electrical or electronic loads are unable to determine its occurrence.
Preferred
Source
Critical
Load
Alternate
Source
Figure 2
Since it is entirely based on solid state technology, the static transfer switch has no mechanical
moving parts and requires minimal maintenance. It has an AC-to-AC efficiency in excess of 99%, a
small footprint and requires no batteries when used as an alternative to the UPS.
70
Thermal Management
• System Cooling Fan Selection
Power Losses
Power losses occur in all electronic components. The effect of these losses becomes greater as
more and more components are squeezed into smaller spaces. The result of this 'miniaturisation' is
higher levels of heat per cubic volume of space. This waste heat can be considered to be system
losses and expressed as follows:
Input power = output power + losses

Efficiency = output power/input power (always
less than 1) 120
100
Losses can be in the form of heat, noise, light or
work, and are expressed in Watts. The heat 80
Temperature (˚C)
generated by a component does not only pass 60
into the air around that component, it is also

40
absorbed by adjacent components, the PCB and
the equipment case. This 'waste' heat affects the 20
performance of the adjacent components causing

0
them to operate in higher ambient temperatures. Direction of heat current
Although modern design aids such as 'fluid

dynamic analysis' can assist in the thermal design Component
Component Mounting
Air-filled Interior
Cabinet Panel
Ambient Area
of equipment, the costs associated with the
system often restrict its use. The majority of
equipment designers must rely upon experience
and knowledge to assist in the selection of a
cooling system.
Representation of the temperature
The designer’s dilemma is whether equipment gradient of a device
can be designed to ensure that all waste heat
can be removed by convection alone, or whether
'best practice' calls for the incorporation of forced
cooling. The thermal control of electronic
equipment should be considered as part of the
overall design specification which, when achieved, will result in a coherent design, exhibiting greater
reliability and life expectancy.
Establish Allowable Temperature Rise
First, it is necessary to establish the maximum operating temperature in which either the power
supply or the electronics can safely operate. For example, this could be 50°C, the typical maximum
operating temperature of a power supply. If the enclosure in which it is contained is to be used in a
non air-conditioned environment, where the maximum temperature could reach as high as 40°C, the
maximum temperature rise allowed is 10°C.
71
Thermal Management
Establish Power to be Dissipated
If our example has all the load within the equipment then the total power dissipated within the
enclosure is the power dissipated by the load, and the power dissipated by the power supply due to
its inefficiency.
The load power is 260W

+
The power lost due to the inefficiency of the power supply,
which is 80% efficient, so the power lost is 65W
=
Total power to be dissipated is 325W
The airflow needed through a system can be calculated as follows:
Air flow (m3/hr) = 2.6 x Power (W)

Tc
Where Tc is the allowable temperature rise of the air in the equipment in °C, calculated as the
maximum air temperature required minus the maximum temperature of air coming into the
equipment (the ambient temperature). Airflow is measured in m3/hr, and the power in Watts is the
amount of heat dissipated into the box.
The power supply often has its flow rate given as a linear figure, while fan manufacturers typically
specify a volumetric flow rate. To convert from one to the other, convert the volumetric flow rate in
m3/hr to m3/s (divide by 3600), then divide the resultant figure by the active area of the fan. The
active area of the fan is the area traced by the tips of the blades minus the area of the central hub
(which is not directly contributing to air movement).
Linear flow rate (m/s) = Volumetric flow rate (m3/s)

Active fan area (m2)
Therefore in our example:
Air flow (m3/hr) = 2.6 x total power dissipated (W)

Allowable temp rise (˚C)
= 2.6 x 325 W
10 ˚C
= 84.5 m3/hr (49.77 CFM)
To convert m3/hr to cubic feet per minute (CFM) multiply by 0.589.
72
Thermal Management
Characteristic Curves for Equipment Pressure Drop
90
Airflow figures published for fans are given in free air.
Pa In practice, an enclosure provides resistance to air
70 movement. It’s rather like trying to blow up a balloon.
This resistance will change with each equipment
60
design due to PCB sizes and positions and the effect
Pressure loss pv
50
of other components which will provide resistance to
40 airflow. Luckily, there is an approximation to back
30 pressure which can be applied. This graph is an
approximation or an average, based on accumulated
20
historical data from fan manufacturers and is
10 p = Rv 2 V 2
applicable to most electronic equipment. The graph
0 shows the flow rate along the horizontal axis in litres
0 10 20 30 40 50 Vs 60
Flow rate V
per second and the back pressure on the vertical axis
in Pascals.
Estimating Back Pressure
The graph above is then used to estimate the back 90
pressure, so for a system which requires 14.2l/s, Pa

the back pressure is 5 Pascals. 70
60
Pressure loss pv
50
40
30
20
10 p = Rv 2 V 2
Using Fan Characteristic Curves 0

0 10 20 30 40 50 Vs 60
Flow rate V
We know that the back pressure is 5 Pa and we
require 84.5m3/hr. Therefore, from the graph below
it can be seen that fan 2 is the suitable selection.
0 20 40 60 CFM 80
50
Pa
3
40 Curve Type
1 FAN 1
pf
30 FAN 2
2 2
Static Pressure
3 FAN 3
20
1
10
0
0 20 40 60 80 100 m3/h 140
Flow Rate V
73
Thermal Management
• Cooling power supplies

Power supplies generate waste heat which has to be dissipated. They typically have either
convection cooled or forced cooled ratings or, in some cases, both. Forced cooled power supplies
may incorporate a cooling fan, or may specify the user cooling required to operate the unit at
maximum load and ambient temperature.
Where user cooling is required it is most important that the power supply cooling is adequate for
both safe operation and adequate service life. It is very application specific and dependent on the
ambient temperature, applied load and physical location with respect to the cooling fan and other
system assemblies.
The main difference between convection and force cooled products is in the power density offered
for a given efficiency, convection cooled products offering a lower power density meaning that they
occupy a larger volume. A power supply on a 3” x 5” industry standard footprint may have a
convection rating of 100W while the force cooled version may have a rating as high as 200W.
Convection cooling
Where the power supply has a convection cooled rating, it is intended to be used in an environment
where there is free air. The system designer must ensure that there is adequate space around and
above the unit for free air convection currents to cool the unit and must also ensure that the ambient
temperature local to the power supply is controlled to a level within its maximum ratings.
Forced cooling
Force cooled products with integral cooling fans are relatively easy to apply as it is a simple matter of
ensuring that the maximum specified ambient temperature is not exceeded for a given load rating
and that the intake and exhaust areas are not obstructed.
Typically, power supplies that require the user to provide forced air cooling will specify a minimum
required airflow. This is usually for operation at 100% of the power rating at the maximum ambient
temperature allowed.
The required airflow is often specified in Cubic Feet per Minute (CFM) which is also the common
rating for cooling fans. The effectiveness of cooling fans installed in enclosures must be given
consideration, as discussed earlier in this section, and the CFM rating deals in volume of air rather
than air speed, which is the important factor. The object is to maintain the components used within
the power supply at a safe operating temperature and to ensure adequate service life.
When the required airflow is specified in CFM it assumes that the power supply is installed in an area
which is relatively similar to it’s own cross sectional area. This is rarely the case as the power supply
is typically used as a sub-assembly within a complete equipment enclosure. It will also assume that
the air is directed at the power supply, which may also not be the case, so converting to Linear Feet
per Minute (LFM) or meters per second (m/s) provides a more valid criterion as linear air speed
measurements specify where the air is flowing and directly relate to heat transfer.
74
Thermal Management
Example based on an EMA212 power supply
5.00 (127.0)
0.22
(5.7) 4.55 (115.6)
CN3
6 5 CN2 CN2
1 1
10 1
TR4 TR11 2 2
AIR FLOW 3.00 2.55 C14 SCREW 6.35 mm

CN2
12 CFM (76.2) (64.8) TYPE FASTONTYPE(2)
+ve
N -ve
CN1
L
T7
C42
SAFETY EARTH - 4.8 mm Faston 4 X ø 0.14 (3.5) MOUNTING HOLES
1.22 0.06
(31.0) (1.5)
0.06 (1.5)
In the case above, the power supply requires forced air of 12 CFM in the direction indicated by the
arrow. The cross sectional area is:-
3” x 1.34” = 4 inches2 or 0.028 feet2
Therefore the air velocity required is:-
12/0.028 = 429 LFM or 2.17 m/s
This air speed can be measured locally to the power supply to ensure that sufficient forced air
cooling is being applied.
Evaluation of the application
As discussed earlier, the object is to maintain the components used within the power supply at a
safe operating temperature and to ensure adequate service life. Given the huge potential for variation
between one application and another, the only real test is measurement of the temperature of the
critical components within the power supply assembly when installed within the end application
under the worst case external ambient conditions. The other option is to model the application
exactly using a suitable software simulation.
The criteria for safe operation will be specified for the power supply in question or can be obtained
from the manufacturer. For the example above, the specific component temperatures for safe
operation for the EMA212 are given on the next page; these are typical for a power supply of this
type.
75
Thermal Management
Temperature Measurements (Ambient ≤50 ºC)

Component Max temperature ºC
TR4 case 110 ºC
C14 105 ºC
C42 105 ºC
TR11 case 110 ºC
T7 coil 120 ºC
While these figures will ensure safe operation they do not give any indication of the service life that
can be expected. The lifetime of a power supply is largely determined by the temperature of the
electrolytic capacitors, which have a wear out mechanism. As a general rule, capacitor lifetime can
be doubled for every 10 °C drop in operating temperature.
The graph below indicates the expected service life of the EMA212 power supply based on
measurement of two key electrolytic capacitors.
185000
165000
145000
125000
Lifetime (Hrs)
105000
85000
65000
45000
25000
5000
105 95 85 75 65 55
C14 & 42 Temperature (°C)
76
Thermal Management
• Cooling Power Modules
Typically, power supplies and DC-DC converters have pre-defined thermal ratings for both convection
and forced cooling. However, when power modules are used, additional thermal management is
normally required.
In order to maximize the benefit of thermal management, it is necessary to begin early in the design
stage with an analysis of which components will be subject to high temperatures and what will be
required to cool these devices. Since operating temperature directly affects the lifetime and reliability
of a power supply, it is necessary to obtain an operating temperature that falls below the maximum
permitted operating temperature. The definition of thermal resistance is:
θ = ΔT
Q
where θ is thermal resistance in °C/W

ΔT is the temperature difference between two reference points in °C
Q is the heat flux or power passing through the two points in Watts.
This definition allows the calculation of junction temperatures using a thermal circuit similar to an
electrical circuit:
• Ta
θSA is thermal resistance

Ta
sa
heatsink–to–ambient ˚C/W
Ts
• Ts cs θCS is thermal resistance

•
Tc Tc case–to–heatsink ˚C/W
Thermal resistance to the flow of heat from the power module to the ambient temperature air
surrounding the package is made up of several elements. These are the thermal resistances of the
case–to–heatsink and heatsink–to–ambient interfaces and can be added together to give an overall
thermal resistance from power supply to ambient θCA.
Just as Ohm’s Law is applied in an electrical circuit, a similar relationship is applied to heatsinks.
TC = TA + PD (θCA)
where TC = maximum power supply temperature

TA = ambient temperature
PD = power dissipation
θCA = case to ambient thermal resistance
From this equation, power module temperature may be calculated, as in the following examples.
77
Thermal Management
Example: A power module package must operate in ambient temperature of

+30 ˚C. What is its baseplate temperature? Let PD = 5 W and θCA = 11.0 ˚C/W.
TC = TA + PD θCA = 30 + (5 x 11.0) = +85.0 ˚C
Where operation in a higher ambient temperature is necessary, the maximum power module
temperature can easily be exceeded unless suitable measures are taken.
Example: The same device to be used at an ambient temperature of +50 ˚C,
what is its case temperature?
TC = TA + PD θCA = 50 + (5 x 11.0) = +105.0 ˚C
This exceeds most power module maximum operating temperatures and therefore some means of
decreasing the case–to–ambient thermal resistance is required.
As stated earlier, θCA is the sum of the individual thermal resistances; of these, θcs is fixed by the
design of device and package and so only the case–to–ambient thermal resistance, θCA, can be
reduced.
If θCA, and therefore θJA, is reduced by the use of a suitable heatsink, then the maximum Tamb can be
increased:
Example: Assume that a heatsink is used giving a θCA of 3.0 ˚C/W. Using this
heatsink the above example would result in a baseplate temperature given by:
TC = TA + PD θCA = 50 + (5 x 3.0) = +65.0 ˚C
It should be noted that these calculations are not an exact science. This is because factors such as
θCA may vary from device type to device type, and the efficacy of the heatsink may vary according to
the air movement in the equipment.
Where it is impossible to improve the dissipation capability of the heatsink, forced air cooling can
become necessary, and although the simple approach outlined above is useful, more factors must
be taken into account when forced air cooling is implemented.
• Baseplate Cooling
The use of power supplies in harsh or remote environments brings with it many fundamental design
issues that must be fully understood if long-term reliability is to be attained.
Under these conditions, it is generally accepted that electronic systems have to be sealed against
the elements. This makes the removal of unwanted heat particularly difficult. The use of forced-air
cooling is undesirable as it increases system size, adds the maintenance issues of cleaning or
replacing filters, and the fan being prone to wear out, particularly in tough environments.
A commonly adopted solution is to use a standard power supply and modify the mechanical design
to enable removal of heat from the sealed system. However, this simple compromise does not really
address the fundamental issues of power supply design for the applications described. A more
practical approach is to select a power supply which has been designed specifically for sealed
enclosure applications.
78
Thermal Management
The power supply design has to take into account two main thermal factors.
The extremes of ambient temperature encountered in remote sites can range from -40 ºC to over
+40 °C. It is common for the temperature within the enclosure to rise some 15 to 20 °C above the
external temperature. The positioning of the power supply within the enclosure can help minimize the
ambient temperature in which it operates and this can have a dramatic effect on system reliability. As
a rule of thumb, MTBF (mean time between failures) halves with every 10°C rise in temperature. The
power supply therefore needs to be able to operate from –40 °C to +65 °C as a minimum
specification.
System enclosures are typically sealed to IP65, IP66 or NEMA 4 standards to prevent ingress of dust
or water. Removal of heat from other electronic equipment and power supplies in a situation with
negligible airflow is the challenge. From the power system perspective, the most effective solution is
to remove the heat using a heatsink that is external to the enclosure. However, most standard power
supplies cannot provide an adequate thermal path between the heat-dissipating components within
the unit and the external environment.
Conventional power supplies dissipate heat into small on-board heatsinks or onto a chassis. The
basic construction is shown in Figure 1. Most of the heat is dissipated within the enclosure in which
the power supply is used. Such units typically have to be derated from 50 °C, delivering 50% of their
full rated power at 70 °C. The derating specification is a general guide based on individual
components within the power supply not exceeding their maximum operating temperatures.
Power transistor
& heatsink Inductor Power transistor
PCB
Figure 1
Construction of typical industrial AC-DC power supply
Fundamentally, the successful design of a power supply for use within sealed enclosures relies on
creating a path with low thermal resistance through which conducted heat can be passed from heat-
generating components to the outside world.
The components that generate the most heat in a power supply are distributed throughout the
design, from input to output. They include the power FET used in an active PFC circuit, the PFC
inductor, power transformers, rectifiers, and power switches. Heat can be removed from these
components by mounting them directly onto a substantial base-plate that in turn can be affixed to a
heatsink, rather than on to the PCB. As mentioned earlier, the heatsink is then located outside of the
enclosure.
79
Thermal Management
Power transistor Baseplate of power supply
Inductor
PCB
External heatsink Ambient
Figure 2. Basic construction of baseplate cooled PSU with all of the major heat-generating
components fixed directly to the baseplate
This construction does demand accurate pre-forming of the leads of the components mounted on
the baseplate, and accurate positioning of the PCB with respect to the baseplate but there is no
significant increase in manufacturing complexity or costs.
With the appropriate heatsink, removal of heat can be so effective that there is no need to derate
the unit until the ambient temperature reaches +70°C. This eliminates the need to over-engineer
the power supply for the application.
Dissipating the Heat: Heatsink Calculations
Three basic mechanisms contribute to heat dissipation: conduction, radiation and convection. All
mechanisms are active to some degree but once heat is transferred from the baseplate to the
heatsink by conduction, free convection is the dominant one.
Effective conduction between the baseplate and heatsink demands flat surfaces in order to achieve
low thermal resistance. Heat transfer can be maximized by the use of a thermal compound that fills
any irregularities on the surfaces. System designers should aim to keep thermal resistance between
baseplate and heatsink to below 0.1 °C/W. This is the performance offered by most commonly used
thermal compounds when applied in accordance with manufacturers’ instructions.
Radiation accounts for less than 10% of heat dissipation and precise calculations are complex. In
any case, it is good practice to consider this 10% to be a safety margin.
The degree of convection cooling depends on the heatsink size and type. Heatsink selection involves
the following steps:
1. Calculate the power dissipated as waste heat from the power supply. The efficiency and worst
case load figures are used to determine this using the formula:
1 -1
Waste heat =
{ 1 Eff%
- Eff%
}x P out
or
{ Eff% }x P out
80
Thermal Management
2. Estimate the impedance of the thermal interface between the power supply baseplate and the
heatsink. This is typically 0.1°C/W when using a thermal compound.
3. Calculate the maximum allowable temperature rise on the baseplate. The allowable temperature
rise is simply:
TB – TA where TA is the maximum ambient temperature outside of the cabinet

and TB is the maximum allowable baseplate temperature.
4. The required heatsink is defined by its thermal impedance using the formula:
θH = TB – TA
-0.1
Waste Power
5. The final choice is then based on the best physical design of heatsink for the application that can
deliver the required thermal impedance. The system’s construction will determine the maximum
available area for contact with the baseplate of the power supply and the available space outside of
the enclosure will then determine the size, number and arrangement of cooling fins on the heatsink
to meet the dissipation requirement.
Conclusion
The reliability of remotely-sited electronic equipment is fundamentally dependent upon power supply
reliability. The most cost-effective approach to power system design is to use power supplies
designed for the application, which conduct heat via large, flat baseplates to heatsinks that can be
mounted outside of the enclosure.
81
Reliability
• Terminology
Failure Rate λ
The failure rate is defined as the percentage of units failing per unit time. This varies throughout the
life of the equipment and if λ is plotted against time, a characteristic bathtub curve (below) is
obtained for most electronic equipment.
A B C
Failure Rate
Time
The curve has three regions, A - Infant mortality, B - Useful life, C - Wear out.
In region A, poor workmanship and substandard components cause failures. This period is usually
the first few tens of hours and a burn-in is often employed to prevent these failures occurring in the
field, although burn-in may not get us to the bottom of the curve. This does not stop the failures
occurring, it just ensures that they happen within the manufacturing location rather than at the
customer’s premises or in the field .
In region B the failure rate is approximately constant and it is only for this region that all of the
following analysis applies.
In region C, components begin to fail through reaching end of life, rather than by random failures.
Electrolytic capacitors dry out, fan bearings seize up, switch mechanisms wear out and so on. Well-
implemented preventative maintenance can delay the onset of this region.
Reliability is defined as the probability that a piece of equipment operating under specified conditions
shall perform satisfactorily for a given period of time.
Probability is involved since it is impossible to predict the behavior with absolute certainty. The
criterion for satisfactory performance must be defined as well as the operating conditions such as
input, output, temperature, load etc.
MTBF – Mean Time Between Failures

