INVERTER 12V - 110/220V 600W Part 2/2

INVESTMEN
T 600/

apv-e
sedu 13
Mwiyiai
SMepes
e"MNWS
hMihükkm2kk.aG
NeMggu
CIRCUIT
My social networks:
Youtube: My YouTube channel, where all my videos are
Twitter: @Terrazocultor
Facebook: 100ciaencasa
Instagram: Photos, diagrams, drawings...

1. Previous part: The transformer

2. Reminder: Features of this inverter

3. Need to rebuild transformer

3.1 By the number of turns
3.2 By cable section

4. Scheme
4.1 General description
4.2 Frequency adjustment 50-60 Hz
4.3 Output voltage regulator
4.4 Low battery protection
4.5 fan cooling

5. Preparing the two heatsinks for the mosfets

6. Test on breadboard with 12V 4A source

7. Test and adjust with real 12V batteries

7.1 Output voltage test and adjustment
7.2 Test and adjust frequency 50-60 Hz
7.3 Thermal stability test with high performance at 600W

8. Three last minute changes to the circuit

9. Component List
10. Design and make PCB
11. The inverter box

12. Machining of the 6 woods of the box

12.1 Floor
12.2 Right side
12.3 Ceiling
12.4 Left side
12.5 Frontal
12.6 Rear

13. The cable to connect the inverter and the battery

14. Internal wiring
15. Final test
16. Final words
17. The video

1. Previous part: The transformer

Fig 1. Toroidal transformer for this inverter

In the previous video and blog, part 1/2, we saw the transformer that will be used for
this inverter, the photo on the left fig 1 corresponds to this transformer in its first
version.

Below I put the two links, video and blog:

Inverter 12V - 220/125V 600W Part 1. The TRANSFORMER:

(Video)

(Blog)

In this second and last part we will make the inverter circuit and
see the necessary adjustments for its proper functioning. We will
finish the task by making a box to accommodate all the
components inside, including a cooling fan, and we will have a
compact unit that we can transport wherever we need it.

2. Reminder: Features of this inverter

The characteristics and benefits of this inverter are:

INVESTOR CHARACTERISTICS

Parameter Worth Unit

input voltage 12 V

Output voltage 220/125 V

Maximum power 600 W

Frequency Adjustable 50-60Hz

Automatic output voltage regulation Yeah

Low battery alarm with disconnection Yeah

Transformer type Toroidal

Internal ventilation Yeah

Overheat protection Yeah

Oversized mosfets mentioned Yeah

Fig 2. Inverter Features
3. Need to rebuild transformer

3.1 BY THE NUMBER OF TURNS

In the first version of the transformer the winding was made for
6+6 volts, that is: 12 volts with a central tap, a 6-0-6. However,
this was a mistake on my part, since the circuit works at 12 volts in
each half cycle, which adds up to 24 volts. For this reason, I was
getting over 400 volts on the high winding.

The solution: do the winding again, but with twice as many turns,
since the voltage will be double (24 instead of 12 volts). If before
there were 13 turns of double thread, now there will be 26 turns of
double thread.

The experimental winding of 10 turns and the calculations carried

out in video 1 of 2, that of the transformer, are still valid (as of
minute 15:18 of that video), only that the rule of three will be done
for 24 volts , and not for twelve.

I take advantage of the fact that I am going to rewind and size to

obtain 240 volts instead of 220, and this for three reasons:

1) everything that works at 220 also works at 240

2) 240 volts allow a downward margin if there is high consumption
3) The voltage can be decreased with potentiometers P2 and P3

After applying an inverse rule of three, I obtain the final value to

wind the toroidal transformer: 24 turns of double wire.
 Number of final turns: 24 turns of double thread
with which there will be a winding of 11+11V approx.

3.2 BY CABLE SECTION

The 6 mm2 cable has insufficient section. The transformer was

capable of lighting 12 12-volt, 50W halogens each (600W total, 50
amps) when supplying 220V to the 220V winding. Which is clearly
seen in the video.

However, the other way around doesn't work as well. If voltage is

supplied to the low voltage winding, when you connect things to the
high winding, the voltage drops in the latter if the consumption is
more than 100W, to the point that the inverter collapses. It doesn't
break, but it falls apart and doesn't work because the energy is
dissipated as heat.

