Gacun SMPS Design Tips
Gacun SMPS Design Tips
Gacun SMPS Design Tips
Before we discuss about the SMPS let's make a special measuring tool to measure
The transformer we wound and the FET we will use. Because this tool plays an important role for
success or not in making a power supply. Actually if we have LCR meter or
The oscilloscope is better, but the device is very expensive & rarely available in the market.
2. Calibrate and wind the coil of the measuring instrument (Transformer & FET Tester)
a. Making a measuring coil
Before continuing to discuss the manufacturing method, it is good that we know the function
from the coil. The function of the coil is to measure Fet
later
Apart from the components written in the picture, try to pay attention to that
there is a picture of the coil, so how do you make the coil above?
The way to make it is: - Prepare a 12mm diameter ferrite rod and enough 0.6mm wire. - Wrap the wire over the ferrite
as much as 40L then make a Center Tap and make it again
40L wrapped clockwise.
Description:
F Frequency
Ra is the resistor from Pin 7 to Vcc
Rb is the resistor from pin 7 to pin 2 & 6
Note: The transformer we make must be 200VDC if measured with the device above
or 500µH if using an LCR meter.
In order not to confuse how to calibrate or test the transformer, you can look at
the following picture.
Gb 1.2 How to calibrate or test the transformer
How to know the magnitude of Rds? Why does Rds need to be known? this
important and you need to know. If Rds is small by itself the price of Id will be large, which
determine the magnitude of the current that flows as a current pulse, this current pulse later
transferred by a transformer. The larger the Id, the greater the force produced.
How to find out the Rds price you should make a tool for a simple Rds test
but accurate, but quite confusing for those who don't understand.
See the tool and how to test the Rds of a FET.
Example:
Vcc is measured to show the value of 12V and Va is 10V and Vb = 2V then Rds FET is ?
Rds = {(Vcc-Va) : Vb} x Ra
= (2:10) x 9,75 = 1,95Ω
CHAPTER II
CALCULATION OF CIRCUITS ON SMPS AND POWER PRODUCED
To get results that are close to measuring accuracy, a test tool is recommended
supplied with the correct voltage of 12V, not more or less, if it is not exactly 12V supplied
power, there is a high possibility that the measurement results are also inaccurate.
If this way you have found the right one then tap/duculi and count
one by one what is the amount, well now you know how many turns to
the primer. how about the other number of turns? Set the mold voltage to 200 Volts
for the input voltage until you get Voltperlilit or lilitpervol or n/v or g/v
after getting n/v, then calculate the number of turns with the transformation comparison.
2. Determine the size of the wire to be used and determine the polarity of the ends
girth
Before wrapping the fixed wrap as explained above, if after untying
it turns out to be 50 laps, so measure the coker first how many layers will be used, for example the size
koker that is available as wide as 2.5cm or 25mm or even bigger, measuring as wide as 3.5cm
or 35mm, then to make a double layer winding, the number 35/50 means the wire used
is ∅= 0,7mm. The purpose is for the circumference to fit tightly around the coker, if you
using 0.6mm wire and 50 turns, then there will be in the first layer
the rest is 5mm, or you can use a multilayer so that ½ of the circumference can be applied
primer first and the size of the wire is 35 : 25 = 1.4mm. Only the secondary and final winding
The 2nd primary wrap continues the first primary wrap 25 wraps and returns 25 wraps
circumference, so that the primary circumference is a constant amount but the size of the wire becomes large.
Based on the number of twists and turns in the discussion above, we can draw a conclusion, if the amount
the windings are the same then the inductance will be the same, double or single, if double means
increase the size of the wire by increasing the amount of wire, if connected in a clear tie
inductance will change because the number of turns is different. maybe it can be better understood about
increase the size of the wire by increasing the number of wires, so the number of turns is constant, only
doubled or duplicated.
Maybe when wrapping there is something disturbing so the wrapping process is stopped
temporary, for example sudden needs or any other matter, you are sometimes confused yourself just now
wrap clockwise or not? confused, will be dismantled almost
finished or even finished, the marks at the ends are not marked, continue the direction
twisted also because the coker is turned upside down so it is dizzy. Often and even always this happens,
What should you do to avoid mistakes or doubts?