MTTF – Mean Time To Failure
MTBF applies to equipment that is going to be repaired and returned to service, MTTF to parts that
will be thrown away on failing. MTBF is the inverse of the failure rate and is often misunderstood. It is
often assumed that the MTBF figure indicates a minimum guaranteed time between failures. This
assumption is wrong, and for this reason the use of failure rate rather than MTBF is highly
recommended.
82
Reliability
The mathematics are expressed as follows:

m= 1
λ
R(t) = e-λt = e-(t/m) Where R(t) = reliability
e = exponential (2.178)
m= t λ = failure rate
log(n){ R1 } (t)
m
t
=
=
mtbf
time
This shows that for a constant failure rate, plotting reliability ‘R(t)’ against time ‘t’ gives a negative
exponential curve. When t/m = 1, i.e. after a
1.0
time ‘t’, numerically equal to the MTBF
figure ‘m’, then
.75
Reliability: R(t)
R(t)=e-1=0.37
.50
This equation can be interpreted in a
number of ways:
.25
a) If a large number of units are considered,

only 37% of them will survive for as long 0.0 1.0 2.0 3.0 4.0
as the MTBF figure. Time
b) For a single unit, the probability that it will work for as long as its MTBF figure is only 37%.
c) The unit will work for as long as its MTBF figure with a 37% Confidence Level.
In order to put these numbers into context, consider a power supply with an MTBF of 500,000 hrs
(or a failure rate of 0.2 failures per 1000hrs), or as the advertisers would put it, an MTBF figure of 57
years. Using the above equation, R(t) for 26,280 hours (three years) is approximately 0.95 and if such
a unit is used 24 hours a day for three years the probability of it surviving is 95%. The same
calculation for a ten year period will give an R(t) of 84%. If 700 units are used, on average
0.2%/1000hrs will fail, or approximately one per month.
Service Life Transatlantic

Cable
There is no direct connection or correlation 100 Human
between service life and failure rate. It is

Service Life (Years)
perfectly possible to design a very reliable 10 Car PSU
product with a short life. A typical example is a

missile, which has to be very very reliable
(MTBF of several million hours), but its service 1.0
Toaster
life is only around 4 minutes (0.06hrs). 25-year-

old humans have an MTBF of about 800 years,
(failure rate of 0.1% per year), but not many 0.1 Missile
have a comparable service life! If something
has a long MTBF, it does not necessarily have
3 4 5 6 7 8 9
a long service life. 10 10 10 10 10 10 10 MBTF
83
Reliability
• Factors Affecting Reliability
The most important factor is good, careful design based on sound experience, resulting in known
safety margins. Unfortunately, this does not show up in any predictions, since they assume a
perfect design.
Many field failures of electronic equipment are not due to the classical random failure pattern
discussed here, but to shortcomings in the design and in the application of the components, as well
as external factors such as occasional voltage surges. These may be outside of the specification but
no one will ever know as all that will be seen is a failed unit. Making the units rugged through careful
design and controlled overstress testing is a very important part of making the product reliable.
The failure rate of the equipment depends on these three factors.
Complexity Keep things simple, because what isn’t there can’t fail but, conversely, what isn’t
there can cause a failure. A complicated or difficult specification will invariably result
in reduced reliability. This is not due to the shortcomings of the design staff, but to
the resultant component count. Every component used will contribute to the
equipment's unreliability.
Stress For electronic equipment, the most prominent stresses are temperature, voltage,
vibration and temperature rise due to current. The effect of each of these stresses
on each of the components must be considered. In order to achieve good reliability,
various derating factors have to be applied to these stress levels. The derating has
to be traded off against cost and size implications. Great care and attention to detail
is necessary to reduce thermal stresses as far as possible. The layout has to be
such that heat-generating components are kept away from other components and
are adequately cooled. Thermal barriers are used where necessary and adequate
ventilation needs to be provided.
The importance of these provisions cannot be overstressed since the failure rate of
the components will double for a 10ºC increase in temperature. Note that
decreasing the size of a unit without increasing its efficiency will make it hotter, and
therefore less reliable.
Generic Generic reliability (also known as inherent reliability) refers to the fact that, for
example, film capacitors are more reliable than electrolytic capacitors, wirewrap
connections more reliable than soldered ones, fixed resistors more reliable than
potentiometers. Components have to be carefully selected to avoid the types with
high generic failure rates. Quite often there is a cost trade-off, as more reliable
components are usually more expensive.
84
Reliability
Estimating the Failure Rate
The failure rate should be estimated and measured throughout the life of the equipment.
During the design, it is predicted.

During the manufacture, it is assessed.
During the service life, it is observed.
The failure rate is predicted by evaluating each of the factors affecting reliability for each component
and then summing these to obtain the failure rate of the whole equipment. It is essential that the
database used is defined and used consistently. There are three databases in common use:
MIL-HDBK-217, HRD5 and Bellcore. These reflect the experiences of the US Navy, British Telecom
and Bell Telephone respectively.
In general, predictions assume that the design is perfect, the stresses known, everything is within
ratings at all times, so that only random failures occur; every failure of every part will cause the
equipment to fail and that the database is valid. These assumptions are wrong. The design is less
than perfect, not every failure of every part will cause the equipment to fail, and the database is likely
to be at least 15 years out of date. However, none of this matters as long as the predictions are used
to compare different topologies or approaches rather than to establish an absolute figure for
reliability. This is what predictions should be used for.
Prediction
Parts stress method In this method, each factor affecting reliability for each component is
evaluated. Since the average power supply has over 100 components
and each component about seven factors (stress ratio, generic,
temperature, quality, environment, construction and complexity), this
method requires considerable effort and time. Predictions are usually
made in order to compare different approaches of topologies, i.e. when
detailed design information is not available and the design itself is still in
a fluid state. Under such circumstances it is hardly worthwhile to
expend this effort and the much simpler and quicker Parts count
method is used.
Parts count method In this method, all like components are grouped together, and average
factors allocated for the group. So, for example, instead of working out
all the factors for each of the 15 electrolytic capacitors used there is
only one entry of capacitor with a quantity of 15. Usually only two
factors are allocated, generic and quality. The other factors, including
stress levels, are assumed to be at some realistic level and allowed for
in the calculation. For this reason, the factors are not interchangeable
between the two methods. In general, for power supplies, HRD5 gives
the most favourable result closely followed by Bellcore, with MIL-217
the least favorable. This depends on the mix of components in the
particular equipment, since one database is ‘unfair’ on ICs, and another
on FETs. Hence the importance of comparing results from like
databases only.
85
Reliability
Assessment This is the most useful and accurate way of predicting the failure rate.
A number of units are put on life test, at an elevated temperature, and
so the stresses and the environment are controlled.
During life tests and reliability demonstration tests it is usual to apply greater stresses than normal, so
that the desired result is obtained more quickly. Great care has to be applied to ensure that the
effects of the extra stress are known and proven to be calculable and that no hidden additional
failure mechanisms are activated by the extra stress. The usual extra stress is an increase of
temperature and its effect can be calculated as long as the maximum ratings of the device are not
exceeded.
Prototype Testing
With all the sophisticated computer analysis available, there is still no substitute for thoroughly testing
products or components. One way of doing this would be to perform HALT testing. HALT (highly
accelerated life test) is used to test as many different conditions as possible and cycling the
temperature, input and load independently.
Manufacturing Methods
Suppliers must be strictly controlled and deliver consistently good product with prior warning of any
changes to processes. Because of the supply chain JIT and QA practices this can be achieved by
dealing with a small number of trusted suppliers.
Manual assembly is prone to errors and to some random, unintentional abuse of the components by
operators, such as ESD. This causes defects, which will show themselves later.
Changing settings produces inconsistency and side effects. A good motto is ‘if it works leave it
alone, if it does not, find the root cause.’ There must be a reason for the deviation and this must be
found and eliminated, rather than masked by an adjustment.
The results from the HALT test can be used to set test limits for production screening. Highly
Accelerated Stress Screening (HASS) uses the same equipment as for HALT tests but knowing the
operating and destruct (where possible) limits can be used to screen HALT tested products in
production. This process differs from conventional stress screening in that the climatic and
mechanical stimuli are much higher and consequently the test times are much shorter. HASS can be
summed up as a process used in manufacturing to allow discovery of process changes and prevent
products with latent defects from getting into the field.
86
Reliability
• System Reliability
There are two further methods of increasing system reliability.
More reliable components MIL standard or other components of assessed quality could be used
but in industrial and commercial equipment this expense is not normally
justified.
Redundancy In a system where one unit can support the load and two units are used
in parallel, the system is much more reliable since the system will still
work if one unit fails. Clearly, the probability of both units failing
simultaneously is much lower than that of one unit failing.
Redundancy has a size and cost penalty so normally an n+1 system is used, where n units can support
the load, but n+1 units are used in parallel, 2+1 or 3+1 being the usual combinations. Supposing the
reliability of each unit under the particular conditions is 0.9826, the system reliability for an n+1 system
where n=2 would be 0.9991, an improvement of 20 times. (Nearly 60 times in a 1+1 system).
There are downsides to this approach. More units, higher cost and the need for faulty units to be
brought to the operator’s attention so that they can be replaced, changing units must not make the
system fail (hot swap). The extra circuitry required to monitor all aspects and ensure reliability in itself
increases the failure rate and cost of the system (see page 58 for more details on redundant
operation).
Comparing Reliability
When comparing reliability figures, the following points must be satisfied.
• The database must be stated and must be identical. Comparing a MIL-HDBK-217F prediction with
a MIL-HDBK-217E prediction or an HRD5 prediction is meaningless as there is no correlation.
• The database must be used consistently and exclusively. The result is meaningless if a different
database is used for some components.
• The external stresses and environment must be stated and be identical. (input, load, temperature
etc). The result is meaningless if all the environmental details are not stated or are different.
• The units must be form-fit function interchangeable. If, for example, the ratings are identical, but
one needs an external filter and the other does not then there is no comparison (although you
could work out the failure rate of the filter and add it to the failure rate of the unit).
There is no magic; if one manufacturer predicts 200,000 hours and another states 3,000,000 hours
for a comparable product, then they must have used a different database, a different stress level
or a different environment.
87
Legislation
• Power Supply Safety
Legislation requires electrical equipment to be designed to reduce the likelihood of injury or damage
due to:
• Electric shock • Energy related hazards

• Fire • Heat-related hazards
• Radiation • Chemical hazards
A safe power supply is an inherent part of any electrical product and must comply with the relevant
safety standard. There are several standards which could be used for power supplies and the
decision on which to use depends on the intended application of the end product.
In future there will be an international product specific standard for power supplies which will need to
be used to demonstrate compliance with the safety requirements and this will be part of the
IEC61204 range of standards. At time of writing, the particular standard relating to low voltage DC
power supplies has not been published in the European Official Journal (OJ) which means it cannot
be used.
Instead, one of the product family standards must be used which are application dependant. Most
power supplies will use an information technology equipment standard (60950), a medical equipment
standard (60601) or less commonly a standard for equipment used for measurement, control and
laboratory use (61010). This latter standard covers equipment intended for professional, industrial
process and educational use such as equipment for testing or measuring non-electrical quantities,
controlling output quantities to specific values or laboratory equipment which measures, analyses or
prepares materials.
Another standard which is sometimes used in conjunction with one of the above is UL508C which
covers industrial equipment intended to power control systems for electrical motors. It is common for
DIN rail power supplies to have approval to this standard.
The standard for information technology equipment, IEC60950, covers a wide range of product
types and is commonly used. Approvals are separately granted by a number of national test
laboratories depending on the target markets. UL (Underwriters Laboratories, UL60950) are
commonly used for approvals in North America, CSA (Canadian Standards Association, CSA22.2
No.60950) for Canada and there are a number of European test laboratories which will grant
approval for EU wide use, EN60950. UL & CSA also operate a scheme to grant approvals for both
markets.
In the major Asian markets other approvals are required. The rules are essentially as laid out in
IEC60950 with some additional testing in some instances, including EMC.
China CCC (China Compulsary Certification)

Taiwan BSMI (Bureau of Standards, Metrology & Inspection)
Japan PSE (Product Safety Electric Appliance & Materials)
Korea KETI (Korean Electrical Testing Institute)
There are many other approval bodies in existence which may need to be considered depending on
the equipment target markets.
88
Legislation
Electrical Safety
An electrically safe system relies on the use of safety earthing, the insulation of hazardous voltage
and the controlling of leakage currents.
Insulation
The five different types of insulation grades are listed below.
Operational/functional Insulation that is necessary only for the correct functioning of the
insulation equipment and does not provide any protection against electric shock.
Basic insulation Insulation applied to live parts to provide basic protection against
electric shock.
Supplementary insulation Independent insulation applied in addition to basic insulation in order to

provide protection against electric shock in the event of a failure of
basic insulation.
Double insulation Insulation comprising both basic insulation and supplementary insulation.
Reinforced insulation Single insulation system applied to live parts which provides a degree
of protection against electric shock equivalent to double insulation.
Creepage and clearance spacing specified in the safety standard must also be met. The requirement
depends on the insulation type, working voltage and pollution degree. The insulation barriers must
then undergo a high voltage test.
Clearance is Creepage is
the distance the distance
between two between two
points through points over
air the surface
Earthing/Grounding
The two types of earth are listed below:
Functional earth This does not provide any safety function, for example the screen on
an external psu output lead.
Protective earth This provides protection against electric shock in a class I system and
must meet certain performance criteria, such as resistance.
89
Legislation
Earth Leakage Current
Current that flows down the earth conductor is defined as earth leakage current. To prevent the risk
of electric shock in the event of the earth becoming disconnected, the maximum value is defined in
the safety standard under touch current and is normally 3.5mA for pluggable equipment. Higher
values are permissible if the equipment is permanently connected. Within the power supply the main
contributor to the leakage current is normally the EMC filter Y capacitors.
L
Primary
N
PE Y Capacitors
Earth Leakage
Current
Class I Systems
Class I systems rely on earthing and insulation to provide a means of protection. In the event of the
basic insulation between live and earth failing the protective earth provides a path for the fault current
to flow, causing a fuse or circuit breaker to trip. The diagram below shows the insulation diagram of
a class I power supply.
Double or Reinforced
Insulation
L
Primary Secondary
N
PE Basic Insulation
PE
90
Legislation
Class II Systems
Class II systems rely on insulation only to protect against electric shock. The diagram below shows
the insulation diagram of a class II power supply.
Insulation
L
Primary Secondary
N
Insulation
PE
• Medical Safety
IEC60601-1 is the generic safety standard for medical equipment used within the patient vicinity.
EMC is considered as safety critical within a medical device. The EMC requirements and tests are
also included within part 1 of the standard. Product specific standards that relate to a particular type
of product are covered under part 2 of the standard.
The principles of a safe system are the same as those applied to non-medical equipment meeting
IEC60950. The differences relate to creepage and clearance distances, test voltages and leakage
current where the requirements are more stringent. There are also additional requirements to
consider if direct contact is to be made to the patient.
Approval to these standards and the national deviations are granted by the same bodies as for
industrial and IT equipment certified to IEC60950.
Leakage Currents
There are four different types of leakage current. Earth leakage, enclosure leakage and patient
leakage current are all determined by the power supply performance.
Earth Leakage Current flowing in the protective earth conductor.
Enclosure Leakage Current flowing to earth via the patient or operator from the enclosure.
Patient Leakage Current flowing to earth via the patient from the applied part.
Patient Auxiliary Current flowing between two applied parts.
91
Legislation
Applied Parts
There are three different types of applied parts. An applied part is a part that in normal use will come
into contact with the patient or needs to be touched by the patient. These are classified as B, BF or
CF. For BF and CF applied parts an additional level of isolation is often required between the output
of the power supply and the patient to meet the patient leakage limits and ensure that the patient is
isolated from earth.
Leakage Current Type B Type BF Type CF
NC SFC NC SFC NC SFC
Earth Leakage Current* 500µA 1mA 500µA 1mA 500µA 1mA
Enclosure Leakage Current 100µA 500µA 100µA 500µA 100µA 500µA
Patient Leakage Current 100µA 500µA 100µA 500µA 10µA 50µA

NC = Normal Conditions SFC = Single Fault Conditions
*US earth leakage current 300µA Figures quoted are for portable equipment
Below is a typical isolation diagram for a power supply meeting the requirements of a BF and CF
applied part. Isolation barrier 1 is contained within a standard 230VAC - 12VDC power supply.
Isolation barrier 2 is contained within a 12V - 48V DC/DC converter.
Signal Ports
Live
MP LP AP
Neutral (230) (12) (48)
B (230) D/R (230) D/R (60) D/R (48)

OP B(230) B(230)
Patient
AP = Applied Part
PE Isolation Isolation
B(xx) = Basic insulation (working
Barrier 1 Barrier 2
voltage)
D = Double insulation
Typical medical device with patient connection LP = Live part
MP = Mains part
OP = Operational insulation
R = Reinforced insulation
92
Legislation
• Electromagnetic Compatibility (EMC)
EMC is a way of describing how pieces of electrical and electronic equipment interact with each
other when they act as either sources or receivers of noise. These two types of interaction are
described as emissions and immunity.
Emissions
Emissions are electrical noise generated by the power supply or its electronic load and transmitted
along the input and output cables as conducted noise or from the outer casing & cables as radiated
noise. If left unchecked electrical noise could interfere with the correct and safe operation of nearby
electrical equipment and it is therefore a requirement to restrict the amount of noise generated. The
EMC directive was introduced in Europe in 1992 (89/336/EEC) with the aim of imposing limits on the
amount of noise that equipment can emit. In the USA, the limits are set by the FCC (Federal
Communications Commission). VCCI (Voluntary Control Council for Interference by Information
Technology Equipment) limits are the Japanese equivalent. In Asia the CISPR and FCC standards are
widely accepted by the various approval bodies.
Conducted Noise
Conducted noise is that which travels along physical routes between pieces of equipment. We
usually think of these paths as being the mains cables which can transmit noise generated by one
piece of equipment along the mains supply (within an installation, a single building or even separate
buildings) and which can then affect other pieces of equipment connected to the same mains
system, or as the cables which directly connect one piece of equipment to another, such as DC
cables or signal and control wires.
The noise takes one of two forms according to whether it is common to the ground system or exists
between differing parts of the electrical circuit.
Common mode noise exists within different parts of the circuit and is common to the ground plane.
On the mains input to a piece of electrical equipment it can be measured between the line conductor
and the earth conductor, or between the neutral conductor and the earth conductor. Differential
mode noise exists between parts of the circuit with different potentials. On the mains input to
electrical equipment it can be measured between the line conductor and the neutral conductor.
X~~~
Line
Common Differential
mode noise mode noise
Neutral
Common
mode noise
Earth
93
Legislation
Differential Mode Noise
Differential mode noise is primarily generated by rapid changes in current.