Solution: Add wire of the necessary section, approximately 16mm2

in section, which corresponds to a diameter of just over 4 mm, a
really thick cable. I was lucky enough to "find" a company that
serves cable in many formats, and best of all: It also serves it in
small quantities, although the final price is a little more expensive
than buying it in bulk, which is quite normal.

The page is this: Prodin-Ferrite

In the video you can see the calculation to know how much copper
to order, since 1, 2, or 3 Kg of copper are used by weight and at
retail. Our choice will be two Kg.
 Fig 3. 4.5mm diameter copper (16 mm2)

The copper is 4.5 mm in diameter (about 16 mm2 in section) and is

of thermal class F155, which is solderable at low temperature.

Regarding the price, between the copper, the shipping costs, the
reel and the inevitable taxes, we are going for €50 more or less, but
there is no other option...

Summarizing the order:

Enameled copper 4 to 4.5 mm diameter
Thermal class: F155
Quantity: 2 Kg.

The finished transformer with that winding looks like this:

 Fig 4. Definitive version finished. It doesn't look pretty, but it's effective and won't get hot

It still retains the two 220V terminals on the left. And at the top you
can see the three terminals of the 12+12, both the central socket
and the two ends.

The copper is so thick that there was no free space inside the core
to later use the fastening of the transformer whose axis passes
precisely through the inside of it, so I will have to hold it with winds
as if it were a mast. I will also put some wood around it as retaining
walls.
4. Scheme
 FUS60A

Fig 5. Inverter scheme

4.1 GENERAL DESCRIPTION

In figure 5, the scheme. We recognize the TF1 toroidal transformer,
the one we made in the previous chapter and we just modified in
the previous point. It will be the one that takes the 12V AC supplied
by the two mosfets Q1 and Q2 and will convert them to 220/125V.

Each of the two mosfets is responsible for a half cycle of the

alternating current, in a setup known as "push-pull". In this way,
each one manages half of the work, because when one is driving,
the other is not.

Integrated IC1, an 8+8-pin SG3525, is the processor of this circuit.

We see the battery on the left side, it has a series switch and also a
60-70A fuse to protect everything. He

battery positive goes (among other places) to the center point of

the primary winding of TF1
4.2 FREQUENCY ADJUSTMENT 50-60 Hz

The inverter must not only provide a certain voltage, but it must do
so at a given frequency, normally 50 or 60 Hz depending on the
country. Potentiometer P1 serves to regulate this frequency. We
really have to work IC1 at 100 Hz to get 50 Hz at the output (pins
11 and 14). This frequency is also determined by the values of C1
and R2

To obtain 60 Hz at the output, IC1 has to work at 120 Hz

We can measure the output frequency with an oscilloscope or a

frequency meter by applying the test leads to the 220/125V output.

Fig 6. Adjusting and measuring the frequency. Practically 50 Hz

4.3 THE OUTPUT VOLTAGE REGULATOR

P2 serves to regulate the output voltage of the inverter. In reality,

IC1 provides a PWM output on its pins 11 and 14, that is, pulse
width modulation. Varying the width of the pulses varies the voltage
on those pins 11 and 14 that manage the gate of the two mosfets,
which in turn will determine the output voltage.

Taking advantage of the fact that this inverter works in PWM, a

system can be implemented to regulate the voltage automatically.
That is, if the output voltage drops, IC1 will increase the width of
the pulses delivered on pins 11 and 14, thus raising the output
voltage and correcting the problem.

Likewise, if the output voltage goes up, IC1 will decrease the width
of the pulses on pins 11 and 14 and the output voltage will go
down, also correcting the problem.

To do this, we use a sample of the output voltage using a technique

known as "feedback": A sample of the output voltage is taken and
applied to a small 220/125 to 9 volt small power transformer (TF2
in the schematic) and is rectified with the diode bridge D1. The
resulting voltage is applied to pin 2 of IC1 which, together with pin
1 forms an internal operational amplifier that IC1 has. This allows
automatic regulation as described in the previous two paragraphs.

Potentiometer P3 must be adjusted so that on pin 3 there is a

voltage slightly less than 5.1 volts when there are about 230 volts
at the output (we will regulate those 230 volts with P2 previously).