If the primary winding is for example 50 windings and the input voltage is 200V
n/v is 50/200 or 0.25 per volt, how much extra voltage is needed? Stay in for 4
only, for example for a 12V fan, the number 12/4 is 3 turns, for example tone control needs 24V then
24/4 as many as 6 turns, well whatever you want is up to you, for example to supply power amplifier
CT60V then the number 60/4 is 15 turns, because CT is 2 times, so make 2x turns.
The calculation above is just an example so it doesn't have to be the primary winding is 50 turns,
but use the reference value of inductance 500µH how many turns, or you can read again
above information.
produces a voltage of around 60Volt then after use is it still 60Volt? If electric welding
the voltage is stable at 60V and 150 Amperes, then the output power is 60x150=9000watts
is it true? Not true. the output voltage will drop to 8 Volts only and the power used
8x150 = 1200Watt. this is the logic, if it is for power amp and only 8Volt, what can sound?
For electric welding that makes the voltage drop is the duty cycle which is regulated by
deadtime, for a fixed frequency, so it is not pwm that regulates the dutycycle,. The smaller
deadtime the greater the current that is infected with a constant voltage, only the current that
change, until current regulation is done by deadtime.
With the description above, it is clear that the power produced by the transformer is not because
the power of the transformer, but the power that enters the transformer, in this case it is the smps driver.
Tips or what efforts should be made to obtain a large output power
from the transformer? The trick that must be followed is to increase the power that goes into the transformer,
for example, the selection of supporting components, the use of high-efficiency circuits up to
the system operates more optimally and resilient.
CHAPTER III
AIR GAP FERTI CORE TRANSFORMER (AIR GAP) AND HOW TO WIND IT
In SMPS types, the use of transformers is very diverse, namely gapped ferrite, non-gap ferrite
or even toroid ferrite, here we will discuss SMPS that works with gap ferrite.
What is a gap? Gap is where there is a gap at the meeting of the upper ferrite and the lower ferrite, maximum
the gap is 2mm, let's see the following picture for more clarity.
a. Measure the coker as in the picture how many millimeters high is the inner coker
the example here measured 27mm
c. Cover with enough insulation, the wrap that was made earlier and we will make it next
the main secondary winding....that is, take a 1mm wire and wrap it 11 times
wrap and leave about 5cm left, cut and release again straighten, and cut the wire 3
again along the measurement... until you get 4 long and measuring wire rods
same...1mm. See pictures.
Gb 3.5 Secondary Winding
d. Wrap 4 wire rods together as many as 11 wraps and mark for separation
the wire will be correct, this is for the main secondary winding. 1mm wire double 2 later
will be the minus and plus poles.
e. If the secondary winding has been made then there are 8 ends of the winding wire and make double 2
and tie it to the legs of the coker with 4 legs...made opposite so that it is easy to
make his CT. From the 4 stanzas, please divide into two, for example, 2 pieces of wire from the beginning of love
mark A 1 and the end is A2, while the remaining 2 pieces of wire are also given
example marks for the beginning of B1 and the end of B2 so as not to be confused in the connection later..
f. Make the 2nd primary winding which will be connected in a draw as much as the target minus the winding
has been made. Example 25 laps target 44 laps, so the remainder is 19 laps. this is for the second primer
and the secondary winding is spread between the 2 primary windings, the purpose of which is to reduce
leakage of induction of magnetism and increase the efficiency of power transfer, wrapping around the winding
The 2nd primer is of course less wire and there is some remaining coker, so wrap it in the middle
just limit the upper and lower edges, the purpose is for the induction to be focused in the middle
gap.