Within a switch mode power supply, the primary circuit is opened and
closed by means of a switching device such as a BJT or MOSFET.
The current flowing through the circuit therefore goes through a
continuous cycle of changing from a maximum value to zero and vice
versa as the switch opens and closes. The rate of current change is
very fast, perhaps in the order of 50ns, and if the primary current was I
in the order of 1A, the change would be 1A in 50ns or put another way,
20 million A/s. The impedance of the printed circuit traces will be
significant at current changes of this magnitude and unwanted voltages
will be generated along the traces in the form of noise.
Common Mode Noise
Common mode noise is primarily generated by changes in voltage.

The same switching device which is breaking the current in the
primary circuit is also breaking a voltage. The voltage could be as
high as 600V and this may be being interrupted in the order of 50ns
meaning that there could be a voltage change rate of 12V/ns or
12,000 million volts per second. The unwanted capacitances found
around the switching element, for example between its case and the V
heatsink to which it is attached will be significant at these levels of Parasitic

capacitance
voltage change and significant voltages in the form of noise will be
generated.
Radiated Noise
Electrical noise can radiate from the enclosure or casing of the equipment and from its connecting
cables. It will escape through the seams, ventilation slots, display areas and so on and travel in any
direction through the air. In order to successfully propagate through air, the wavelength will be shorter
than for conducted emissions meaning that frequencies will be higher. While conducted emissions
are measured up to a frequency of 30MHz, radiated emissions are measured up to 1GHz.
Standards
In the US, EMC standards are written and enforced by the FCC. FCC 20870 covers both radiated
and conducted noise. The FCC standard is harmonized with CISPR standards, and these are
sometimes used instead to show compliance.
In Europe, the EMC directive does not define what the required levels are which need to be met so
we must rely on international standards. There are three different types published. Product-specific
standards define the allowable EMC performance of particular types of product. If a product-specific
standard exists, then it MUST be used. Where a type of equipment doesn’t have an associated
product standard, generic standards can be used. As the term generic suggests, they contain
requirements which cover many types of equipment and therefore some of the tests listed cannot be
relevant or even adhered to. The product specific and generic standards refer to basic standards.
These are the ones which define the exact test set up as well as the limits allowed. In Asia the
CISPR and FCC standards are widely accepted.
94
Legislation
For power supplies, the product-specific standard, IEC61204-3, will take precedence over the generic
standards. For emissions, it defines the following basic standards:
CISPR22 for conducted emissions (maximum of level B)

CISPR22 for radiated emissions (maximum of level A)
IEC61000-3-2 for harmonic currents
IEC61000-3-3 for voltage flicker
Sometimes there are other basic standards which need to be applied. For example, EN55014 is
applicable to motor operated household equipment, CISPR11 is applicable to industrial, scientific and
medical equipment. These basic standards will be called into use by product family standards which
may be applicable to end user equipment.
Methods of Measurement
Noise measurement techniques are defined by the relevant basic standard. The techniques will be
generally similar whether it is an IT standard such as CISPR22 which is applicable or a military
standard such as MIL 461 or DEF STAN 59-41.
Conducted Noise
Conducted noise values will largely be dependent upon the local impedance of the mains system at
the location at which the measurement is being done. Mains impedances will vary throughout a
network and they could be vastly different throughout the world. A Line Impedance Stabilization
Network (LISN), also known as an Artificial Mains Network (AMN) is used to give a defined mains
impedance to the measurement system of 50 Ohms. In the case of the IT standard CISPR22, the
noise will be measured from 150kHz to 30MHz and two readings must be taken. These are a quasi
peak measurement and an average measurement. Both must be under their respective limit lines in
order for the equipment to pass.
Radiated Noise
The services of a dedicated test house will normally be required to measure radiated noise. This is
because the test should be performed on a large area known as an Open Area Test Site (OATS) which
will not only be free of reflecting surfaces but will also be calibrated so that the influence of any
reflections from far away is known as the reflections will either add to the original signal, or detract
from it depending upon the phase shift of the reflection. The measuring equipment will consist of an
antenna which will feed into a receiver. The emissions from all sides of the equipment must be taken
and for each face the antenna will be moved between heights of one and four meters to obtain the
worst case reading. In addition to this, the antenna will be positioned with its elements alternately
horizontal and vertical, again to obtain the worst case reading.
As this setup is impractical for most companies, alternative techniques are normally used to give an
indication of the radiated emissions. This may consist of using near field probes to ‘sniff’ around the
enclosure of the equipment or using conducted emission techniques to measure at frequencies into
the hundreds of MHz band. This is a relevant test as it will often be the cables themselves that are the
source of the radiation.
95
Legislation
EMC Filtering
L
A power supply or DC/DC converter will have an in-built input filter to
reduce the conducted emissions. It will have two parts; one to reduce Y
the common mode noise, the other to reduce the differential mode
noise. Common mode noise can be reduced by use of Y capacitors
between line and ground and another one between neutral and ground Y
in conjunction with a common mode inductor.
N
L
Differential mode noise can be reduced by use of an X capacitor
between the line and the neutral in conjunction with a differential mode
inductor. In some instances the differential mode inductor is formed
X X
from the leakage inductance of the common mode inductor so that
there is only one visible wound component.
L
When combined the resulting filter may look like this:
Y
X X
Sometimes the built-in filter will give an inadequate performance for a given application. This may be
where the power supply is designed to meet the lesser requirements of an industrial environment but
is being used in the more stringent light industrial or residential environment. Perhaps several power
supplies are being used in a single piece of equipment and the resulting emissions must be reduced,
or perhaps noise from the load itself is being coupled into the input of the power supply. In all these
instances some form of external filtering will be required.
Filter Selection
There are some basic steps to follow when choosing a filter, some of which are straightforward and
others less so.
Mechanical format Is the filter going to be mounted within the equipment where it can be
fixed to a panel or should it also provide the extra functions of being the
mains input connector and perhaps contain an on/off switch? If it is the
former, a chassis mount filter can be used. These will generally have
faston terminals for easy connection but may also come with flying
leads. IEC inlet filters can have built-in on/off switches and even fuse
holders. They can be mounted by either screwing them down to the
equipment or by use of self locking lugs. Generally, for metal chassis
equipment, the bolt-down variety will provide a lower impedance earth
path for the circulating noise down to ground.
96
Legislation
Input current The filter should be able to pass the maximum working current of the
equipment so as not to overheat but generally the lower the current
capacity within a filter series, the higher its filtering performance.
Attenuation required A filter will be required to reduce the noise at certain frequencies. By
how much and at which frequencies is information which will not readily
be known without having first performed a conducted noise
measurement. Filters have differing amounts of attenuation and, for a
given current rating, the higher the attenuation the larger the filter. As
there will be a practical limit on the size of filter components, large
amounts of attenuation will require the use of multi-stage filters.
Immunity
Immunity is concerned with how a piece of equipment will behave when subject to external electrical
or magnetic influences in the form of noise. The noise will exist as either conducted or radiated noise
and will be from natural sources such as lightning, electrostatic build up or solar radiation or may be
from man made sources such as radio or mobile phone transmissions, commutation noise from
electrical motors or emissions from power supplies and other switching devices.
Conducted Immunity Phenomena
A power supply or piece of electrical equipment will be subject to conducted noise either via the
mains connection, a DC output or via the signal and control lines. The noise could take various forms
from brown-outs of the mains, to single short duration but high voltage spikes, to RF frequency noise
coupled into the cables and conducted into the equipment.
Radiated Immunity Phenomena
Noise can also directly enter a system via the air in the form of electrical or magnetic fields. The field
is picked up by the cables attached to a piece of equipment or by the internal PCBs themselves and
can be in the form of electromagnetic fields generated by a mobile phone or the magnetic field
generated from a nearby transformer.
Standards
The product standard for power supplies, EN61204-3, lists all of the basic immunity standards that
are applicable to a power supply. These are listed below. For each type of test there are two
important factors: the test severity level and the performance criteria which defines how the
equipment operates while the test is being carried out.
Performance criteria A There is no change in operating status of the equipment. For a power
supply this means that it will continue to operate within specification and
no signals will change state.
97
Legislation
Performance criteria B There is a loss of function while the test is being applied, but when the test
stops, the operating parameters automatically return to normal. For a
power supply, this means that the output may go out of regulation and
signals may change state but only during the test.
Performance criteria C There is a loss of function while the test is being applied and a manual reset
or intervention is required to restore the original operating parameters.
Electrostatic discharge: IEC61000-4-2
There are three types of test specified in the standard; contact discharge, air discharge and
discharge onto a coupling plane. The test is to simulate the effect of a person charging themselves
up (to many kV) and then touching either the equipment directly or adjacent equipment which could
in turn affect the equipment’s behavior. For open frame power supplies, this test is not normally
applicable but for other power supplies, the pass conditions are ±4kV for contact discharge and
±8kV for air discharge and coupling plane discharge, all with minimum performance criteria B.
RF electromagnetic field: IEC61000-4-3
This test simulates the fields given off by mobile phones and DECT phones. The field is generated by
a sweeping signal generator with a 1kHz modulation function. The signal is amplified and radiated
using an antenna. The field strengths are high enough and in the frequency band (80MHz to 1GHz)
to prevent local radio and TV stations and more importantly emergency services communications
from working so the test must be performed in a screened chamber. For power supplies intended to
operate in a light industrial or residential environments, the field strength is 3V/m but for industrial
power supplies the required field strength is 10V/m. Minimum performance criteria is B in both cases.
Electrical fast transients: IEC61000-4-4
This test is to simulate switching transients generated by motor or solenoid activation or perhaps
from fluorescent lighting. The pulse is very short, only 50ns with a 5ns rise time and is applied
between the two lines and the earth. Generally, the test is only applied to the AC input as the DC
lines and the signal and control lines on a power supply are normally too short. For power supplies
intended to operate in light industrial or residential environments the pulse is ±1kV but for industrial
power supplies the required pulse is ±2kV. Minimum performance criteria is B in both cases.
Voltage surge: IEC61000-4-5
This test is to simulate the effects of a lightning strike. The duration and energy content of the pulse
are much greater than for the EFT test with the duration being 50µs with a 1.2µs rise time. The pulse
is applied between each line and earth and also between lines themselves. For power supplies the
pulse is ±2kV common mode, ±1kV differential, with a minimum of performance criteria B in both
cases.
RF conducted: IEC61000-4-6
This test is similar to the RF radiated electromagnetic field test and must be applied under similar
conditions within a screened chamber though the frequency range is 150kHz to 80MHz. For power
supplies intended to operate in a light industrial or residential environments, the coupled noise is
3Vrms but for industrial power supplies the coupled noise is 10Vrms. Performance criteria B is the
minimum applicable in both cases.
98
Legislation
Voltage dips and interruptions: IEC61000-4-11
A voltage dip represents the brown-out conditions experienced from time to time on the power grid,
while a voltage interruption represents a complete black out condition. There are 3 parts to the test;
a 30% dip for 10ms with minimum performance criteria B, a 60% dip for 100ms with minimum
performance criteria C and a >95% interruption for 5 seconds with minimum performance criteria C.
• CE marking
CE marking within Europe was established as a means of identifying a product as meeting all the
relevant European directives. These directives have been introduced as a way of allowing free trade
within the EU member states as individual members are no longer allowed to prevent trade on
technical grounds. By displaying the CE mark, the product is identified to customs and border
controls as complying with the necessary directives. There are many directives which are applicable
for CE marking and these include:
- Low voltage equipment - Simple pressure vessels

- Toys - Construction products
- Electromagnetic compatibility - Machinery
- Personal protection equipment - Non automatic weighing machines
- Active implantable medical devices - Gas appliances
- Hot water boilers - Civil explosives
- Medical devices - Potentially explosive atmospheres
- Recreational craft - Lifts
- Refrigeration appliances - Pressure equipment
- Telecommunications terminal equipment - In vitro diagnostic medical devices
- Radio & telecommunications terminal equipment
For power supplies, only two directives are applicable, the Low Voltage Directive (LVD) and the
EMC directive.
Low Voltage Directive (LVD) 2006/95/EC
This is applicable to equipment designed for use with a voltage rating of between 50 and 1000 VAC
and between 75 and 1500 VDC. The directive itself does not define how to comply with it but by
conforming to one of the relevant standards, such as the IT safety standard EN60950, compliance is
demonstrated. The route to compliance is by generating a Technical Construction File (TCF) which
includes the following:
- General description of the electrical equipment

- Conceptual design and manufacturing drawings
- Description and explanation of these designs and drawings
- Listing of the product standards used as safety reference
- Results of design calculations and examinations
- Test reports
99
Legislation
Technical File (Annex IV.3) EC declaration of conformity

Manufacturer establishes (Internal production control -
MANUFACTURER the technical documentation Annex III B). Manufacturer
covering the design, ensures and declares
manufacture and operation compliance of manufactured
of the electrical equipment products with the technical
documentation and with
directive requirements.
Module A
The file must be kept within Europe by the manufacturer or importer of the equipment and be
available for inspection within 48 hours. For power supplies, it is common for the TCF to be based
on a competent body test certificate such as TUV.
Electromagnetic Compatibility Directive 89/336/EEC
MANUFACTURER
Other than
Radio Transmitters Type of product
Radio Transmitters
Application of
harmonised standards
Non-application or
EC Type examination
partial application of Full application
(Art 10.5) Module B(*)
harmonised standards
Technical construction
Self certification
file (Art. 10.2)
(Art. 10.2). Module A(*)
by manufacturer
EC declaration of Technical report or EC declaration of

conformity with type certificate by a competent conformity by
approved. Module C(*) body (Art. 10.2) manufacturer
(*) These procedures were approved before the adoption of Council Decision 90/683/EEC (as amended by Decision 93/465/EEC)
on conformity assessment procedures (modules). Their provisions may therefore not be identical to those of the modules.
100
Legislation
This directive is applicable to apparatus liable to cause electromagnetic disturbance or the

performance of which is liable to be affected by such disturbance. Again, the directive does not state
how compliance should be achieved, but there are two routes to compliance. The first is the
standards route whereby the product is tested against either product specific or generic standards.
The second is the technical construction file route. This would be chosen where a piece of
equipment may be too large to undergo testing, or it may be that some of the tests are just not
relevant. The arguments for this would be laid down in the TCF which would be assessed and
signed off by a competent body.
The CE mark should be accompanied by a declaration of conformity. This will list:
- Name & address of the manufacturer