4.4 LOW BATTERY PROTECTION

The battery that will power our inverter will discharge as it is used,
unless it is charged. If there comes a time when the voltage at the
battery terminals drops below 10.5 volts, the battery and inverter
will continue to work for a while longer, but at the cost of a
phenomenon in the battery known as "sulfation", so that battery is
no longer It does not support being loaded or it will do so very
poorly, which means it will be rendered useless.

This should be avoided at all costs, since batteries have a high cost

Fortunately, the IC1 SG3525 has a feature on its pin 10: When we
introduce a positive voltage to this pin, the IC1 stops and stops
working: The battery does not continue to run out and we avoid
sulfation.

We are going to obtain that positive voltage with IC2, an

operational 741.
At its non-inverting input (pin 3) we are going to put a fixed
reference voltage, obtained with the zener diode DZ1. This voltage
will not change no matter how much you change the battery since it
is only 3.3 volts.

In its inverter input (pin 2) we put a fraction of the battery voltage

obtained with potentiometer P5, and this fraction of voltage will
change (decrease) as the battery depletes. This potentiometer
should be adjusted so that the output of IC2 (pin 6) changes from
"low" to "high" when the battery voltage drops below 10.5 volts.

When a high state occurs at the output of IC2, that signal is sent to
pin 10 of IC1, which immediately goes into stop and the inverter
stops. At the same time, the high output of IC2 turns on LED1,
warning of the "low battery" circumstance.

We are going to avoid the following phenomenon: 1) The battery

drops its voltage below the preset voltage (10.5V) 2) The IC2 circuit
acts and cuts off the operation of IC1
3) The battery, when at rest, momentarily restores its voltage
4) IC2 stops sending the stop signal to IC1 and everything works
again
5) The battery's voltage drops again due to consumption 6) Return
to point 1 and so on, in a loop, until the battery is destroyed.

To avoid this loop entry, we make a feedback in IC2 through R9 and

P4. Thus, when a high output occurs on pin 6, part of this voltage is
reinjected into input 3, latching the circuit, which will remain that
way until we turn the inverter off and on, presumably with an
already charged battery.

4.5. FAN COOLING

This does not appear in the general diagram, nor do two of the
three LEDs that are optional, which is why I document it here:

At first we have the question: Where to connect the fan? At 12V

batteries? Or to the 220/125 AC outlet?

It seems more efficient to do it at the input, at 12V, since doing it

at the output of the inverter forces it to do work that will be
subtracted from the maximum power it can deliver.

In some installations I have installed the fan not fixed, but driven
by a thermostat, so that it is only connected when it is really
needed: When the equipment heats up. This time, even more so,
since battery power is limited and it is always good to save. Why
run the fan if we are only consuming 150W and the inverter barely
heats up? In addition, we will avoid noise.

We will use a 12V fan and we will connect it to the 12V input of the
batteries (on the connection strip, and be careful, because this fan
does have polarity ) and we will put a thermostat of about 70º-
100º in series with the fan that we will attach to the mosfet
heatsink.

Fig 7. 12V continuous cooling fan

5. Preparing the two heatsinks for the mosfets

Each mosfet will have to perform at a power of up to 600W,

although only for half the operating time due to the push-pull
configuration. It must also be taken into account that there will be
forced cooling with a fan. But even so, the mosfets have to be
mounted on a heatsink to release the heat.

The measurements of this heatsink are:

Length: 18cm
Width: 7cm

Based on the results obtained, I would dare to say that this

heatsink can be made much smaller, with a length of almost half:
10 cm.
This heatsink also comes from scrap aluminum from metal
carpentry (windows, frames, doors...) at a cost of almost zero. I
have a video dedicated to this: Super-cheap heatsinks.

Fig 8. Aluminum heatsink obtained from scrap in metal carpentry workshops. Efficient and free

6. Breadboard test with laboratory source

I am going to assemble the circuit on a breadboard to check that

the circuit works. First of all, I will supply 12 volts not with a battery
but I will use the laboratory source (Useful Circuits 5) that can
provide those 12 volts up to 4 amps since it is adjustable in voltage
from 0 to 30 volts.

Now we have "only" 4 amps (those provided by the source), so we

should not connect anything to the output that has more than 12V x
4A = approximately 48W. A small 10-25W lamp will be ideal for
testing. I used two: One 24W incandescent and another 11W LED

Four amps is not all that a well-charged battery can provide in

terms of intensity (far from it) but that will come in handy precisely
in case there is a problem: We will avoid burning up the circuit
and/or blowing a 70A fuse in case there is something wrong.