Gb 3.7 Closure of the final winding
g. Cover with insulation and note the picture of the end of the connection coming out of the coker slot
to make it easier to connect...
h. Solder the winding connection of the first primer with the second primer. That is the end of the
connect is the end of the first primary wire to the beginning of the second primary, while the beginning
The first primary will be connected to 300VDC and the second primary will be connected later
Gacu red cable. fold it in so it looks neat.
i. We have wrapped everything necessary and covered it with insulation as tightly as possible
l. Test the transformer with a transformer test tool and determine the frequency of 40-60khz
displayed volts? If the voltage indication is 200V and the frequency is between 40-60khz then the transformer
manufacturing is successful, but if the price is not as recommended then
make a gap in the ferrite to get the price. if this is successful then it's done
make smps transformer
CHAPTER IV
FABRICATION OF SMPS WITH PWM GACUN
depending on the force of the load. Block emi filter if complete then the order is
before entering the emi filter there is a VDR that parallels the network, its function
limiting the source voltage rise (PLN) in the 240VAC range.
The output of the emi filter usually gets a thermistor (PTC/NTC) which is connected to
load (diode bridge). Its function is to reduce the starting current so that
MCB does not trip. After passing through the thermistor, the AC current is directed
with 4 high-voltage bridge diodes in parallel with c pillar
10nF/103, this c function is to bypass/pass the high frequency
causing a cutoff diode, so that the high frequency does not follow the same direction,
unless it is damped/removed by the filter capacitor to ground. That's why at
need c ceramic 104 in parallel with elko filter or thunder in each
AC voltage rectifier. Good low frequency (using a step down transformer)
or high frequency (with chopper/smps transformer).
Rectifier block
The rectifier block works to direct AC current into DC.
The rectifier components consist of diodes and electro capacitors, which
each role is a diode to direct the AC current and then be smoothed
ripple by the capacitor (filter). Common types of rectifiers
using the bridge rectifier method which has the advantage of high isolation
between the DC voltage produced and the input voltage.
The input voltage is around 220VAC after being directed and passing through the electromagnet
turns out to be around 308VDC, because 1.4 x 220.
If the elko on the dry rectifier the voltage 308 becomes unreachable
at the same time there is a ripple, the worst consequence is that the SMPS will be hotter
(because trying to stabilize the output and voltage of the pulse form by Dc
ripple). The easiest way to detect this is to measure the 308VDC voltage
or the appearance of a screeching / screeching sound on the main transformer.
the transformer becomes a magnet for a moment until the transformer is turned off, the transformer will transfer
energy
magnet to the secondary winding until the transformer is turned on again and so on. Will
but not all magnetic energy can be transferred (due to its imperfection
transformer) resulting in the remaining magnets in the core of the transformer,
The remaining magnetic energy directly enters the Transistor/FET
through the legs of the Collector/Drain with a voltage that may be higher than the capability
the work of the Transistor/FET. The main function of the snubber is to
removes the voltage (accelerates demagnetization) in addition to the snubber
also used to determine/adjust the working frequency of the transformer. Because
the nature of cutting the smps is finally called snubber. The main feature of the snubber circuit
is the arrangement of resistors and capacitors in series in general, but there are
also which uses diodes and capacitors arranged parallel to the winding
transformer primary.
Secondary Rectifier
The secondary voltage on the smps transformer is not in the form of AC voltage but
DC voltage in the form of a pulse. The voltage appearing on the secondary at
rectified and filtered to obtain the secondary Dc voltage. Characteristics
rectifier/diode should have Fast rectifier properties eg UF4002 not IN4002.
Fast rectifier is meant to be able to direct the pulse with
high frequency, leveling elko is sufficient using hundreds to 2200µF measurements
only, because the voltage frequency that comes out of the transformer is quite high, so it's fast
merecovery/reload quickly.
For pictures of the snubber and secondary rectifier, you can see the picture below
this.
Block Switching
Because smps uses a working frequency between 30-40Khz and frequency
is not found in the DC voltage, then the smps system must be made
or self-stimulating the pulse, the most common method found
is with the self oscillating method. In this type of smps is like
as a network of high power frequency oscillators. It is not uncommon to find smps
which uses IC to make the pulse, for example TDA8380, TEA2661
or STR-group and so on.