- Description of the equipment
- Reference to harmonized product standards used
- Name of the signatory and their position
- Last two digits of the year in which the CE marking was affixed
This is a self declaration that the manufacturer (or person who places the product on the market in
the EU) has taken all necessary steps to ensure compliance with all the relevant directives.
The two directives treat power supplies in different ways. The LVD states that all apparatus, which
includes open frame power supplies as well as externals, should be CE marked. However, the EMC
directive states that components, which includes open frame power supplies must not be CE
marked. Therefore, the CE mark on an open frame power supply is showing that it complies only to
the LVD, whereas the CE mark on an external power supply is showing that it complies with both
the LVD and the EMC directive.
• No Load Power Consumption and Efficiency of External Power Supplies
Two important reasons for wanting to control the power taken by an external power supply are
continuity of the energy supply and reduction of environmental impacts. Targets are given for
external supplies because of the quantity in use. They normally do not have an off button and they
are commonly left plugged into the AC supply.
At the time of writing there are around 1.5 billion external power supplies in the US which account for
6% of the national electric bill and it is estimated that left unchecked this could rise to 30% by 2010.
In 1992 the US Environmental Protection Agency (EPA) started a voluntary program to promote
energy efficiency and reduce pollution which became the Energy Star program. The California Energy
Commission (CEC) declared that these requirements would be mandatory and from 1st July 2006
limits applied to applicable external power supplies. Washington State and Arizona State also now
have mandatory requirements and it is expected that other States will follow. These requirements are
aimed at consumer appliances and therefore it is single output external power supplies which are
used with laptop computers, mobile phones, printers, print servers, scanners, PDAs and digital
cameras which are applicable. From July 2007 the scope is increased to include other applications.
It is not expected that industrial or commercial applications will require the limits to be met from a
legal point of view though they may do from a marketing stand point. This position may change over
time.
101
Legislation
In Europe, security of energy supply is paramount as energy reserves are used up. Reducing the no
load power consumption of external supplies is expected to save between 1-5 TWh per year by
2010 and defining the minimum active mode efficiency will save another 5 TWh per year. In 2005 the
EU drafted a Code of Conduct (C of C) on external power supplies which is a voluntary agreement
limiting standby and active mode power losses. The active mode efficiency is closely based on the
Energy Star requirements however the no load power consumption limits are more stringent than the
US requirements.
On July 6th 2005 the EU released the Energy Using Products (EuP) Directive. It is currently expected
that this will come into force in 2008 at which point there will be mandatory limits imposed. It should
be noted however that this is applicable to high volume products i.e. ones that would ship more than
200,000 pieces per year.
In all cases, the limits will become more severe over time. In Europe the C of C limits changed on 1st
Jan 2007, in the US the Energy Star and CEC limits change on 1st Jan 2008.
The following tables show the limits imposed by the three bodies. The average efficiency is taken as
the mean of individual efficiencies at 25%, 50%, 75% and 100% loads
Summary of Limits
No load power limits

Rated power No load consumption
0W to <10W 0.5W
≥10W to 250W 0.75W
Active mode power limits
Rated power Average efficiency
0W to 1W 0.49 x Rated power
>1W to 49W ≥[0.09 x Ln(Rated power)] + 0.49
>49W ≥0.84
Energy Star - Jan 1st, 2005 limits

0W to <10W 0.3W
≥10W to 250W 0.5W
0W to 1W Not yet known
>1W to 49W Not yet known
>49W Not yet known
Energy Star - Jan 1st, 2008 limits
102
Legislation

0W to <10W 0.5W
≥10W to 250W 0.75W
>49W ≥0.84
California Energy Commission - July 1st, 2006 limits

0W to <10W 0.3W
≥10W to 250W 0.5W
>49W ≥0.85
California Energy Commission - Jan 1st, 2008 limits
EU Code of Conduct

0.3W to <15W 0.3W
≥15W to <50W ≤0.5W
≥50W to <60W ≤0.75W
≥60W to <150W ≤1.00W
0W to <1.5W ≥0.3
≥1.5W to <2.5W ≥0.4
≥2.5W to <4.5W ≥0.5
≥4.5W to <6W ≥0.6
≥6W to <10W ≥0.7
≥10W to <25W ≥0.8
≥25W to 150W ≥0.9
Jan 1st, 2005 limits
103
Legislation

0.3W to <60W ≤0.3W
≥60W to 150W ≤0.5W
Rated power Efficiency (1)
0W to <1W 0.49 x Rated power
>1W to <49W ≥[0.09 x Ln(Rated power)] + 0.49
>49W ≥0.84 (2)
Jan 1st, 2007 limits
Notes
(1) Readings can be taken at 100% load or, to be in full harmonisation with Energy Star requirements
are an average of 4 points of load (25%, 50%, 75% and 100%).
(2) Externals above 75W having a PFC circuit have a 4% allowance i.e. min efficiency can be
reduced to 80%.
Measurement Technique
The US EPA has devised a procedure for measuring the no load power consumption and active
mode efficiency of external supplies. This procedure has been adopted as an acceptable test
method to demonstrate compliance with Energy Star, California Energy Commission and the EU
Code of Conduct. The document can be found on the www.energystar.gov website and is titled ‘Test
Method for Calculating the Energy Efficiency of Single-Voltage AC-DC and AC-AC Power Supplies’.
This document sets out a standardized test method including test room conditions, accuracy of
measuring instruments, quality of applied mains voltage and accuracy of load conditions. The
document also details the information that is required for the test report.
Marking Requirements
For the EU’s C of C there is currently no requirement to mark the product. However, when the EuP
Directive becomes mandatory, the CE mark applied to the product will indicate that it is compliant
with all applicable Directives and this will include the no load power and active mode efficiency
requirements. To demonstrate compliance with the Energy Star and CEC requirements a mark must
be placed on the product. Before this can be done, the product details and test results must be
submitted to the EPA. The mark is made up of a Roman numeral and should now be a minimum of
III to show compliance with July 1st 2006 limits or a IV to show compliance with the January 1st
2008 requirements.
Summary
Whilst we believe that the customers and applications that utilize XP’s external power supplies do not
require them to be compliant with these requirements, as an environmentally aware company we
have plans to introduce compliant product as soon as reasonably possible and will have most of the
power range covered by the end of 2007.
104
Legislation
• Military EMC and Immunity Standards
For power supplies operated in a military environment there are standards maintained by government
or international organisations. Examples of EMC specifications are:
US Department of Defense: MIL-STD 461 C/D/E

UK Ministry of Defence: DEF STAN 59-41
French Military: GAM-EG 13B
NATO: AECTP500 (proposed replacement for MIL-STD 461E)
Many countries use the MIL-STD series of standards, maintained by the US Department of Defense,
but have national deviations covering specific conditions or equipment used by their armed forces.
Military EMC standards are typically organized by: Service (Air Force, Army, Navy etc.); environment
(e.g. above or below decks); test details, equipment and limits.
Military EMC standards typically contain both immunity and emissions limits for both conducted and
radiated noise. The measurement techniques differ from commercial specifications.
For example: CE (Conducted Emissions)

MIL-STD 461C Measurement made in dBµA using a current probe.
MIL-STD 461D/E Measurement made in dBµV using a LISN.
DEF STAN 59-41 Measurement made in dBµA using a LISN and current probe.
Example of Immunity and Input Voltage Specifications
US Department of Defense: aircraft: MIL-STD 704 A-F, military vehicles: MIL-STD-1275 A/B
UK Ministry of Defence: military vehicles, naval vessels and aircraft: DEF STAN 61-5
Typically, military standards such as MIL-STD-1275 A do not state pass or fail criteria for the power
system this is up to the user to define. For example, a power supply being damaged by a surge
voltage would generally be deemed a failure, but a power supply showing higher levels of output
ripple during a conducted susceptibility test will be deemed to have passed or failed by the tester.
The main susceptibility tests are: Abnormal operating voltage such as generator-only or emergency
power; surges, spikes and dropouts, usually defined as a Volt second product with differing source
impedances and input ripple, usually defined in amplitude across a wide range of frequency. The
tables on the following page outline the input voltage variations specified in MIL-STD-704 and
1275 A.
105
Legislation
Aircraft 28 VDC Steady State 28 VDC DC

Power Cat High Transients Low Transients power
Systems Normal Abnormal Emergency (max) sec
Voltage Time Voltage Time
A 24.0-28.5 23.5-30.0 17.0-24.0 80 0.05 10 0.05 0.05-7.0
MIL-STD-704A B 23.0-28.5 22.5-30.0 16.0-24.0 80 0.05 8 0.05 0.05-7.0
C 22.0-29.0 21.5-30.0 15.0-24.0 80 0.05 7 0.05 0.05-7.0
MIL-STD-704B 22.0-29.0 20.0-31.5 18.0-29.0 50 0.01 18 0.015 5.0-7.0
MIL-STD-704C 22.0-29.0 20.0-31.5 16.0-29.0 50 0.01 18 0.015 5.0-7.0
MIL-STD-704D 22.0-29.0 20.0-31.5 16.0-29.0 50 0.01 18 0.015 7.0
MIL-STD-704E 22.0-29.0 20.0-31.5 18.0-29.0 50 0.01 18 0.015 7.0
Normal Gen. Only Batt. Only
28 V vehicle systems
25.0-30.0 23.0-33.0 20.0-27.0 100.0 0.05 15.0 0.5
MIL-STD-1275A(AT)
Aircraft 270 VDC Steady State 270 VDC

Power High Transients Low Transients
Systems Normal Abnormal Emergency
MIL-STD-704B 250-280 240-290 475 0.01 125 0.05
MIL-STD-704C 250-280 240-290 475 0.01 125 0.05
MIL-STD-704D 250-280 245-285 240-290 475 0.01 125 0.05
MIL-STD-704E 250-280 240-290 240-290 475 0.02 200 0.01
Aircraft 400 Hz AC Steady State AC Transients DC

Power Cat High Transients Low Transients power
Systems Normal Abnormal Emergency (max) sec
A 110-118 104-124 106-122 180 0.1 64 0.05 0.05-7.0
MIL-STD-704A B 108-118 104-124 104-122 180 0.1 58 0.05 0.05-7.0
C 104-118 98-124 100-122 180 0.1 48 0.05 0.05-7.0
MIL-STD-704B 108-118 100-125 102-124 180 0.01 80 0.01 7.0
MIL-STD-704C 108-118 100-125 104-122 180 0.01 80 0.01 7.0
MIL-STD-704D 108-118 100-125 104-122 180 0.01 80 0.01 7.0
MIL-STD-704E 108-118 100-125 108-122 180 0.01 80 0.01 7.0
106
Technology Editorial 1
• An Innovative Topology for
Configured Power Supplies
Contents
• Introduction
• A different approach
• Taking the heat
• Design flexibility
• What can you do with it?
• There can be benefits to OEMs’ end customers too
• Summary
• Introduction
Power supplies are traditionally the last sub-system to be specified and the first to be needed for
system prototype testing. They are also frequently impacted by late changes elsewhere in the
system, whether by the need for more current, or even for additional voltages. Designers have
traditionally struggled to balance project budgets and the need for fast time-to-market, where
custom designs become inappropriate.
As companies focus on core competences designing power supplies in-house has become less
attractive, even when using building-block modules or combining multiple self-contained units. What
is needed is a flexible topology that can be configured to customer requirements, using off-the-shelf
tested and proven modules, and shipped as a complete sub-system.
This editorial explores an alternative approach to this problem that simplifies configuration while
assuring full agency approvals, even for medical equipment. The result is not a kit of parts but a
complete, configured and tested sub-system using versatile packaging to accommodate up to
fourteen outputs and with a wide mix of output voltages and currents. This production-ready
topology also enables maximum benefit to be gained from combining innovative design with low cost
module manufacturing in Asia and local final assembly.
• A different approach
Using DC/DC modules as building blocks can produce configurable power supplies but in many
applications it is issues of heat, mechanics, control and monitoring that need customization. It’s just
these aspects that designers have to sort out for themselves. Although standard brick DC/DC
modules handle raw power conversion, design work is still required to create the power sub-system.
True configurability revolves around dividing the system into two parts: input and output. Within its
total power rating, the input section is capable of supplying any number of customer-specified output
combinations. Each output is specific for a given voltage and maximum current. Many technical and
commercial issues influence the division between input and output, perhaps nowhere more crucially
than at the transformer. Unfortunately, building the transformer primary as part of the input and making
the secondary part of the output module means that every configuration requires assembly of a
custom transformer. XP uses a different topology that allows for easy local configuration at recognised
centers around the world.
107
The problem is resolved by separating input and output at the off-line 380 VDC rail, making each
module fully testable electrically, and making configuration a simple plug-in exercise. The input
includes protection, EMC filtering and a power factor correcting boost converter, to deliver stabilized
bulk power at 380 V DC from mains input between 90 – 264 V AC, 47 – 63 Hz. Although different
models can support different combinations of output sections, depending on their power rating and
physical size, modules will fit in any chassis within a family, using a mechanical and electrical
interface. This has the additional benefit that as new output modules are added to the current list of
single and dual models, they become immediately useable with any input stage in the family.
• Taking the heat
As equipment gets smaller, waste heat remains a key consideration for any power supply
sub-system so thermal and mechanical design is critical to flexibility. To emphasize the importance of
packaging, lower power models using this topology are designed to fit within a 1U system chassis
and are just 39.7mm high, despite offering configurable power up to 450 W in total. The same
thermal design has been applied to higher power models, enabling the enclosure of a 2,400 W multi-
output, configurable PSU with less than 124 mm overall height by packaging 2 chassis together in
an X10 ‘double decker’.
Fan and cover assembly

Fan
Main chassis - side view

Input Output
Main chassis - plan view
O/P
O/P
Input
module
O/P
O/P
Plug in output modules with voltage and current as required
Figure 1
Each complete PSU consists of three elements (Figure 1):
1) Main power chassis

2) Cover and fan assembly
3) Plug-in output modules
Input power conversion, protection and filtering is included in the main chassis which also provides
the plug-in location for the output modules. It is designed to deliver maximum airflow through the
plug-in modules, irrespective of configuration. Chassis and module construction both provide a rigid,
flat interface for mechanical location and efficient thermal management. Although the higher power
models combine the cover and fans into the chassis, lower power chassis may be supplied without
the fan and cover, further reducing cost where forced air is available within the end equipment or
system.
108
The flexibility of this modular topology is apparent when considering the fleXPower series of power
supplies. There are currently 33 single and dual output modules that use only two mechanical
formats across the whole range. Every fleXPower chassis can handle every module format.
Thousands of configurations are possible based on the options shown in Tables 1- 4.
Table 1: Chassis Options

Chassis Designations, Power & sizes
Code Power Slots
X4 400 W Industrial 10
XM4 400 W Medical 10
Table 2: fleXPower - Single Output Modules
Single Output - Module Voltage/Current Rating

Voltage Current Slots Code
3.3 V 40.0 A 2 2C
3.3 V 60.0 A 3 3C
5.0 V 40.0 A 2 2D
5.0 V 60.0 A 3 3D
12.0 V 17.0 A 2 2J
12.0 V 25.0 A 3 3J
15.0 V 14.0 A 2 2L
15.0 V 20.0 A 3 3L
24.0 V 10.5 A 2 2P
24.0 V 17.0 A 3 3P
28.0 V 9.0 A 2 2Q
28.0 V 14.0 A 3 3Q
36.0 V 7.0 A 2 2U
36.0 V 11.0 A 3 3U
48.0 V 5.2 A 2 2W
48.0 V 8.5 A 3 3W
60.0 V 4.2 A 2 2Y
60.0 V 7.0 A 3 3Y
109
Table 3: fleXPower - Dual Output Modules

Dual Output - Module Voltage/Current Rating
Output 1 Output 2
Slots Code
Voltage Current Voltage Current
5.0 V 10.0 A 5.0 V 10.0 A 2 5A
5.0 V 10.0 A 3.3 V 10.0 A 2 5B
12.0 V 10.0 A 12.0 V 8.0 A 2 5D
15.0 V 8.0 A 15.0 V 6.0 A 2 5E
15.0 V 8.0 A 12.0 V 8.0 A 2 5F
12.0 V 10.0 A 5.0 V 10.0 A 2 5G
12.0 V 10.0 A 3.3 V 10.0 A 2 5H
12.0 V 10.0 A 2.0 V 10.0 A 2 5J
15.0 V 10.0 A 5.0 V 10.0 A 2 5K
15.0 V 10.0 A 3.3 V 10.0 A 2 5L
15.0 V 10.0 A 2.0 V 10.0 A 2 5M
24.0 V 6.0 A 5.0 V 10.0 A 2 5N
24.0 V 6.0 A 3.3 V 10.0 A 2 5P
24.0 V 6.0 A 2.0 V 10.0 A 2 5Q
24.0 V 6.0 A 12.0 V 6.0 A 2 5T
Table 4: fleXPower Options

Parallel Option Codes Other Option Codes
Code Description Code Description
00 No parallel required 01 Reverse Air
12 Modules 1 & 2
02 Global Enable - Logic 1
13 Modules 1 to 3
14 Modules 1 to 4 03 Option 01 & 02
23 Modules 2 & 3 04 Global DC OK - Logic 1
24 Modules 2 to 4 05 Option 01 & 04
25 Modules 2 to 5
06 Option 02 & 04
34 Modules 3 & 4
35 Modules 3 to 5 07 Option 01, 02 & 04
40 Modules 1 & 2, 3 & 4 08 Global AC OK - Logic 1
09 Option 01 & 08
Series Option Codes
10 Option 02 & 08
Code Description
00 No series required 11 Option 01, 02 & 08
12 Modules 1 & 2 12 Option 04 & 08
13 Modules 1 to 3 13 Option 01, 04 & 08
23 Modules 2 & 3
14 Option 02, 04 & 08
24 Modules 2 to 4
40 Modules 1 & 2, 3 & 4 15 Option 01, 02, 04 & 08
110
• Design flexibility
A 15 V supply adjusted to 12 V is still capable of its maximum current rating at 15 V. This means that
20% of its space and capability could be redundant if not used. In order to avoid this while
maximizing power and space efficiency, each module features optimum output adjustment and is
available at many different voltage levels between 2.0 V and 60 V. To further increase space efficiency
dual and triple output models are available, offering either multiple voltages (e.g. 12 V, 12 V and 5 V)
or a single voltage (e.g. dual 5 V). The LP series, for example, offers 23 different multiple output
modules. Since all outputs are fully floating, it is also possible to series connect outputs to further
enhance the flexibility and capability of the power system.
For example, a customer required +10V, +12V and +36V outputs. These were supplied from an
off-the-shelf LP series module with standard rails of 24V, 15V & 5V as shown in Figure 2.
Customer example with medical approval

T9 Module V3 +10 V 3 A
5 V - 16 V
Adjusted to 10 V
+36 V 1.5 A
T9 Module V1
(24 V)
+12 V 2 A
T9 Module V2
5 V - 16 V
Adjusted to 12 V 0V
Figure 2
Overvoltage, overcurrent, remote sensing, current sharing, remote inhibit, AC and DC OK signals,
together with 1% ripple, 1% load regulation and 0.1% line regulation performance are included as
standard to meet virtually all requirements.
• What can you do with it?
As well as avoiding the engineering costs of custom PSU design, the flexibility of a locally configured
power supply can clearly reduce the time, cost and risks normally associated with changes to power
requirements. However, it is useful to remember that these changes can come from three sources:
1) Last-minute changes in system design