If we do not obtain any voltage at the output (total failure), it is

time to investigate the cause. It will have to be solved. In case of
total failure, it is interesting to look at what intensity the source is
supplying. This will give us clues: If consumption is zero,
(benevolent failure) we will have forgotten some line or contact. On
the contrary, if the consumption is the maximum allowed by the
source, it indicates something more serious: There is a high
consumption despite which there is no outlet. It must short circuit
something. Pay attention to the polarity of semiconductors and
mosfets, that we have not forgotten any connections, etc.

It happened to me that I forgot to connect pin 12 of IC1 to ground,

and the circuit did not work despite which it consumed the
maximum that the source allowed: 4A. If I had put the batteries in,
who knows what would have happened...

I connected that pin to ground and the circuit worked flawlessly.

 Fig 9. Everything ready to test on a breadboard with a 12V source limited to 4A.
Note: The toroidal in this photo is version 1, which has been replaced by the final version in this video.

7. Test and adjust with real 12V batteries

When we are sure that the circuit works by feeding it with the low-
intensity source, we will move on to testing it with batteries. We
have already excluded the risk of short circuit or serious
malfunction.

7.1. TEST AND ADJUSTMENT OF OUTPUT VOLTAGE

With the circuit in operation, and the 24W lamp connected to the
output of the inverter, that is, to the secondary of TF1, we put a
selected multimeter in alternating current. We must obtain about
240V (125V depending on the case).
We put P3 in its central route more or less.
We regulate P2 to obtain the desired voltage at the output
From here, we will regulate with P3 to correct the tension.

Fig 10. Voltage at the output of the toroidal: 227 volts with a load of 400W (halogen)

7.2. FREQUENCY TEST AND ADJUSTMENT 50-60 Hz

Potentiometer P1 is responsible for adjusting the working frequency

of this inverter. It may be interesting to put a multiturn
potentiometer, to have greater precision in the adjustment.

I made the measurement with the ammeter clamp connected to the

output of the inverter, at 220V. This clamp also measures
frequency. It cost me a little bit, but it is proving to be a pretty
serious tool.

I move P1 until I get 50 Hz or a very close value, like

49.98 Hz which also works. The frequency remains quite stable.

7.3. THERMAL STABILITY TEST WITH HIGH PERFORMANCE 600W

In a similar way to how I tested the toroidal transformer, now I will

test the entire inverter, including the heatsink, the mosfets, the
cables... everything, with a consumption close to the maximum,
between 500 and 600W, and thus make sure that the inverter is
viable and it will not break due to heating.

To monitor the operation I had three instruments: - Polymeter on

the battery terminals, to see how its voltage decreases
- Polymeter measuring temperature, that of the mosfets heatsink
- Clamp ammeter measuring the intensity consumed
Fig 11. Everything prepared for the FIRE TEST. Almost seven minutes of uninterrupted operation at 600W

The test was a resounding success. The mosfets, consuming 50

amps, barely reached 38ºC after six and a half minutes that the
small battery lasted. In reality, the battery would have lasted
longer, but when the battery voltage dropped to 10.6 volts, I
activated the P5 potentiometer, looking for the stop of the circuit to
adjust the "low battery" function, which also worked.

The transformer barely heated up. The cables, cold. Anyway, things
look very good.

8. Three last minute changes to the circuit

I had to make three changes to the circuit to solve some problems

that arose. The circuit in figure 5 is the final one, with those
changes included.

Change ONE : The resistor R10 that limits the intensity that passes
through the red LED LED1, the one that signals the "Low Battery"
condition, was initially 1K, I raised it to 4K7 to prevent LED1 from
shining without said condition being activated.

In any case, LED1 does not work well, because when the "low
battery" situation occurs, the IC2 chip sends its output pulse from
pin 6 to pin 10 of IC1, which immediately obeys, producing a stop.
However, LED1 does not light up until the battery voltage drops
even further, which will not happen since the system has been
stopped by IC1 and the battery will not be used up any more.

Change TWO : Resistor R2 also originally 1K, I raised it to 10K.

Here something a little strange was happening: With the original 1K
resistor and feeding the inverter with the source (not the battery)
the frequency obtained was correct, and could be regulated with P1

But if I powered the inverter with batteries, it could not go below 83

Hz even if I set P1 to the limit. This problem disappeared when I
changed R2 from 1K to another 10K resistor, so the frequency could
be adjusted within a range that allowed values of 50 and 60 hertz.