In every oscillator system, an initial voltage is needed as a trigger, which
serves as the initial trigger of the oscillator network to oscillate. Pinch voltage
this appears a few seconds after the smps gets the input voltage (AC). Big
This trigger voltage depends on the type of smps network that is used. For example
on STR-F665x the oscillator will work if the trigger voltage has reached 16v.
Because its nature is only as a trigger, this voltage is not used anymore when
smps is working. In general, the trigger voltage is taken directly from
308VDC through a resistor or transistor start up.
If the oscillation is stable and emits a high frequency then the switching block
will work as a main switch. Transformers generally use
Transistors or FETs. The characteristics of the switcher must be able to withstand the current
collector/drain large enough to withstand the voltage on the primary winding
transformer. This current is not a constant current but a momentary current depending on the width
the pulse that moves.
In addition to current capability, the transistor/FET switcher must have a frequency
enough work to be employed as a switcher.
Because the block we are discussing is already a bundle from start up & switching
which is a gacun product so we don't have to bother making it, you can see the schematic
click below to learn it.
Gb 4.4 Switching Block Schematic
The shape of the signal on the Gate (FET) is very important for us to investigate, because it will
determine the output signal.
Measurements on the Gate FET are as follows:
1. If measured with an AVO DC meter, it will show a result of 0.05Volt only
2. If measured using an oscilloscope, it will look like a picture
For a discussion of how to install this network, we discuss it in the chapter on how to modify
plaster
An additional network apart from the 6 points above is the smps network
protector, but we're sorry we can't display the picture, please
find out at http://google.com
Protection blocks that are important for the perfection of smps include:
1. OVP (Over Voltage Protector)
Functioning to protect from excessive tension, this block will
turn off the smps when it detects a voltage above the smps working voltage
2. OCP (Over Current Protector)
This network works to detect more charges, smps will be off if
detected excessive use
3. OHV (Over Heat Protector
This network works to protect smps if used
cause excessive heat, then the smps will automatically shut down
Almost all the blocks have gone into IC smps, for example STR-w575x, STR-f665x
and so on.
CHAPTER V
MODIFICATION OF GACUN
A lot of people complain or have problems in applying it, actually GACUN already
designed as simply as possible to be easy to use, but because there are many of us
his desire and of course who made this kit does not know your desire, so by the way
whatever kit gacun is forced to satisfy your taste. Tip what you do with the kit
GACUN? If you apply it incorrectly to the spec, of course the glue will fail or fail
don't want to be ruled by you, how can GACUN accept your order? Treat
GACUN deservedly with the menu he likes "Gacun Modification"
Do you know how many GANCU frequencies are used? Maybe not all the gacun
operate with the same frequency, how to know the frequency? And if you
want to change the operating frequency of which part needs to be changed or replaced? Look at the picture
the following.
What needs to be modified? There are 3 locations that need to be adjusted. see the picture, which is given in the box
red is the part that may need to be modified or changed.
Whatever your needs, this is a single fet type, so the size of the single fet is still the same
lose with double fet(halfbridge) and 4 fet(fullbridge).
Gacun operates at 80khz frequency, what about the transformer you made? Do you still remember?
with an inductance value of 500uH? Well the key to success starts from here, is it still memorizing numbers?
200V, well you absolutely know these two numbers and can translate them, because these are numbers
single fet smps application.
If you use L or primary winding of 500uH then the transformer you make can
operated from 40khz to 60khz frequency. A frequency as big as that is actually less suitable for
gacun, then the gacun frequency is changed or modified to be able to operate at a frequency of 40-60khz, use it
formula. Take the number between 40 - 60 ie 50khz. how do you do it? If C10nF we increase
then the frequency will decrease and what is the exact value of C? F = 1,2/R+C
If you want a frequency of 50 khz then C needs to be increased according to the formula to get 16nF, is it
C this big in the store? If there is none you can use 15nF and how much if you use 15nF the frequency? According to
the formula, a frequency of 53.3 khz is obtained. Can this be applied? Of course
can ! let's look at the following picture.