2) Updates or upgrades later in the system life-cycle
3) System functionality configured to individual customer requirements
With configuration and burn-in test requiring only 24 hours for sample quantities, designers can
ensure that the PSU configuration matches the latest system needs, yet when specifying today’s
requirement, they can be confident that future system upgrades will not make the whole power
supply obsolete. As system builders look more closely at lifetime costs, such built-in future proofing
can be an important benefit for end-users.
111
• There can be benefits to OEMs’ end customers too
Using the configurable power topology described, a supplier of laser surgery equipment, for
example, is able to benefit their customers who can select only those functions that they need from
their laser system to minimize cost, confident that this can be reflected in the power supply fitted to
their equipment. Cost is controlled, since unnecessary power supply capability is avoided, but if the
customer wants to upgrade or amend the system the power supply configuration can be simply
reworked accordingly. In another instance, a medical company cut costs by replacing a custom
power supply with a configurable solution, working with XP to develop a custom cable harness that
would fit into the existing application.
• Summary
Whether it is to meet low volume needs, where engineering costs for conventional custom designs
are hard to justify, or where volume requirements contain variation or uncertainty and where time to
market is key, the architecture described provides the flexibility to configure multiple output power
supply subsystems and to deliver complete, tested and fully safety approved (UL, CSA & TUV) units.
In-house design and dedicated configuration facilities ensure that designers have access to the level
of technical support they demand, while the topology enables maximum benefit to be gained from
low-cost manufacture of all sub-assemblies. The result is a flexible configuration service delivered at
competitive cost and with built-in future proofing.
112
• Designing Smaller, More Efficient
AC/DC Power Supplies
Contents
• Introduction
• Input filter
• Power factor correction circuit (PFC)
• Main converter
• Output rectifier
• Control circuit
• Summary
• Introduction
In the case of AC/DC power units, it is not dramatic technology breakthroughs that drive the trend; it
is good engineering and the inventiveness to combine the best of a whole range of techniques and
technologies that separate a really innovative power supply from an average one. This article looks at
AC/DC power supply design in the popular 100W to 200W range. It considers a combination of
design approaches that can be brought together to minimize the size and cost of the power unit,
whilst maximizing efficiency and application flexibility.
Let's start by defining some typical design goals. The power supply should be as small as possible to
save space or leave room for added system functions. It should make minimal contribution to the
waste heat in the system. In practice, it is microprocessors that now create most system heating but it
is still important that power supplies are designed for high efficiency; smaller heat sinks can then be
INPUT FILTER PFC MAIN CONVERTER SYNCHRONOUS RECTIFIERS

+
+
+
L N
-
CONTROL CIRCUIT STANDBY CONVERTER
- +
+
Frequency &
PWM Control
Figure 1
113
used to save space. For a 100W to 200W power supply, efficiency goals of 90% are not unrealistic.
A 1% efficiency improvement represents 10% less heat dissipation at the upper end of the range and
this can make a significant difference to the degree of cooling needed for the power supply. Cost, of
course, is an ever-present consideration, both in terms of bill of materials and manufacturing
complexity. Keeping the design as simple as possible is an important consideration in this respect.
Finally, functionality should not be compromised. Control and alarm signals, current sharing with
similar units and the ability of the power supply to maintain its performance over a wide range of AC
input conditions are all important.
Looking at the main stages within an AC/DC power supply shown in Figure 1, here are some proven
ways in which size and cost can be minimized without compromising performance or functionality.
• Input filter
A two-stage filter design using high permeability cores will minimize size while providing high
common mode and differential noise reduction. Stacking some components vertically can save board
space and improve cooling.
L N
Figure 2
• Power factor correction circuit (PFC)
The use of silicon carbide diodes has become economically feasible in the last few years as
component prices have fallen. Their reverse current characteristics mean that they don’t require a
snubber circuit, saving on 5 or 6 components. Furthermore, they contribute to a 1% typical efficiency
boost. Using a stepped gap inductor provides high inductance at high input line and supports
maximum flux density at low line. Using continuous conduction mode (CCM) operation throughout
the input range keeps the peak switching current and input filter requirements to a minimum.
Figure 3
114
• Main converter
A resonant topology can virtually eliminate switching losses. This not only improves power supply
efficiency but also enables smaller heat sinks to be used. In fact, compact ceramic heat sinks can
sometimes be used for power transistors, rather than metal ones. Their advantages include a
reduction in noise and consequently simplified filtering. This is because the heat sinks do not have
capacitive coupling with the drain connections of the switching MOSFETS. In addition, smaller
creepage distances, compared with those needed for metal heat sinks, can be used. This gives
further savings in board space.
Current
Voltage
0
Transition
Figure 4
• Output rectifier
Using switched MOSFETS rather than output rectifier diodes improves efficiency through a significant
reduction in power dissipation. For example, at 20 Amps a diode with 0.5V forward voltage gives a
power dissipation of 10W. Using a MOSFET with an ‘ON’ resistance of, say, 14mOhms at 100 °C
dissipates just 5.6W – a 44% improvement. Once again, ceramic substrates can replace
conventional heat sinks.
Figure 5
115
• Control circuit
Semiconductor manufacturers have been developing increasingly integrated control circuits for power
supplies in recent times. This means savings in component count, manufacturing costs and board
space, even where the integrated circuits themselves may be more expensive than a discrete
component approach. One example is the IR1150 – a PFC chip that operates as a one-cycle control
(OCC) device, which allows major reductions in component count without reducing power system
performance. Similar, application-specific chips can provide main converter voltage control plus over-
current protection, over-voltage protection and over-temperature protection. They can also control
the output rectifier switching. Other desirable control options for increased application flexibility
include power sharing with synchronous monotonic start-up, an inhibit circuit to shut down the
power supply via logic control, a ‘power good’ signal, and the control functionality needed for a
standby converter. The standby converter provides an independent 5V output whenever AC power is
present.
• Summary
Today’s best-in-class AC/DC switchers are typified by XP Power’s EMA212 power supply. Using
some the techniques described above, this packs 212.5W output from a 3 x 5 inch footprint with a
maximum height of 1.34 inches. That’s a power density of 10.55 W per cubic inch in an industry
standard footprint that fits within a 1U high enclosure. It delivers 200W from its main 12V, 24V or 48V
output, plus 12V at 1A for driving fans and a 5V standby output. The unit needs just 12CFM of
forced-air cooling, which is easily achievable using standard 40 x 40mm fans. Forced-air cooling is
now the norm in many communications systems and 12CFM is easily achievable without complex
mechanical arrangements. Finally, it achieves an efficiency of 91% at full rated load.
The possibilities for improvements in AC/DC power supply design will continue to be driven largely by
improvements in semiconductor performance and functionality. Better magnetic and passive
components also have a role to play, but here progress is more evolutionary than revolutionary. The
best power supplies are developed from a deep understanding of the latest proven component
technologies, plus a determination to explore how these technologies can be combined in new and
innovative ways to achieve ever more challenging design objectives.
116
• Design Considerations for Compact
& Flexible Power Supplies
Contents
• Introduction
• Mechanical flexibility
• Thermal considerations
• Efficiency and topology issues
• Output considerations
• Safety and EMC
• Manufacturing
• Agency approvals
• Introduction Figure 1
Many electronic systems require DC power rails outside the standard 3.3, 5, 12, 24 and 48 V ratings,
or require combinations of outputs not available from standard off-the-shelf products. Further, where
multiple products or product variants are produced there are often differing requirements for each
type. Commissioning custom designs can be an expensive, risky and lengthy process, especially in
medium volume applications. In higher power applications modular, configurable power supplies are
available from a number of vendors but these tend to become uneconomic below 300-350 W. Many
applications require 100-200 W and these solutions are too expensive and typically too large.
Flexible design, from a power supply perspective, means flexibility in mechanical options, electrical
options, output ratings and configurations, and careful consideration to agency requirements and
manufacturing.
• Mechanical flexibility
Some applications require open frame power supplies in this power range, others are better served
by U-channel or chassis mount construction. Some will require a safety cover; some will have
internal fan cooling while others will require a complete sub-assembly with integral cooling. Some
systems will be required to be entirely convection cooled due to noise or maintenance concerns.
These issues need to be considered from the outset of the design from the PCB through thermal
management and mechanical parts, as this level of flexibility is difficult to add later. Figure 1 shows 3
of the 4 mechanical versions of the same standard power supply platform, XP Power’s RCL175
AC/DC switcher.
Connection to the power supply can be a key consideration depending on the volume of products to
be manufactured. For simplicity and cost, screw terminals will usually be the preferred method in low
volume products. For medium & higher volume products push fit connectors – usually PCB headers
– offer the benefit of faster system assembly and reduced risk of interconnection errors. Designing
the power supply PCB to accommodate both options adds flexibility in this area.
117
• Thermal considerations
Cooling arrangements and environmental requirements differ considerably from application to

application so a flexible power supply design needs to maximize efficiency, minimize size to ease
integration, minimise waste heat in the end application, and maximize lifetime. The aim is to maximize
power density with a high convection-cooled rating and minimize forced cooling requirements to
obtain higher power levels from one flexible product. Products requiring 20 CFM and above can
present a huge challenge for the system designer, especially as systems and equipment continue to
shrink and audible noise levels come under pressure. 10 – 12 CFM is much more achievable without
incurring high costs and the resulting moderate levels of audible noise can be easily managed. The
combination of a substantial convection cooled rating and higher power force-cooled ratings also
allows the use of variable speed or thermally controlled fans which operate only at higher ambient
temperatures or under high system load, enhancing reliability and further reducing the audible noise.
• Efficiency and topology issues
Maximizing efficiency starts with the power factor correction (PFC) stage. Here, the availability of
silicon carbide diodes for this application benefits efficiency and reduces component count at a
similar cost to traditional diodes and their associated snubbing components (Figure 2).
Figure 2
The choice of power converter topology for a flexible product, designed to perform in a variety of
applications, is important. The fly-back converter is often the topology of choice in the 100-200 W
range but does have some limitations. The power throughput is limited by the size of the transformer
and hence the ability to provide peak power to complex loads such as motors, solenoids, lamps,
downstream DC/DC converters and systems with large amounts of capacitance is also limited.
Other topologies, such as forward or half bridge converters have a transformer size limited only by
temperature rise and are therefore capable of providing peak load requirements which occur on a
momentary basis or during power up. The half bridge converter has further advantages over the
forward converter. It brings a reduction in transformer size and allows the use of schottky barrier
rectifiers in the higher output voltage rails further enhancing overall efficiency. The use of a coupled
output inductor also enhances control of auxiliary output voltages over wide load variations.
118
• Output considerations
Many applications require multiple outputs and combinations of positive and negative polarity.
Keeping outputs isolated from one other allows flexibility in connection to provide such combinations,
as well as allowing parallel or series connections to provide higher output voltages or increased
output current. Isolated outputs also provide the facility for separate returns. These may be desirable
for digital and analogue circuits.
To allow the power supply to be easily configured to a specific requirement, the full range of outputs
must be considered at the outset; designing for both the highest voltage and the highest current is
key and requires care to avoid compromising overall performance. Typical requirements fall into the
range from 3.3 V to 60 VDC, higher output voltage requirements being served by the ability to
connect outputs in series. The challenge is to provide real flexibility in output voltages without
unnecessary complexity or additional post-regulation stages that increase cost and reduce efficiency.
Here a fractional turn transformer (FTT) can provide an effective solution.
Standard transformers are able to offer 1/2 turns by using E shaped cores, this not only severely
restricts the choice of output voltages but can lead to problems with balancing and regulation. XP
have used a different approach when designing the RCL175. This uses a separate transformer just
for fractional turns, allowing the main transformer to deal only with integer turns where it is more
efficient, smaller and provides better regulation.
The FTT provides significant advantages in multiple output supplies in terms of space saving, power
transformer optimization, cost cutting and efficiency.
Take a power supply requiring outputs of 5 V and 12 V. Ideally the 5 V winding will be a single turn
on the power transformer but the requirement for the 12 V winding becomes 2.3 turns which cannot
be readily achieved. The normal solution is to increase the 5 V winding to 2T and then increase the
12 V winding to 5T, by using diodes with differing forward volt drops an approximate 12 V output can
be realised. This means that the transformer size and losses are compromised, particularly as the
lower voltage windings are typically copper foils and the diodes are not optimized for minimum
losses.
Another common scenario in many applications is the need for 3.3 V and 5 V output combinations.
The FTT largely eliminates cross regulation and, with suitable voltage sensing and control, there is no
need for the post regulation usually employed to regulate the 3.3 V output. This approach removes
cost, enhances efficiency and reduces size. The principle is illustrated in Figure 3.
Figure 3
119
• Safety and EMC
The power supply is critical when it comes to achieving the necessary safety agency approvals and
compliance with electromagnetic compatibility requirements in any system.
The ideal scenario for a flexible power supply is to carry approvals for IT, industrial and medical
applications, and allow for both class I (with a protective earth) and class II (without a protective
earth) systems. The ability to deal with both class I and class II applications is increasingly important
in worldwide markets and as medical equipment in the home and residential clinics is becoming
commonplace and ground connections are not always as good as they should be. This demands
careful consideration of EMI filtering, PCB layout and topology to meet the requirements for
increased creepage and clearance, reduced leakage current and non-earthed systems while
maintaining compliance with international emissions and susceptibility standards. With appropriate
design & component selection, it is possible to achieve compliance with level B conducted
emissions, level A radiated emissions and level 3 susceptibility requirements.
• Manufacturing
The key consideration for manufacturing is to design the power supply on a single PCB covering all
requirements. This minimises both cost and delivery times and allows part built assemblies to be
configured to the application in a short time frame without compromising agency approvals. The
potential to configure modified standards in a short time, produce pre-production quantities on a
short lead time and manufacture volume requirements in a low cost area is a key consideration in the
initial design phase.
• Agency approvals
One of the most significant delays in bringing new equipment or systems to market can be gaining
the relevant agency approvals. With the right approach to design, safety testing & part numbering it
is possible to gain approval for a wide range of potential output voltages and configurations without
applying for re approval for each variation. This approach saves cost and, more crucially, saves
significant time in removing the need for additional approval cycles of the power supply during
equipment and system development.
All of the techniques described in this article have been combined in XP Power’s RCL175 series
shown in figure 1. The basic specification:
• 1-4 outputs
• Universal AC input with Active PFC
• Up to 120 W convection cooled, 175 W with 12 CFM
• Common PCB’s for any potential variation
• Compact 3.7” x 5.5” footprint
• Agency approved for output voltage to 60VDC on individual outputs
• Industrial, IT & medical agency approvals
• Class I & class II safety approvals
• Level B conducted & level A radiated emissions
• Four optional mechanical/cooling formats
• Choice of PCB header or screw terminal connectors
• Options for conformal coating, overload characteristics & remote inhibit/enable
120
• Synchronous Rectification Joins the Mainstream
Contents
• Introduction
• What is synchronous rectification?
• What are the benefits of synchronous rectification?
• The low voltage challenge
• Synchronous rectification for the mainstream
• Hidden costs
• Ring-free zero voltage switching
• Synchronous rectification for all
• Introduction
Synchronous rectification delivers higher efficiency and therefore more compact power conversion
equipment. Early products were aimed at the premium end of the market, with premium price tags.
However, using patented technology and a new topology, it has proved practical to bring the benefits
of synchronous rectification to a wider range of more price-sensitive applications, and this article
illustrates how high efficiency power supplies are now available at 5 V and below, within a
mainstream 130 W range.
• What is synchronous rectification?
Rather than using output rectifier diodes to conduct when forward biased, synchronous rectification
uses switched MOSFET transistors. The success of this technique has been driven by the rapid
reduction in the cost of high current MOSFETs, combined with a significant reduction in the available
‘ON’ resistance values. The benefits become clear by considering losses at 20 A:
A diode with nominal 0.5 V forward voltage gives a power dissipation of:
(Using P = IV) 20 x 0.5 = 10 W
Whereas a MOSFET ‘ON’ resistance of 14 mOhms at 100 °C, gives a power dissipation of:
(Using P = I2R) 20 x 20 x 0.014 = 5.6 W
This results in a 4.4% increase in efficiency in a 100 W power supply.
Although losses are nearly halved, it is worth noting that higher voltage MOSFETs typically have
higher RDS values, so synchronous rectification is not ideal for all output voltages.
The MOSFET must be switched, and conventional designs use relatively complex ICs to provide the
timing and control necessary, often offering precision performance, but at a price. Of particular
importance is synchronising the switching to ensure that only one switch is ever ON and to minimize
switching losses and noise generation. We shall come back to these issues later, but let us review
the core benefits of the technique.
121
• What are the benefits of synchronous rectification?
Although efficiency is often used as the headline feature of PSUs, the benefits become much clearer
by focusing on losses rather than efficiency. A change of 10% in efficiency from 80% to 90%
represents losses dropping from 20 W to 10 W at 100 W. That represents half the power loss, and
therefore half the waste heat. Even moving from 85% to 90% reduces losses from 15 W to 10 W, a
33 % reduction in waste heat.
Less dissipated power may mean no forced cooling and a reduction in the physical size of the power
supply (PSU). In addition, when the PSU is mounted close to other electronics, it will not just be
heating itself. Equipment failure rates typically double for every extra 10 °C increase in ambient
temperature, so reduced PSU losses can improve overall system reliability. Less heat may also mean
that the mounting or orientation of the PSU is less critical, since it is no longer vital to align special
heatsinks with unimpeded airflow.
Because diode losses depend on I x VFD (where VFD is the diode forward volt drop), paralleling
diodes does not reduce dissipation; therefore large diodes are needed at high currents. Conversely,
MOSFET losses depend on I2 x RDS (where RDS is drain-source ON resistance), so splitting current
between two MOSFETs reduces each current by two and the dissipation in each device by two
squared, i.e. to a quarter, halving the total dissipation. The positive temperature dependence of RDS
also means that MOSFETs tend to share current well. Using multiple, smaller devices gives significant
mechanical flexibility to adjust form factor, height etc, and enables the use of lower cost devices.
• The low voltage challenge
A quick check of most power supply ranges will underline the difficulties of delivering power at 5 V
and below, and it is important to check the datasheet to see whether power ratings are reduced,
significantly greater airflow is specified, or the designer is left to work out how to keep a heatsink
below a certain maximum temperature.
The limitations of PCB track losses for sub-3V supplies are leading to the widespread adoption of
point-of-load regulation, but there is still a significant requirement for 5 V and 3.3 V PSU modules
that are easy to use,compact, reliable and low-loss. By focussing on meeting this need at low cost,
XP has introduced a new topology incorporating synchronous rectification and several patented
innovations to reduce noise and provide power factor correction to further simplify system design.
The JPS130 power supply uses the new topology.
• Synchronous rectification for the mainstream
Given the need to control MOSFET switching, most existing designs use complex control chips,
often delivering sophisticated control and performance features, but placing the core benefits of
synchronous technology outside the budget of mainstream systems. These control systems typically
combine both input and output control in a single device. However, by separating the input and
output control (See Figure 1) and ensuring that the primary DC bus voltage is held stable, limiting the
range of duty cycle within the switching power stage, the output MOSFETs can be switched using
relatively simple, and therefore inexpensive, control. Since this input stage can drive either a
conventional diode output, or a synchronous rectifier output, it is possible to produce a single range
that matches the output section to the rated voltage, i.e. synchronous rectification for 5 V and 3.3 V
units and conventional output rectifier diodes for higher voltages, maximizing volume cost advantages.
122
Diode
Rectifcation
400 V Patented
Boost Primary Transfomer O/P
Regulator Drive
Synchronous
Rectifcation
Boost regulator controls Simple control delivers LV models use