Change THREE : The DZ2 zener diode was 3V3, I changed it to

4V7.
The reason: The 3V3 diode allowed leakage from pin 6 of IC2 which,
applied to pin 10 of IC1, made the "stop" function work at all times.
This is because an ideal operational amplifier is not the same as a
real one (IC2).

One of the differences between an ideal one and a real one is that
the output on pin 6 is not actually "zero" when it is zero. There is a
small residual voltage, which, when applied to IC1, caused it to
stop.

By removing the 3V3 zener and putting in a 4V7 one, the problem
was solved.

Note: Low battery protection circuit is not mandatory. It's optional.

The inverter will work just as well without it, although without said
battery protection feature. If you choose not to put it, pin 10 of IC1
must be grounded. If left unconnected, open, the inverter will
operate erratically.
9. Component List

I will divide this list into three sections:

1. To make the box

2. For the PCB
3. Annexes

1. To make the box

- 6 wood cut according to this sketch:

 220

3 roof

6
Rear

Floor
1

5
Frontal DM type wood
1, 2, 3 and 4 in thickness 16 mm
5 and 6 in thickness 4 mm
[ 252 All dimensions are in mm

Fig 12. Measurements of the wood to make the inverter box

- 4 wooden blocks of 60 mm x 60 mm in 16 mm thickness to

attach to the toroidal transformer as retaining partitions. (if you can
use the trafo fastening hardware you do not need these plugs)

- Hardware: You can change to your liking, but I used:

- 36 screws to join woods 1, 2, 3 and 4 and make the box. 9 on
each edge
- 20 screws to join the front [5]
- 20 screws to join the rear [6]
- 8 screws to hold the 4 60 x 60 plugs (2 for each one)

- Six rubber feet for the base

2. For the PCB:

R1 Resistor 22Ω (red-red-black)

R2 Resistor 1KΩ (brown-black-red)
R3 Resistor 4K7Ω (yellow-violet-red)
R4 Resistor 1KΩ (brown-black-red)
R5 Resistor 100Ω (brown-brown-brown)
R6 Resistor 100Ω (brown-brown-brown)
R7 Resistance 22Ω (red-red-black)
R8 Resistor 1KΩ (brown-black-red)
R9 Resistor 100KΩ (brown-black-yellow)
R10 Resistor 1KΩ (brown-black-red)
R11 Resistor 10KΩ (brown-black-orange)

All resistors can be 1/2 W

P1, P2, P4: 10K adjustable potentiometers

P3, P5: 10K linear potentiometers (with knob) for recessing

C1 Polyester capacitor MKT 1µf/50V

C2, C3, C4, C5 Polyester capacitor 100 nf /50V
C6, C7 Electrolytic Capacitor 4.7µf/25V

D1 Rectifier bridge with 4 diodes. Little. With 1A support, okay.

D2 Diode 1N4007

DZ1 1/2W 3V3 zener diode (can be 1W)

DZ2 1/2W 4V7 zener diode (idem)

Q1-Q2 Mosfet IRF1405 or any other with similar performance

IC1 Integrated circuit SG3525

IC2 Operational integrated circuit 741

TF2 Transformer 220/125 (depending on country) to 9V.

This transformer can be very small, a 9V secondary winding that
supports only 0.2 Amp is sufficient.

1 x 120 x 80mm single-sided blank PCB board

At the last minute (it is not included in the diagram) I included a 1A

fuse for the PCB with its corresponding printed circuit version fuse
holder.
3. Annexes :

- The 600W toroidal transformer that we made in the previous

chapter
- The support for that transformer (or if it cannot be, the 4 60x60
plugs)
- 12cm x 12cm fan for 12V continuous
- M10 screws for 12 volt input from the battery:
- 2 screws long about 70 mm
- 4 grower washers
- 4 nuts
- 4 wide wing washers
- 4 large ring terminals, through which hole the M10 screw
passes
- Fuse holder and 70 amp fuse for 12V input
- Fuse holder and 2.5-3A fuse for 220V output (6A if
you use 125V)
- Red LED1 (low battery function)
- Green Led2 + 680Ω resistor to put it in series
- Blue LED3 (220V version) or failing that, a neon.
- Recessed base for 220V/125V
- Switch for 220/125V output
- PCB pieces to make connection strips
- Optional: Two sturdy handles for carrying the inverter
- 16 mm2 section cable in red and black colors
- "Normal" section cable for connections other than 50 Amps.
- Cable clamps for batteries that actually have a section of 16mm2

10. Design and make the PCB

I already know that at this point I should be making the circuits

with a KiCad-type design program and using methods such as sun
exposure, or at least ironing, but... lack of time. Someday it will
come...