Gb 5.2 R C Gacu frequency
F =1,2 / R C
Gacu standard resistor 1K5Ω
Gacu standard capacitor 10nF
So the standard gacun frequency is 80Khz
Almost everyone who applies gacun complains why gacun is used for
Why does the amplifier drop the bass? Some try to use an optocoupler but installed where and
how to install it?
See the picture above. There are 4 red boxes that need to be modified.
1. Make the frequency 50khz which was originally 10nF changed to 16nF or 15nF.
2. Optocoupler
How to make GACUN out stable? Use feedback that implements an optocoupler
or photocoupler or PC. Where and how to install it? Follow the directions, don't
until it goes wrong, unplug the 20k ohm VR. And the former draw the middle and right legs connect
or in the jumper and then connect with the optocoupler leg no. 3 and leg no 4 connect with
the remaining VR container hole, or the left VR container leg (seen from the top of the front, which is there
trimpot place) and legs no. 1 and 2 of the opto connect to the out voltage reference that will be in
regulation. How many volts do you want. Gacun let me follow you, don't you follow Gacun.
How and what is the out voltage reference?
By installing feedback, the output voltage can be stable, even if affected by bass,
no shaking at all, as long as it doesn't exceed the ability of its power.
3. Force adjustment
That is to change the price of R0,22ohm/2watt, if you buy a gun, don't install it right away
to electricity but test it first with the test tool we discussed in chapter 1, if you haven't done it yet
The risk is not knowing the capability of the FET used, the size of the R price depends on the FET
which is installed, so don't rely on that. use the appropriate FET
Do you want, test the FET to find out its Vds, if you know the Vds then the application will
know
How to install an optocoupler with an output voltage reference.
look at this picture.
Gb 5.3 Optocuppuler installation (voltage reference)
By fitting this reference circuit, you can adjust the output voltage
plaster We don't discuss how much power it has, please try it and
you can prove it yourself.
With the addition of this reference, what is the output voltage of the gacun produced?
look at the diagram above, every 1k ohm there is a voltage of 1.5Volt, how much do you want?
How much R above is used? Just add up R, the minimum voltage is
(10+47)=57x1.5V =85.5V means this is the same as CT42.75V, and the maximum is
67x1.5=100.5V or CT50.25V, if the trimpot position is in the middle then the output voltage is 93V
or CT46,5Volt. With this reference you can arrange the gacun according to your wishes. If
want a different arrangement you can replace R47K according to the voltage reference above.
1K=1,5Volt. Do not change
zenner voltage later the voltage reference will change, so do the same as the schematic
with Dz 15V. Guaranteed you can.
Why install references to (+) and (-) only and not to CT? CT does not need to be connected,
this censors the total tension, if the tension is CT then each wing gets half
the total voltage.
4. Resistor 0,22Ω/2Watt.
What is the function of R 0,22Ω/2watt? Internally there is a comparative that works for
me-Reset pwm, this comparator works after the voltage in RS reaches 1Volt pulse, if R that
installed at 0.22 ohms means the current will be limited to 1/0.22 or less than 5A
pulse To be made into rms ampere, the size is ⅓ of the pulse current up to around 1.8 ampere
rms, and if so then the gacun kit is only able to absorb power of 1.8x220 = 396VA,
and if the transfer system has 80% effectiveness then the maximum output power is 316VA.
Gacun gives the maximum power limit that can be absorbed is 396VA and this is according to
i'm bigger, but if you want more power then you have to
adjust the characteristics of the FET used.
What specifics should be applied? To produce large power choose FET
with high Vds at least Vds=800V besides that the price of Id is at least 10A, if you can
meet the characteristics like this then you can change the size of the original RS
0.22 Ω/2Watt can be reduced to 0.11 Ω or R0.22 Ω by 2 pieces in parallel. This will
increase the pulse current to 10 Amperes and 3.3A rms current until the absorbed power can
reach 800VA and how do we know the price of the Vds FET in the Gacun KIT? You can test
Vds Gacu in this way.
Gb 5.4 Test Vds