400 VDC to simplify drive industry standard synchronous rectification,
and rectification control performance from either higher voltages use
rectifier type diodes
Figure 1
The input section is designed as a boost sub-system, enabling the design to maintain the DC line
across the universal range of 85 – 264 VAC, from 47 – 63 Hz, with built-in active power factor
correction to give a typical power factor >0.9 and enabling the low-cost output stage to deliver
industry standard performance. All the usual features such as overcurrent and overvoltage protection,
remote sense and high MTBF are achieved, but there are some additional issues needed to make
the PSU simple and economical to use.
• Hidden costs
Most system designers have met the additional cost of having to provide forced airflow, with
attendant concerns such as fan reliability and excessive noise, so performance should be simple to
understand. The same size module is rated at 130 W in 18 CFM air, and 100 W for convection
cooling, whether 48 V or 5 V output, thanks to the use of synchronous rectification for the 5 V model
and below, and unlike other ranges where 5 V and 3.3 V have the same output current, the JPS130
3.3 V model delivers 25 % extra current at 25 A, compared to 20 A for the 5 V unit. As can be seen
from Figure 2, convection cooling enables 100 W up to 50 °C ambient for 48, 24, 15 and 12 V
models, and up to 40 °C for 5 V, with 80 W available up to 50 °C from the 3.3 V model.
130 A JPS130 with 18 CFM fan.

A
100 B JPS130PS12, PS15,
PS24, PS48 with
L o a d (W)
80 B convection cooling &

D
60 JPS130PS03 with 18
CFM fan.
40 C
C JPS130PS05, PT30,
20
PT31 with convection
cooling.
-10 0 10 20 30 40 50 60 70
Ambient Temperature (˚C) D JPS130P03 with

Figure 2 convection cooling.
123
Power supplies are critical to equipment EMC performance, but although system performance will
naturally depend on issues such as loading and wiring, it is clear that if a power supply only just
meets the required EMC standards, any locally created noise is likely to cause problems. The lower
the EMI noise of the PSU, the easier it will be to achieve EMC compliance for the whole system, and
since switching contributes significantly to EMI noise as well as providing another component of
power loss, this aspect of PSU design is key.
• Ring-free zero voltage switching
Zero voltage switching (ZVS) has been widely explored as a way to deliver compact, efficient power
conversion. Resonant circuits typically required either a large gap for the main transformer core, or
an external inductor, but parasitic ringing between leakage inductance and stray capacitance
introduce additional losses and unwanted noise.
By using a patented ZVS switching circuit, which automatically clamps ringing for both inductive and
capacitive components, a 6 dB improvement in noise has been achieved, reducing power losses,
lowering the reverse voltage rating for the secondary rectification and increasing the practical
operating frequency. These benefits apply whether the output is synchronous rectification (5 V or
3.3 V) or conventional diodes (12 V and above) and contributes to the 6.7 W/in3 power density for
the 5 V to 48 V models.
• Synchronous rectification for all
When introduced, synchronous rectification was a revolution, but perhaps it is time for system
designers to move on and focus on what a given power supply will do for their system and the
overall design process. With this new topology, the technology is now affordable for mainstream
systems, so designers can return to looking at the fundamentals of their system needs, rather than
viewing synchronous rectification as a ‘holy grail’.
Although efficiency is a common headline specification, comparing the actual power loss will give a
better comparison between the system-level impacts of different PSUs, and provide a more valuable
insight into factors such as cooling requirements and overall reliability. It is also important to address
issues such as EMC margins and input power factor to ensure that the final system meets
international standards for EMC immunity and emissions.
124
• Military Power from
Commercial Modules
Contents
• Introduction
• Environmental considerations
• Input & output voltages
• EMC compliance
• COTS solutions
• Introduction
COTS (Commercial Off The Shelf) parts can provide fast, cost-effective solutions for power system
design, but only when key industry standards are fully understood.
Designers of electronic equipment for military applications work under the same pressures and
constraints as their commercial counterparts. The rate of product development is increasing, leading
to increased time-to-market pressures. Budgets are being squeezed, so the adaptation of COTS
parts is an attractive option, providing overall system performance and uncompromised reliability and
ensuring that the final power units meet the requirements of the appropriate DEF-STAN, MIL-STD or
relevant industry standard, depending on the end use.
Power supplies are critical system components - if power fails, the whole system fails. Also, the
power supply is often key in protecting sensitive electronics from excessive input voltage variation
and providing input filtering for the system. Good power supply design is a pre-requisite of reliable
system design.
Some military systems designers opt to build their own power supplies from discrete components. It
can appear to be a relatively cheap solution at the outset, but issues of unpredictable performance,
time for experimentation and potential delays in product qualification can all add to the real costs of
the product.
Buying-in a custom-designed power supply is another option. This can deliver an optimum technical
solution, but with the risks of high up-front engineering charges and a lengthy design cycle. The latter
often runs to between four and six months and the small volumes required by many military
applications can make unit costs unacceptably high.
In many applications, the ideal solution is found in a combination of standard AC/DC or DC/DC
power modules, or 'bricks', to which customised input and output circuits and customised
packaging are applied. Key to the successful implementation of such a solution is a detailed
understanding of military specifications and the special conditions under which the power supplies
will be required to operate.
125
• Environmental considerations
Extremes of temperature are encountered in military applications and it’s not just high temperatures
that cause concern. Some applications require equipment that needs to be capable of operating at
–40 °C or even –55 °C and, while a first glance at product datasheets might indicate that this is
possible, it’s an entirely different matter when the detailed performance and EMC specifications at
these temperatures are considered. Compensation for degraded performance at low temperatures
needs to be included in the overall power solution.
High temperatures cause similar problems but with additional implications for reliability. As a rule of
thumb, the mean time between failures (MTBF) halves for every 10 °C increase in operating
temperature. In some applications, designers must avoid the use of forced-air cooling. Fans are
electromechanical components that can compromise system reliability. They require monitoring
circuits that add to system complexity and cost, and their associated filters are prone to clogging in
many environments, reducing cooling effectiveness and increasing maintenance work. Power
modules with baseplates are most effective in dissipating excess heat and they need to be carefully
designed into packaging that effectively conducts and convects heat away from sensitive electronics.
By using appropriate power modules, baseplate operating temperatures of between 85 and 100 °C
can be achieved.
Resistance to shock and vibration is another important environmental consideration. Normal potting
compounds used to encapsulate DC/DC power modules are prone to crystallisation at very low
temperatures, leading to damage to internal components. Specially developed soft-potting
compounds and spin-fill techniques can be used to overcome these problems. Finally, the ingress of
dust and moisture needs to be avoided.
• Input and output voltages
The input voltages needed for military applications rarely coincide with those used in commercial
ones. Military specifications are stringent for other input characteristics too, including low and high
line conditions and the capability of power supplies to handle voltage spikes, surges and excessive
input ripple. Not only do these special input requirements demand flexibility from the DC/DC power
module manufacturer but they also mean that input conditioning circuits are invariably required.
Output voltages are often non-standard when compared with commercial products. A power module
manufacturer that offers a very wide product choice may well be able to provide something with an
output that is trimmable to the required voltage but it’s important to remember that over- and
undervoltage trips will probably not move to accommodate the change in nominal output voltage.
A few manufacturers will adapt DC/DC power modules for user-defined output voltages, with
appropriate over- and under-voltage trip points. Trip points can be set or deleted depending upon
design requirements.
• EMC compliance
Military standards typically specify limits from 10Hz to 1GHz as well as a range of conducted and
radiated susceptibility and emissions requirements. Once again input and output conditioning are
needed to achieve compliance. EMC filtering can be achieved using discrete components or filter
modules. Active filters remove spikes and filter both conducted emissions and conducted
susceptibility such as transients or input ripple appearing at the output. Radiated emissions are dealt
with by complete screening of the final power supply.
126
• COTS solutions
Companies such as XP can produce reliable and cost-effective power supplies for military
applications based upon COTS DC/DC power modules. The block diagram of a typical design is
shown below. Successful designs demand an in-depth understanding of the special requirements of
military systems and access to the widest possible range of reliable DC/DC power modules.
XP provides analogue design for the input and output circuits and 3D mechanical modelling of power
supplies based on COTS units. The company also carries out pre-compliance testing and assists
customers in the final product qualification process.
V&I
Monitoring
Transient EMI Active Surge DC/DC O/P

Power Passive O/P
Suppression Filtering Protection
Module Filter
Parallel Power
Inhibit Function Modules
127
Glossary level of protection, and the test voltage used is
1500 VAC. Quite frequently, safety standards call
Abnormal Failure for basic insulation between secondary circuits
An artificially induced failure of a component, (e.g. between a telecom network and SELV
usually as a result of ‘abnormal’ testing for circuits).
regulatory agency safety compliance.
Bode Plot
Ambient Temperature A graphic plot of gain versus frequency for a
The still-air temperature in the immediate vicinity control loop, typically used to verify control loop
of a power supply. stability, including phase margin.
Apparent Power Breakdown Voltage

A value of power for AC circuits which is The maximum AC or DC voltage which may be
calculated as the product of RMS current times applied from input to output and/or chassis of a
RMS voltage, without taking the power factor power supply. See Figure 2.
into account.
Autoranging Input
An input voltage sensing circuit in the power RECTIFIER
INPUT FILTER & OUTPUT
supply which automatically switches to the REGULATOR
appropriate input voltage range (90-132 VAC or
180-264 VAC).
BREAKDOWN VOLTAGE
Balun
A transformer which presents a high impedance Figure 2
to common-mode signals and a low impedance
to differential-mode signals. It is commonly used Bridge Rectifier
on the input of switching power supplies to A full wave rectifier circuit employing four
suppress common-mode noise. See Figure 1. rectifiers in a bridge configuration.
Brown-out
Condition during peak usage periods when
SWITCHING
electric utilities reduce their nominal line voltage
IN OUT
SUPPLY by 10% to 15%.
BSMI
Bureau of Standards Metrology & Inspection.
Figure 1
Certification body for Taiwan.
Burn-in
Bandwidth Operating a newly manufactured power supply,
A range of frequencies over which a certain usually at rated load and elevated temperature,
phenomenon is to be considered. for a period of time in order to force component
infant mortality failures or other latent defects
Basic Insulation before the unit is delivered to a customer.
According to international safety standards (e.g.
UL60950, EN60950) basic insulation provides Capacitive Coupling
basic protection against electric shock i.e. one Coupling of a signal between two circuits, due
128
Glossary
to discrete or parasitic capacitance between current at some maximum value whenever an

the circuits. overload of any magnitude is experienced.
CCC Constant Current Power Supply

China Compulsary Certification. Certification A power supply which regulates its output
body for China for product safety and EMC. current, within specified limits, against changes
in line, load, ambient temperature and time.
Center Tap
An electrical connection made at the center of a Constant Voltage Power Supply
transformer or inductor winding, usually so as to A power supply designed to regulate the output
result in an equal number of turns on either side voltage for changes in line, load, ambient
of the tap. temperature and drift resulting from time.
Centering Creepage Distance

The act of setting the output voltage of a power The shortest distance between two conducting
supply under specified load conditions, usually parts measured along the surface or joints of the
an auxiliary output of a multiple output power insulating material between them.
supply with all outputs at half load.
Crest Factor
CISPR In an AC circuit, crest factor is the mathematical
International Special Committee on Radio ratio of the peak to RMS values of a waveform.
Interference. Crest factor is sometimes used for describing
the current stress in AC mains supply wires for a
Clearance Distance given amount of power transferred, the RMS
The shortest distance (through air) separating value and hence the losses become greater with
two conductors or circuit components. increasing peak values.
Common-mode Noise Cross Regulation

The component of noise that is common to both In a multiple output power supply, the
the live and neutral conductors with respect to percentage voltage change at one output
ground, also the component of noise that is caused by the load change on another output.
common to both the DC output and return lines
with respect to input ground. Crowbar
An overvoltage protection circuit which rapidly
Compliance Voltage places a low resistance shunt across the power
The output voltage of a constant current power supply output terminals if a predetermined
supply. voltage is exceeded. See Figure 3.
+5V
Conducted Immunity
The immunity of a product to bursts of short SCR Vz = 6.2 V
duration, fast rise time transients that may be
generated by the switching of inductive loads,
TO POWER
contactors etc. SUPPLY OUTPUT
OUTPUT
R
Configurable
See Modular
COM
Constant Current Limiting Circuit
Current-limiting circuit which holds output Figure 3
129
Glossary
CSA provides two levels of protection and the test

Canadian Standards Association. An voltage is 3000 VAC for IT and industrial
independent Canadian organization concerned equipment, and 4000 VAC for medical
with testing for public safety. equipment.
Current Limiting Distributed Power Architecture (DPA)

See Output Current Limiting. This is a power distribution system where the
conversion to lower voltages is effected locally
Current Limiting Circuit near the load. An interim DC voltage is provided
A circuit designed to prevent overload of a from the AC mains or DC bus by a converter. This
constant-voltage power supply. It can take the is then distributed to smaller DC/DC converters.
form of constant, foldback or cycle-by-cycle Some versions of this system are also known as
current limiting. Intermediate Bus Architecture (IBA). See page 6.
Current Share Drift

The accuracy with which two or more power The change in output voltage of a power supply
supplies share a load current. An active share over a specified period of time, following a
control connection is sometimes employed warmup period, with all other operating
which may be described as a current share or parameters such as line, load, and ambient
power share connection. temperature held constant.
Derating Dropout
The specified reduction in an operating The lower limit of the AC input voltage where the
parameter to improve reliability. Generally for power supply begins to experience insufficient
power supplies it is the reduction in output input to maintain regulation.
power at elevated temperatures. See Figure 4.
Dynamic Current Allocation
IF1
A system for dual positive outputs such as 5V &
3.3V where the full amount of current may be
OUTPUT CURRENT
taken from either output in whichever

I1
combination is required. For instance, in a 6A
system any value of current from 0A to 6A may
be taken from the 3.3V output and the
remainder from the 5V or vice versa.
0 T1 T2
OPERATING TEMPERATURE
Figure 4 Dynamic Load Regulation
See Transient Response.
Differential Mode Noise
The component of noise measured between the Earth Leakage Current
live and neutral conductors, and also the The current that flows through the earth
component of noise measured between the DC conductor of a piece of equipment under normal
output and output return. See Ripple and Noise. conditions. This is limited by legislation. Limits
depend upon the application.
Dips and Interruptions
Short input interruptions to simulate the utility Efficiency
supply under various conditions. The ratio of output power to input power. It is
generally measured at full-load and nominal line
Double Insulation conditions. In multiple output switching power
Insulation comprising both basic insulation and supplies, efficiency is a function of total output
supplementary insulation. Double insulation power.
130
Glossary
EFT/Burst Filter
See Conducted Immunity. A frequency-sensitive network that attenuates
unwanted noise and ripple components of a
Eighth Brick rectified output.
An industry standard package size and pin-out
for DC/DC converters. The package size is 2.3” Floating Output
x 0.9” with the pins on a 2.0” spacing. The An output of a power supply that is not
height is typically 0.3”. connected or referenced to any other output
usually denotes full galvanic isolation. They
Electromagnetic Interference (EMI) generally can be used as either positive or
Also called radio frequency interference (RFI), negative outputs. Non-floating outputs share a
EMI is unwanted high frequency energy caused common return line and so are referenced to
by the switching transistors, output rectifiers and one another.
zener diodes in switching power supplies. EMI
can be conducted through the input or output Fly-back Converter
lines or radiated through space. The fly-back converter is the simplest type of
switcher. In most cases, it uses one switch and
Enable only needs one magnetic element - the
The ability to turn on electrically the output of a transformer. Practical output power from flyback
power supply from a remote location. converters is limited to less than 150W.
See Figure 5 and page 2.
Equivalent Series Resistance (ESR)
The amount of resistance in series with an ideal
capacitor which exactly duplicates the OUTPUT
performance of a real capacitor. In high
frequency applications low ESR is very INPUT
important.
PWM ISOLATION
Electrostatic Discharge (ESD)