In the meantime, I will use, yes, once again, the marker method,
which by the way has given me more than good results in the
circuits I have built so far, including some with several integrated
ones.

Perhaps the marker pen method is a little more laborious, but it is

also true that much less means are needed (no computer, no
software, no knowledge of how to use them, no sunscreen, no
special products...). The chickens that come in, the ones that come
out...

I am not going to detail in detail the process of making the PCB in

each video, I already did that in the Basic Electronics Tutorial
chapter 15 , but I am going to show below the most relevant photos
of said process to facilitate the assembly of this circuit:

INITIAL OR "ROUGH" DESIGN:

Fig 13. Initial design

We trace the tracks, trying not to use bridges or jumpers. Here we

will also determine the ideal size of the PCB which turned out to be
120 x 80 mm

You must leave space in the four corners for the 3 mm drills and not
close the margins of the PCB with the components.

SCALE DESIGN
 Fig 14. Definitive scale design

We converted the initial design to this one, now to scale. With the
help of graph paper in tenths of an inch, the standard distance
between pins on the components.
PADS AND TRACKS ALREADY LABELED

Fig 15. Tracks and pads labeled on the copper side

Tracks and pads drawn with permanent marker on the copper side.
You have to make two or even three strokes of the marker so that
the line is able to withstand the action of the acid, leaving a minute
between strokes.

No matter how good your eyesight is, it is good to help yourself

with a magnifying glass: Your hand pulse will improve and the line
will be of higher quality.

COPPER TRACKS, TIN PLATED

 Fig 16. PCB attacked by acids, tinned tracks

The copper side, tinned, to give it more resistance over time,

improve conductivity and "cover" imperfections in the tracks.

On that occasion, some tracks did not come out as clean and clear
as on other circuits. Tin plating solves it.

SCREEN PRINTING (COMPONENTS SIDE)

 Fig 17. Homemade screen printing

It will always be clearer when assembling or repairing the PCB with

good indications about the components: Location, polarity, code,
values...

THE CIRCUIT, SEEN AGAINST THE LIGHT (COMPONENTS +

TRACKS)

Fig 18. In case it helps: PCB showing component side and track side simultaneously
...AND THE CIRCUIT ALREADY ASSEMBLED

Fig 19. Circuit finished. Beware of the typo: The IC1 chip must be the opposite of how it is seen in this photo

The previous photo, figure 19, contains an error: The IC1 chip is
upside down.
As soon as I put the last component in I took several photos of the
PCB. Later, before connecting it I gave it a once over and
discovered that IC1 was backwards. I put it in correctly (easy
operation thanks to using a socket instead of soldering it), but I
forgot to take new photos of the PCB...

11. The inverter box

We will make the inverter box by joining the 6 woods with screws.
It is advisable to list (or name) each of the woods because,
although they come in equal pairs, the marks and drills that we
make will not be exactly the same, and if we exchange them we will
have adjustment and assembly problems. In the photo you can see
that they are numbered inside so as not to confuse them.

If we make the investor look (part of 220 towards us) we have

Wood 1: Floor
Wood 2: Right wall
Wood 3: Roof
Wood 4: Left wall
Wood 5: Front: 220V output
Wood 6: Rear: 12V input

I label those numbers on the wood not only to identify each one of
them, but also (and very important) to know which side goes on
which side, and in what position the wood goes. This identification
of the position of each wood will also help us give a better finish to
the box. For example, it is not the same to start drills on one side of
the wood or the other. The side where the drill bit comes out tends
to be uglier than the side where it enters, which is much cleaner
and more defined.