Discharge of static electricity built up when two
insulating materials are rubbed together.
Figure 5
ETSI
The European Telecommunications Standards Foldback Current Limiting Circuit
Institute (ETSI) is a non-profit-making Current limiting circuit that gradually decreases
organization whose mission is to determine and the output current under overload conditions
produce the telecommunications standards that until some minimum current level is reached
will be used for decades to come. It is an open under a direct short circuit. See Figure 6.
forum which unites 696 members from 50
countries, representing administrations, network
operators, manufacturers, service providers and
users.
Vout
FCC
The FCC (Federal Communications Commission)
is an independent United States government
agency, directly responsible to Congress and Iout 100%
load
charged with regulating interstate and
international communications by television, Figure 6
radio, wire, satellite and cable.
131
Glossary
Forward Converter mounting holes are provided for the attachment

Similar to a fly-back converter but the forward of heatsinks and to the customer’s board.
converter stores energy in the output inductor
instead of the transformer. See page 3. Half Bridge Converter
A power switching circuit similar to the full
Front End bridge converter except that only two transistors
A particular type of AC/DC converter (usually (or diodes) are used, with the other two replaced
high power) used in distributed power by capacitors. See page 4.
architecture (DPA) and Intermediate Bus
Architecture (IBA) systems which provides the Harmonic Currents
DC voltage that is bussed around the system. Current distortion generated by non-linear loads
such as the input to a switch mode power
Full Brick supply.
An industry standard package size and pin out
for DC/DC converters. The package size is 2.4” Heatsink
x 4.6” with the pins on a 4.2” spacing. The Device used to conduct away and disperse the
height is typically 0.5” without a heatsink. Four heat generated by electronic components.
mounting holes are provided for the attachment
of heatsinks and to the customer’s board. Hiccup Mode
See Trip & Restart Current Limiting
Full Bridge Converter
A power switching circuit in which 4 transistors Hi-Pot Test
are connected in a bridge configuration to drive High potential test. A test to determine if the
a transformer primary. See page 4. breakdown voltage of a transformer or power
supply exceeds the minimum requirement. It is
Galvanic Isolation performed by applying a high voltage between
Two circuits which have no significant ohmic the two isolated test points.
connection are considered to be “galvanically
isolated” from each other. Galvanic isolation Hold-up Time
(separation) is achieved by using a transformer, The time during which a power supply’s output
opt-coupler, etc. voltage remains within specification following the
loss of input power.
Ground
An electrical connection to earth or some other Hot Swap
conductor that is connected to earth. Redundant units which may be removed and
Sometimes the term “ground” is used in place of replaced without the need to power down
“common”, but such usage is not correct unless equipment.
the connection is also made to earth.
IEC
Ground Loop International Electrotechnical Commission.
An unwanted feedback condition caused by two
or more circuits sharing a common electrical Induced Noise
ground line. Noise generated in a circuit by a varying
magnetic field produced by another circuit.
Half Brick
An industry standard package size and pin-out Inhibit
for DC/DC converters. The package size is 2.40” The ability to electrically turn off the output of a
x 2.28” with the pins on a 1.90” spacing. The power supply from a remote location via a logic
height is typically 0.50” without a heatsink. Four level signal.
132
Glossary
Input Line Filter Line Regulation

A low-pass or band-reject filter at the input of a The variation of an output voltage due to a
power supply which reduces line noise fed to change in the input voltage, with all other factors
the supply. This filter may be external to the held constant. Line regulation is expressed as
power supply. the maximum percentage change in output
voltage as the input voltage is varied over its
Input Voltage Range specified range.
The high and low input voltage limits within
which a power supply or DC/DC converter Linear Regulator
meets its specifications. A common voltage-stabilization technique in
which the control device (usually a transistor) is
Inrush Current placed in series or parallel with the power
The peak instantaneous input current drawn by source to regulate the voltage across the load.
a power supply at turn-on. The term 'linear' is used because the voltage
drop across the control device is varied
Inrush Current Limiting continuously to dissipate unused power.
A circuit which limits the inrush current during
turn-on of a power supply. Load Regulation
Variation of the output voltage due to a change
Intermediate Bus Architecture (IBA) in the output load, with all other factors held
See Distributed Power Architecture (DPA). constant. It is expressed as a percentage of the
nominal DC output voltage.
Inverter
A power converter which changes DC input Local Sensing
power into AC output power. Using the output terminals of the power supply
as sense points for voltage regulation.
Isolation
The electrical separation between input and Logic Enable
output of a power supply by means of the The ability to turn a power supply on and off
power transformer. The isolation resistance with a TTL signal. A logic low generally turns the
(normally in mega ohms) and the isolation supply off; logic high turns it on.
capacitance (normally in pico farads) are
generally specified and are a function of Long Term Stability
materials and spacings employed throughout Power supply output voltage change due to time
the power supply. with all other factors held constant. This is
expressed in percent and is a function of
Isolation Voltage component ageing.
The maximum AC or DC voltage that may be
applied for a short, defined duration from input Magnetic Amplifier
to output and/or chassis of a power supply. A magnetic device used to improve the cross
regulation of multiple output AC/DC converters.
KETI
Korean Electrical Testing Institute. Certification Margining
body for safety and EMC in Korea. Adjusting a power supply output voltage up or
down from its nominal setting in order to verify
Line Frequency Regulation system performance. This is usually done
The variation of an output voltage caused by a electrically by a system-generated control signal.
change in line input frequency, with all other
factors held constant. This effect is negligible in
switching and linear power supplies.
133
Glossary
Minimum Load supporting the components and making

The minimum load current/power that must be electrical connections.
drawn from the power supply in order for the
supply to meet its performance specifications. Operational Insulation
Less commonly, a minimum load is required to Operational insulation is needed for the correct
prevent the power supply from failing. operation of the equipment, but does not
protect against electric shock. Operational
Modular insulation provides no levels of protection and
A physically descriptive term used to describe a typically the test voltage is ≤ 500 VDC.
power supply made up of a number of separate
subsections, such as an input module, power Operating Temperature Range
module, or filter module. See Temperature Range, Operating.
MOSFET Operational Power Supply

Metal oxide semiconductor field effect transistor. A power supply with a high open loop gain
The device of choice for the main switch in regulator which acts like an operational amplifier
many switch mode power supplies, having and can be programmed with passive
much better switching characteristics than components.
bipolar transistors.
Output Current Limiting
MTBF An output protection feature which limits the
Mean time between failures. The failure rate of a output current to a predetermined value in order
system or component, expressed in hours, to prevent damage to the power supply or the
established by the actual operation load under overload conditions. The supply is
(demonstrated MTBF) or calculated from a automatically restored to normal operation
known standard such as MIL-HDBK-217. following removal of the overload.
Noise Output Good

Noise is the aperiodic, random component of A power supply status signal which indicates
undesired deviations in output voltage. Usually that the output voltage is within a certain
specified in combination with ripple. See PARD tolerance. An output that is either too high or
and Ripple. too low will deactivate the output good signal.
Nominal Value Output Impedance

The stated or objective value for a quantity, such The ratio of change in output voltage to change
as output voltage, which may not be the actual in load current.
value measured.
Output Noise
Off-line Power Supply The AC component which may be present on
A power supply which operates off the AC line the DC output of a power supply. Switch-mode
directly, without using a power transformer prior power supply output noise has two major
to rectification and filtering. components; a lower frequency component at
the switching frequency of the converter and a
Open Frame high frequency component due to fast edges of
A power supply with no external metal chassis; the converter switching transitions. Noise should
the power supply is provided to the end user always be measured directly at the output
essentially as a printed circuit board which terminals with a probe having an extremely short
provides mechanical support as well as grounding lead. See page 45.
134
Glossary
Output Voltage Over Voltage Protection (OVP)

The nominal value of the DC voltage at the A power supply feature which shuts down the
output terminals of a power supply. supply, or crowbars or clamps the output, when
its voltage exceeds a preset level. See Crowbar.
Output Voltage Accuracy
For a fixed output supply, the tolerance in Parallel Operation
percent of the output voltage with respect to its The connection of the outputs of two or more
nominal value under all minimum or maximum power supplies of the same output voltage to
conditions. obtain a higher output current than from either
supply alone. This requires power supplies
Output Voltage Trim specifically designed to share the load.
The adjustment range of a power supply or
DC/DC converter via a potentiometer or external PARD
programming of voltage, current or resistance. Periodic and random deviation. A term used for
the sum of all ripple and noise components
Overload Protection measured over a specified bandwidth and
An output protection feature that limits the stated in either peak-to-peak or RMS values.
output current of a power supply under overload See Figure 8.
conditions so that it will not be damaged.
Overshoot SWITCHING POWER SUPPLY

A transient change in output voltage, in excess
of specified output accuracy limits, which can
OUTPUT VOLTAGE
occur when a power supply is turned on or off

or when there is a step change in line or load. LINEAR POWER SUPPLY
See Figure 7.
(TIME NOT TO SCALE)

TRANSIENT
RECOVERY
TIME OVERSHOOT
TIME
Figure 8
UNDERSHOOT
LOAD CHANGE
OUTPUT
ACCURACY Peak Power
LIMITS
The absolute maximum output power
that a power supply can produce without
immediate damage. Peak power capability is
TIME typically well beyond the continuous reliable
Figure 7 output power capability and should only be used
within the defined specification.
Over Temperature Protection
A protection system for converters or power Power Factor Correction (PFC)
supplies where the converter shuts down if the Standard AC/DC converters draw line current in
ambient temperature exceeds the converter’s pulses around the peaks in line voltage. This
ratings. OTP is intended to save the converter may be undesirable for several reasons. PFC
and any downstream equipment in the event of circuits ensure that the line current is drawn
a failure of a fan or such. OTP usually measures sinusoidally and in phase with the sinusoidal line
the hottest item on board the converter rather voltage. See page 30.
than ambient temperature.
135
Glossary
Pi Filter (π filter) supply to stabilise its operation. A bleed resistor

A commonly-used filter at the input of a usually provides a pre-load. See also Minimum
switching supply or DC/DC converter to reduce Load.
reflected ripple current. The filter usually consists
of two parallel capacitors and a series inductor Pre-regulation
and is generally built into the supply. The regulation at the front-end of a power
See Figure 9. supply, generally by a type of switching
regulator; this is followed by output regulation,
usually by a linear type regulator.
L
PSE
INPUT OUTPUT Product Safety Electric Appliances and
C1 C2 Materials. Certfication for Safety and EMC in
Japan.
Primary
Figure 9
The input section of an isolated power supply
that is connected to the AC mains and hence
Post Regulation has dangerous voltage levels present.
A linear regulator used on the output of a
switching power supply to improve line and load Programmable Power Supply
regulation and reduce output ripple voltage. A power supply with an output controlled by an
See Linear Regulator. external resistor, voltage, current or digital code.
Power Density Pulse Width Modulation

The ratio of output power per unit volume. A method of voltage regulation used in switching
Typically specified in W/In3. supplies whereby the output is controlled by
varying the width, but not the height, of a train of
Power Factor pulses that drive a power switch.
The ratio of true power to apparent power in an
AC circuit. In power conversion technology, Push-Pull Converter
power factor is used in conjunction with A power switching circuit which uses a center
describing the AC input current to the power tapped transformer and two power switches
supply. which are driven on and off alternately. This
circuit does not provide regulation by itself.
Power Fail Detection
A power supply option which monitors the input Quarter Brick
voltage and provides an isolated logic output An industry standard package size and pin-out
signal when there is loss of line voltage. for DC/DC converters. The package size is 1.45”
x 2.28” with the pins on a 2.0” spacing. The
Power Foldback height is typically 0.50” without a heatsink.
A power supply feature whereby the input power Four mounting holes are provided for the
is reduced to a low value under output attachment of heatsinks and to the customer’s
overload conditions. board.
Power Sharing Radiated Electromagnetic Interference

See Current Share. Also called radio frequency interference (RFI),
EMI is unwanted high-frequency energy caused
Pre-load by the switching transistors, output rectifiers and
A small amount of current drawn from a power zener diodes in switching power supplies. The
136
Glossary
portion that is radiated through space is known Reliability

as radiated EMI. Most radiated EMI measure- The ability of a system or component to perform
ments are done between 30MHz and 1GHz. its required functions under stated conditions for
a specified amount of time.
Radiated Immunity
The immunity of a product to electromagnetic Remote Enable
fields. Power supply interface signal, often TTL
compatible, which commands the power supply
Rated Output Current to start up one or all outputs.
The maximum load current which a power
supply was designed to provide at a specified Remote Inhibit
ambient temperature. Power supply interface signal, often TTL
compatible, which commands the power supply
Redundancy (N+1) to shut down one or all outputs.
Power supplies connected in parallel operation
so that if one fails, the others will continue Remote ON/OFF
delivering enough current to supply the One or other of remote enable or remote inhibit,
maximum load. This method is used in or a combination of both.
applications where power supply failure cannot
be tolerated. See page 58. Remote Sensing
A technique of regulating the output voltage of a
Reference power supply at the load by means of sensing
The stable voltage, generally a zener diode, from leads which go from the load back to the
which the output voltage of a regulated supply is regulator. This compensates for voltage drops in
controlled. the load leads. See Figure 10 and page 43.
Reflected Ripple Current POWER SUPPLY

OUT
The AC current generated at the input of a SENSE
power supply or DC/DC converter by the REGULATOR LOAD

CIRCUIT SENSE
switching operation of the converter, stated as
OUT
peak-to-peak or RMS.
Figure 10
Reinforced Insulation
Single insulation system applied to live parts Resolution
which provide a degree of protection against For an adjustable supply, the smallest change in
electric shock equivalent to double insulation. output voltage that can be realised by the
Reinforced insulation provides two levels of adjustment.
protection and the test voltage used is 3000VAC
for IT and industrial equipment, and 4000VAC Resonant Converter
for medical equipment. A class of power converter topology which
reduces the level of switching losses by forcing
Regulation either zero voltage across, or zero current
The ability of a power supply to maintain an through the switching device when it is turned
output voltage within a specified tolerance as on or off.
referenced to changing conditions of input
voltage and/or load. Return
The name for the common terminal of the output
of a power supply; it carries the return current
for the outputs.
137
Glossary
Reverse Voltage Protection supply is designed or tested to withstand, such

A feature which protects a power supply against as 20G shock for 11 milliseconds and 10G
a reverse voltage applied at the input or output random vibration for 2 hours over a 2-2000 Hz
terminals. bandwidth.
RFI Short Circuit Protection

See Radiated Electromagnetic Interference. A feature which limits the output current of a
power supply under short circuit conditions so
Ripple and Noise that the supply will not be damaged.
The magnitude of AC voltage on the output of a
power supply, expressed in millivolts peak-to- Signals
peak or RMS, at a specified bandwidth. This is Output interface, often at TTL level, of various
the result of feed through of the rectified line operational conditions such as power fail and
frequency, internal switching transients and other DC OK.
random noise. See also PARD & Noise.
Sixteenth Brick
Rise Time An industry standard package size and pin-out
The time required for the voltage in a switching for DC/DC converters. The package size is 1.3”
electronic circuit to rise from 10% to 90% of its x 0.9” with the pins on a 1.1” spacing. The
nominal final value. height is typically less than 0.4”.
Safety Approvals Soft Start

Third party or agency approvals to internationally A technique for gradually activating a power
recognised safety standards. supply circuit when the power supply is first
turned on. This technique is generally used to
Safety Ground provide a gradual rise in output voltages and to
A conductive path to earth that is designed to limit inrush current.
protect persons from electrical shock by
shunting away any dangerous currents that Standby Current
might occur due to malfunction or accident. The input current drawn by a power supply
when shut down by a control input (remote
Secondary inhibit) or under no load.
The output section of an isolated power supply
which is isolated from the AC mains and Start-up Rise Time
specially designed for safety of personnel who The time between the output voltage starting to
might be working with power on the system. rise and reaching the desired level.
SELV Start-up Time (Start-up Delay)

Safety extra low voltage. A term generally Time between the application of input voltage
defined by the regulatory agencies as the and the output voltage being within regulation.
highest voltage that can be contacted by a
person and not cause injury. It is often Supplementary Insulation
specifically defined as 42.4 VAC or 60 VDC. Independent insulation applied in addition to
basic insulation in order to provide protection
Sequencing against electric shock in the event of a failure of
The desired order of activation of the outputs of basic insulation. Supplementary insulation
a multiple output power supply. provides one level of protection and has a test
voltage of 1500VAC.
Shock and Vibration
A specification requirement for which a power
138
Glossary
Surface Mount Technology (SMT) Topology

A space-saving technique whereby special lead- The design type of a converter, indicative of the
less components are soldered onto the surface configuration of switching transistors, utilization
of a PCB rather than into holes in a PCB. The of the transformer, and type of filtering.
parts are smaller than their leaded versions and Examples of topologies are fly-back, forward,
PCB area is saved. half-bridge, full-bridge, and resonant.
Surge Tracking
Part of the conducted immunity suite of tests, A characteristic of a dual or other multiple
designed to simulate a nearby lightning strike. output power supply whereby one or more
outputs follow another output with changes in
Switching Frequency line, load and temperature, so that each
The rate at which the DC voltage is switched on maintains the same proportional output voltage,
and off during the pulse width modulation within specified tracking tolerance, with respect
process in a switching power supply. to common.
Synchronous Rectifiers or Rectification Transient Response

A circuit arrangement where the output rectifier The time required for the output voltage of a
diodes of a power supply are replaced with power supply to settle within specified output
active switches such as MOSFETs. The switches accuracy limits following a step change in output
are turned on and off under control and act as load current or a step change in input voltage.
rectifiers. This results in considerably lower
losses in the output stage and subsequently Trip & Restart Current Limiting
much higher efficiency. They are particularly Current limiting circuit which switches off the
useful with low voltage outputs. output when an overload condition is reached.
The unit will then try to restart periodically until
Temperature Coefficient the overload is removed.
The average percent change in output voltage
per degree centigrade change in ambient TUV
temperature over a specified temperature range. TUV Rheinland Product Safety Group. An
independent German organization which tests
Temperature Derating products for safety.
Reducing the output power of a power supply
with increasing temperature to maintain reliable UL
operation. Underwriter’s Laboratories Incorporated. An
independent, U.S. organization which tests
Temperature Range, Operating products for safety.
The range of ambient or case temperatures
within which a power supply may be safely Undershoot
operated and meet its specifications. A transient change in output voltage, below
output accuracy limits, which can occur when a
Temperature Range, Storage power supply is turned on or off, or when there
The range of ambient temperatures within which is a step change in line or load. See Overshoot.
a non-operating power supply may be safely
stored with no degradation of its subsequent Universal Input
operation. A power supply's ability to accept a wide input
voltage range (90VAC to 264VAC) without the
Thermal Protection selection of input range, either manually or
See Over Temperature Protection. electronically (as in auto-ranging input).
139
Glossary
UPS Voltage Balance

Uninterruptible power supply. A power supply The percentage difference in magnitude
that continues to supply power during a loss of between the two output voltages of a dual
AC input power. This is accomplished by means output power supply where the voltages have
of a back-up battery and a DC/AC inverter or equal nominal values with opposite polarities.
DC/DC converter.
Warm-up Drift
Under Voltage Lock Out (UVLO) The initial change in output voltage of a power
A protection system for power converters where supply from turn-on until it reaches thermal
the converter is deliberately shut down if the equilibrium at nominal line, full load, 25°C
input voltage drops below a pre-defined level. ambient temperature.
Some hysteresis is usually present to prevent
the converter oscillating on and off. UVLO is Warm-up Time
usually needed with battery systems where the The time required, after initial turn-on, for a
voltage decreases gradually with time rather power supply to meet its performance
than turning off quickly. specifications.
VDE Zero Current Switching (ZCS)

Verband Deutsche Elektrotechniker. A German See Resonant Converter.
organization which tests equipment for public
safety and emitted noise. Zero Voltage Switching (ZVS)
See Resonant Converter.
• Prefix Codes
Prefix Multipliers
Prefix Symbol Multiplier
exa- E 1,000,000,000,000,000,000
peta- P 1,000,000,000,000,000
tera- T 1,000,000,000,000
giga- G 1,000,000,000
mega- M 1,000,000
kilo- k 1,000
hecto- h 100
deca- da 10
deci- d 0.1
centi- c 0.01
milli- m 0.001
micro- µ 0.000 001
nano- n 0.000 000 001
pico- p 0.000 000 000 001
femto- f 0.000 000 000 000 001
atto- a 0.000 000 000 000 000 001
140
Glossary
• SI Unit Codes
SI Base Units
Quantity Name Symbol
Length meter m
Mass kilogram kg
Time second s
Electric Current ampere A
Thermodynamic Temperature kelvin K
SI Derived Units
Quantity Name Symbol
Area square meter m2

Volume cubic meter m3
Speed/Velocity meter per second m/s
Acceleration meter per second squared m/s2
Current Density ampere per square meter A/m2
Magnetic Field Strength ampere per meter A/m
SI Derived Units With Special Names
Quantity Name Symbol Expression in terms Expression in terms

of other units of SI base units
Frequency Hertz Hz s-1
Energy/ Work/ Joule J Nm m2 kg s-2

Quantity of heat
Power/ Radiant Flux Watt W J/s m2 kg s-3
Electric Potential/
Potential Difference/ Volt V W/A m2 kg s-3 A-1
Electromotive Force
Capacitance Farad F C/V m-2 kg-1 s4 A2
Electric Resistance Ohm Ω V/A m2 kg s-3 A-2
Magnetic Flux Weber Wb Vs m2 kg s-2 A-1
Magnetic Flux Tesla T Wb/m2 kg s-2 A-1

Density
Inductance Henry H Wb/A m2 kg s-2 A-2
Celsius degree Celsius °C K

Temperature
Electric Field Volt per meter V/m m kg s-3 A-1

Strength
141
Index
Term Page Term Page
19” Rack Kits 65 Conducted Noise 93, 95
AC Generator 8 Conduction Cooling 80
AC Input Harmonic Currents 29 Configured PSUs 107
AC Input Voltage Protection 23 Constant Current Limit 47
AC Motor Load 34 Constant Current Sizing 18
AC OK 50 Constant Power Charging 20
AC Power Sources 8, 59 Constant Power Limit 46
AC Resistive Load 36 Control and Monitoring 60
AC Topologies 62 Control Area Network (CAN) bus 55
AC UPS Systems 62 Control Interfaces 51
Active Power Factor Correction (PFC) 31 Control Area Network Bus 55
Active Power Sharing 57 Convection Cooling 74
Airflow 72 Cooling Fan Selection 71
Alternator 8 COTS 125
Ampere Hour Rating 15 Creepage Distance 89
Aperiodic Noise 45 Cross Regulation 42
Apparent Power 33 Crow-bar 49
Applied Parts 92 Current Share 52, 57
Auto Bypass 64 Cycle Life 17
Automotive Batteries 14 DC/AC Inverters 67
Autonomy 62 DC/DC Converter Input Circuit 24
Back Pressure 73 DC Generator 13
Bandwidth Limiting 45 DC Input Voltage Protection 23
Bank 15 DC OK 51
Baseplate Cooling 78 DC Output 42
Basic Insulation 89 DC Power Sources 13
Batteries 14-22, 60 DC Standby Systems 59
Battery Charging 21-22 Delta Connection 11
Battery Internal Resistance 61 Device Characteristic Curve 73
Battery Memory 17 Differential Mode Noise 94
Battery Protection 61 Digital Communication Interfaces 55
Battery Sizing Methods 18-20 Dips and Interruptions 96
Battery Standby Time 18 Discharge Currents 19
Boost Charge 17 Discharge Power 20
Boost Converter 5 Distributed Power Architectures (DPA) 6, 7
Bypass 62, 64 Earth Leakage Current 89, 90
C Rating 15 Earthing 86, 89
California Energy Commission 103 Earthing for EMC 41
CAN Bus 55 Electrical Fast Transient 98
CE Marking 99-101 Electrical Safety 89
Cell 15 Electrolyte 17
Charge 17 Electromagnetic Compatibility 93
Chargers 60, 64 Electronic Protection 27
Circuit Breakers 25-26 Electrostatic Discharge 98
Circuit Breaker Protection 48 EMC Directive 100
Class I System 90 EMC Filtering 96
Class II System 90 Emissions 93
Clearance Distance 89 Enable 51
Common Mode Noise 93 Enclosure Leakage Current 91-92
Compliance, Routes to 100 End of Discharge (EOD) Voltage 16
Conducted Immunity 97 End of Life Factor 16
142
Index
Term Page Term Page
Energy Star 102 Isolated Fly-back Converter 2
Enterprise Network 6 Isolated Signal Outputs 54
Equalize Charge 17 Lead Acid Batteries 14, 21-22
EU Code of Conduct 103 Leakage Currents 91
Extended Battery Packs 65 Legislation 88
Failure Rate 82, 85 Level of Protection 40
Fans 71-73 Lightning Conductors 39
Fast Charge 17 Line Interactive 63-64
Fast Transients 27 Line Regulation 42
Filter Selection 96 Line Voltage 11
Float Charge 17 Linear Power Supply 1
Fold-back Current Limit 48 Load Rating 18
Forced Cooling 74 Load Regulation 42
Forward Converter 3 Low Frequency Ripple 45
Frequencies (Worldwide) 12 Low Voltage Directive 99
Frequency 9 Low Voltage Disconnect 60
Frequency Converters 69 Magnetic Trip Circuit Breakers 26
Front Ends 6 Maintenance Bypass Units 65
Full Bridge Converter 4 Measurement Techniques (Ripple & Noise) 45
Functional Earth 89 Medical Safety 91-92
Functional Isolation 89 Memory 17
Fuses 25, 48 Metal Oxide Varistor 27-28
Gas Discharge Tube (GDT) 27, 28 Military EMC Standard 105
Generators 8 Minimum Volts per Cell 19
Generic Reliability 84 Monitoring 60
Ground Resistivity 40 MTBF 82
Grounding 39, 89 MTTF 82
Half Bridge Converter 4 Mutual Inductance 9
HALT Testing 86 N+M Redundancy 58
Harmonics 29 Nickel Cadmium NiCd 14, 21-22
Harmonic Distortion 29-30 Nickel Metal Hydride NiMH 14, 21-22
HASS Testing 86 Nominal Voltage 17
Heatsinks 77-81 NTC Thermistor 24
Hiccup Mode 47 Off-Line UPS 63, 64
High Frequency Ripple 45 On-Line UPS 62, 64
I2C Bus 55 Open Collector Signals 54
Immunity 97 Operational Insulation 89
Impedance Phase Angle 33 Opto-coupler Signal Output 54
Industrial Batteries 14 Output Filter 44
Inhibit 51 Output Margining 53
Input Circuit 24 Output Protection (DC) 46
Input Considerations 8 Output Regulation (DC) 42
Input Fuse 23, 25 Over-voltage Protection 49-50
Input Protection 23-28 Parallel Operation 56
Inrush Current 24 Parts Count Prediction 85
Insulation 40-41, 89 Parts Stress Prediction 85
Inter Integrated Circuit (I2C) Bus 55 Passive Power Factor Correction (PFC) 30-31
Interfaces 50 Patient Auxiliary Current 91
Inverters 67 Patient Leakage Current 91-92
Inverter Alarms and Controls 68 Performance Criteria 97
143
Index
Term Page Term Page
PFC Boost Converter 6 Shunt Diodes 28
Phase Angle 35 Signal Outputs 50
Phase Shift 34-35 Signals 50
Phase Voltage 11 Silicon Controlled Rectifier (SCR) 70
Phasor Diagram 35 Single Phase AC 10
Portable Batteries 14 Single Wire Parallel 52
Power Dissipated 72 Slow Charge 18
Power Distribution 10 SNMP Adapters 65
Power Factor 33-34 Standby Batteries 14
Power Factor Correction (PFC) 30-31 Star Connection 11
Power Fail 50 Static Converters 69
Power Good 51 Static Switch 62, 64
Power Losses 69 Static Transfer Switches 69-70
Power Share 52 Status Signals and Controls 50
Power Sharing (Active) 57 Stress 84
Power Sources 8 Supplementary Insulation 89
Pressure Drop 73 Surge Rating 68
Primary Power Source 59 Surges 27, 28
Probability 82 Switch Mode Power Supply 2
Product Specific Standard 95 Switching Noise 45
Protective Earth 89 Synchronization Signal 68
Prototype Testing 86 Synchronous Rectification 121-124
PTC Thermistor 49 System Reliability 87
Pulse Width Modulation 2 System Reset 51
Push-Pull Converter 5 System Sizing 61
Radiated Immunity 97 Temperature Compensation 17, 60
Radiated Noise 94, 95 Temperature Derate Factor 16
Radiation Cooling 80 Temperature Gradient 71
Reactive Power 33 Thermal Circuit Breakers 26
Real Power 32 Thermal Management 71
Rechargeable Batteries 14 Thermal Resistance 77
Rectifiers 6, 60, 64 Thermal Runaway 18
Redundancy 6 Three Phase AC Source 10
Redundant Operation 58 Transient Load Response 44
Reinforced Insulation 89 Transient Protection 27
Relay Signal Output 54 Transorb 27, 28
Reliability 82 Trickle Charge 18
Remote On/Off 51 Trip and Restart 47
Remote Sensing 43 TTL Compatible Signals 53
Resistivity 40 UPS 62-66
Reverse Input Protection 68 VME Signals 51
Reverse Polarity Protection 28 Volt free Cards 65
RF Conducted 98 Voltage (Worldwide) 12
RF Electromagnetic Field 98 Voltage Adjustment 52
Ring-free Zero Voltage Switching 124 Voltage Dependent Resistor (VDR) 27 (28)
Ripple and Noise 44 Voltage Programming 53
Rotary Converter 69 Voltage Surge 98
Self-discharge 17 Worldwide Voltages and Frequencies 12
Series Diodes 28 Y (Wye) Connection 11
Series Operation 56 Zero Current Switching 2
Service Life 83 Zero Voltage Switching 2
144
World’s Strongest
Standard Products
Portfolio
With a global network of sales and engi-
neering, XP can provide power
solutions for almost every application.
Complementary to this technical guide,

our 208 page Power Supply Guide
contains full specifications of our
standard product ranges.
Contact us by telephone or via our

website for your FREE copy.
• AC/DC Open Frame

• AC/DC PCB Mount
• DC/DC Converters
• Rack Mount
• Configured
• Desktop
• DIN Rail
www.xppower.com
Full product listings and specifications

can also be found on our website.
www.xppower.com
North American Sales Offices North American HQ
Toll Free . . . . . . . . . . . . . . . . . . +1 (800) 253-0490 Eastern Region . . . . . . . . . . . . +1 (973) 658-8001 XP Power
Central Region . . . . . . . . . . . . . +1 (972) 578-1530 Western Region . . . . . . . . . . . . +1 (408) 732-7777 990 Benecia Avenue
Sunnyvale, CA 94085
Phone : +1 (408) 732-7777
European Sales Offices Fax : +1 (408) 732-2002
Email : nasales@xppower.com
Belgium . . . . . . . . . . . . . . . . .+33 (0)1 45 12 31 15 Netherlands . . . . . . . . . . . . . .+49 (0)421 63 93 3 0
Denmark . . . . . . . . . . . . . . . . . . . . +45 43 42 38 33 Norway . . . . . . . . . . . . . . . . . . . . . +47 63 94 60 18
France . . . . . . . . . . . . . . . . . .+33 (0)1 45 12 31 15 Sweden . . . . . . . . . . . . . . . . . +46 (0)8 555 367 00 European HQ
Germany . . . . . . . . . . . . . . . . .+49 (0)421 63 93 3 0 Switzerland . . . . . . . . . . . . . . . +41 (0)56 448 90 80 XP Power
Italy . . . . . . . . . . . . . . . . . . . . . . . +39 039 2876027 United Kingdom . . . . . . . . . . .+44 (0)118 984 5515 Horseshoe Park
Pangbourne
Berkshire, RG8 7JW
Distributors Phone : +44 (0)118 984 5515
Fax : +44 (0)118 984 3423
Australia . . . . . +61 2 9809 5022 Amtex Japan . . . . . . . . +81 48 864 7733 Bellnix Email : eusales@xppower.com
Austria . . . . . . +43 (0)2282 3144 Vitecom Portugal . . . . . .+34 93 263 33 54 Venco
Czech Rep. . +420 235 366 129 Vums Powerprag Russia . . . . . . . +7 (495)234 0636 Prosoft
Asian HQ
Czech Rep. . +420 377 824 986 Vienna Trading South Africa . . . +27 11 453 1910 Vepac
Finland . . . +358 (0)9 2906 1990 Cool Power Spain . . . . . . . .+34 93 263 33 54 Venco XP Power
Israel . . . . . . . . +97 2 9 749 8777 Appletec 401 Commonwealth Drive
Haw Par Technocentre
Singapore 149598
Global Catalogue Distributors Phone : +65 6411 6900
Fax : +65 6741 8730
Americas . . . . . . . . . . . . .Newark www.newark.com Email : apsales@xppower.com
Europe & Asia . . . . . . . . . . Farnell www.farnell.com

Power Supply Technical Guide - XP (072-152)

Uploaded by

Copyright:

Available Formats

Power Supply Technical Guide - XP (072-152)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Power Supply Technical Guide - XP (072-152)

Uploaded by

Copyright:

Available Formats

AC UPS Systems

A typical software screen shot and CD are shown below.

Modified Sine Wave (quasi-sine)

Nominal Battery Voltage (V) Voltage Ranges (V)

Reverse Input Protection

Alarms & Controls

However, there are some applications incorporating motors/compressors/timers/avionics etc. where

Typical frequency conversions are:

50Hz to 60Hz or 60Hz to 50Hz

• Static Transfer Switches

DC Source Inverter Critical

Static transfer switch system

The basic configuration is back-to-back SCRs on

version. The three-phase and/or three-source STS

During normal operation, SCRs associated with the Alternate

Input power = output power + losses

into the air around that component, it is also

performance of the adjacent components causing

Although modern design aids such as 'fluid

Establish Allowable Temperature Rise

Establish Power to be Dissipated

The load power is 260W

The airflow needed through a system can be calculated as follows:

Air flow (m3/hr) = 2.6 x Power (W)

Linear flow rate (m/s) = Volumetric flow rate (m3/s)

Therefore in our example:

Air flow (m3/hr) = 2.6 x total power dissipated (W)

= 84.5 m3/hr (49.77 CFM)

To convert m3/hr to cubic feet per minute (CFM) multiply by 0.589.

Characteristic Curves for Equipment Pressure Drop

Estimating Back Pressure

The graph above is then used to estimate the back 90

pressure, so for a system which requires 14.2l/s, Pa

Using Fan Characteristic Curves 0

• Cooling power supplies

Example based on an EMA212 power supply

AIR FLOW 3.00 2.55 C14 SCREW 6.35 mm

3” x 1.34” = 4 inches2 or 0.028 feet2

Therefore the air velocity required is:-

12/0.028 = 429 LFM or 2.17 m/s

Evaluation of the application

Temperature Measurements (Ambient ≤50 ºC)

C14 & 42 Temperature (°C)

• Cooling Power Modules

where θ is thermal resistance in °C/W

θSA is thermal resistance

• Ts cs θCS is thermal resistance

where TC = maximum power supply temperature

Example: A power module package must operate in ambient temperature of

TC = TA + PD θCA = 30 + (5 x 11.0) = +85.0 ˚C

TC = TA + PD θCA = 50 + (5 x 11.0) = +105.0 ˚C

TC = TA + PD θCA = 50 + (5 x 3.0) = +65.0 ˚C

Power transistor Baseplate of power supply

External heatsink Ambient

Dissipating the Heat: Heatsink Calculations

TB – TA where TA is the maximum ambient temperature outside of the cabinet

MTBF – Mean Time Between Failures

The mathematics are expressed as follows:

a) If a large number of units are considered,