The way to join these woods: Each one, the method that he likes
best, but in this assembly we have a fairly heavy component: The
toroidal transformer. We cannot make concessions in excessive
savings, nor allow ourselves to make insecure unions, they must be
robust. I will use my preferred method:

a) Through drilling to the sides. 9 holes down, 9 holes up

b) I put the floor and the sides together, as they should be

c) Through the hole made in the sides, frame on the ground

d) I drill into the ground so the screw doesn't have to go through

(the wood can break)

e) I repeat the operation for the ceiling, turning everything over so

I can put said ceiling on the table, as if it were the floor.

f) I present the front and rear woods, which are thinner, and fix
them directly when the floor, ceiling and walls are joined and the
screws are well tightened.

If the wood was cut on a robotic saw that makes the cuts perfectly
square and yet when you go to put the front or back you see that it
is not square, tighten the screws on the sides well: Everything It
will square perfectly. It happened to me. After tightening the 36
side screws, it was perfectly square, and then I presented, marked
and drilled the front and rear holes.

Fig 20. Woods 1, 2, 3 and 4 assembled

Once the box is assembled to ensure that it looks good, we
disassemble it to begin assembling all the elements since all the
woods have something to do to them. We see it in the next point.

12. Machining of the 6 woods of the box

12.1 FLOOR
 Fig 21. Rubber floor and feet for good footing

I recommend fixing six rubber feet on the outside, on the wood [1], on the side that
will touch the floor, to prevent the inverter from sliding or scratching the surface. One
leg in each corner and two in the center. Pay attention to where we place these legs
that are secured with screws: They should not interfere with other drills already made
in this wood, or drills that we are going to make. Be careful with this error.

In this same wood, on the top side they will go (Photos 22 and 23):

- 4 pieces of PCB as connection boxes. Two for the input of the

battery current (+) and (-) and another two for the alternating
current delivered by the toroidal transformer.

- the toroidal transformer (fastened with thin but strong strings),

and also fastened with 60 x 60 mm wooden blocks

- The heatsink with its two mosfets + the thermostat

- The inverter circuit

Fig 22. Wood "floor" with its fixed elements. Left side (as viewed from the front)
 Fig 23. Wood "floor" with its fixed elements. Right side (as viewed from the front)

12.2 RIGHT SIDE

We are not going to attach any device to this wood, but we will
make a matrix of drills to serve as interior ventilation, especially in
case the fan starts up.
 Fig 24. Wood 2 on the right side (and 4, on the left side) with its matrix of holes for interior ventilation

12.3 CEILING

Fig 25. Fan attached to wood No. 3, the ceiling

The ceiling or wood number 3 has the fan (placed on the inside).
You also have to make a matrix of drills that allows air to pass
through. The fan will be fixed with four screws of sufficient length
with their respective normal washers, grower washers and nuts.

12.4 LEFT SIDE

Same as wood 2. It does not have any component attached, we will

also make a grill of holes so that the air that the fan forces into the
box can escape (The fan is mounted on the ceiling). See figure 24.

12.5 FRONTAL

This wood has a good number of attached elements to be able to

control and monitor the operation of the inverter:
 Fig 26. Wood "5" Front, with the 10 components attached.

- DC voltmeter with scale for 15 volts

To monitor the battery status. It will be of great help to us to adjust
the "Low battery" function.

- AC voltmeter for 300 volts

We will know at all times what voltage we have in the AC line.

- Switch.
To connect/disconnect the AC outlet.
If we are not going to use the inverter for a long time, it is better
that we disconnect the battery clamps. Do not forget that the
inverter has a small consumption (about 100 mA) when nothing is
connected, but this small continuous consumption can end up
discharging a battery in a matter of a few days (unless it is being
charged simultaneously by some energy source, such as solar or
wind).

- AC Fuse
3 amps if you use 220-240V
6 amps if using 125V
In addition to the 70 amp fuse for the battery circuit, I find this
additional protection convenient. There is a third fuse on the PCB of
only 1 amp for said PCB. This inverter therefore has three fuses.

- Built-in AC socket
To connect the devices that we are going to operate, or a multiple
socket strip to connect several things. If it is going to be used to
electrify a house, this outlet can be connected to an electrical
control panel with differential and all the protections we deem
appropriate.

- The two potentiometers, the voltage regulation one (P3) and the
"Low Battery (P5)" trigger point regulation. Fastened with their
respective nuts.

- The three LEDs: The red one for the "Low Battery" function, is
connected with two cables directly to the PCB connector. The green
LED for 12V presence has a series resistance of 680 Ohm. The third
blue LED, for the presence of 220/125V AC, can be of the 220/125V
type or you can use a normal 3 volt one (as I have done) but
putting a 1N4007 diode in antiparallel and in series with both a
470K resistor

In the following figure 27, the front with the holes to house the
components. Having a 38 mm diameter crown drill saved me a lot
of marquetry work: I used it to make the large holes for the two
voltmeters and the 220 socket. All that was needed was a few small
touches with a file.

What you see in the outline of that wood are nails, but they are only
to mark the position of the drills in woods 1, 2, 3 and 4. After these
nails are removed, the pre-drilling is done
in said wood, and finally screws are put (and not nails).

Fig 27: Drills for the instruments in the wood "5" Front

12.6 REAR

Here the connectors will be placed to plug in the 12V battery and
introduce that voltage into the inverter. After several searches it
became clear to me that a good solution is to use a combination of
screw, washers and nuts.

It is a cheap system, it is fail-safe (the clamps here were dangerous

to me, as they can come loose and create a short circuit). The
banana connections didn't convince me either.

For each of the two poles:

ON THE INSIDE:
- A metric 10 screw
- A grower washer
- A ring type connector whose ring allows the M10 screw to pass
through
- A wide brim washer

ON THE OUTSIDE:
- A wide brim washer
- A nut, to firmly fix the screw to the wood.
- Other ring connector (from battery cable)
- A grower washer
- A nut

The battery cables have battery clamps at one end, and ring
terminals at the other end that we will attach to the previous
assembly.
Fig 28. Rear part with the connection terminals for the cables coming from the battery
Fig 29. Detail of the connection coming from the battery. Red positive wire. Cable section: 16 mm2
Fig 30. Connector used for 16 mm2 section battery cables. Its large hole allows the M10 screw to pass through.
Its bushing has an ideal diameter to accommodate and solder the cable. I prefer to weld it instead of riveting it.

13. The cable to connect the inverter and the battery

You have to get a pair of cables to connect the battery to the

inverter.
The section must be at least 16 mm2 to avoid losses.
The length is up to each individual, allowing it to reach from the
battery to the inverter, but always looking for the minimum possible
also to minimize losses.

I built myself a pair of cables, one red for the positive and one black
for the negative. At one end of the cable, large sized clips to apply
to the battery. At the other end, to connect to the inverter, a
soldered ring terminal, like those seen in point 12 above.
14. Internal wiring

All that remains is to join the different elements using cables:

It seems not, but making a decent internal connection is not that

easy. First, the connections must be good, both mechanically and
electrically. They do not oxidize due to heat and do not generate
resistance over time. They should be easy to put on and take off.
And reliable over time. Among the many options, I am going to use
the same system that I used in the "control panel" of my biodiesel
factory, which has never given me the slightest problem:

It consists of attaching small square or rectangular pieces of printed

circuit to the wood floor (using one or two screws), with the copper
facing upwards, almost completely tinned. In each of these squares
we will solder a specific cable and from there we will branch to
where necessary.

We need to make FOUR pieces of printed circuit:

- Battery positive input

- Battery negative input
- One pole of the alternating current (220/125V) coming out of the
toroidal
- The other pole of the alternator (220/125V) that comes out of the
toroid

Fig 31. Connection "box" for the positive pole

We fasten these pieces of printed circuit with two screws to the

wooden floor No. 1, and in those pieces of printed circuit we will
make the connections to the different points of the inverter.

In figure 31, on the left, one of those pieces of PCB to make the
connections, specifically the one for the positive pole of the battery.
Well tinned.

For example, on the positive battery input pole we have six cables:

- A 16 mm2 cable that is the battery positive input

- Another 16 mm2 cable that will go to the central socket of the
12+12 of the toroidal
- A normal-small section cable to the positive of the PCB
- Another normal-small section for the (+) of the voltmeter
- Another normal-small section for the (+) of the cooling fan
- Another normal-small section for the LED anode
 Fig 32. Left side view

15. Final test

The inverter is tested again with a group of devices, and they work
without problem. The devices have been:

- 24W incandescent lamp

- Same, 100W
- 550W drilling machine
- Brushless induction motor for 220V
- Mobile charger
- A DVD player
 Fig 33. The investor, finished.

16. Final words

The power of an electric current is given by the product of its

voltage and its intensity. Therefore, it is the same, in terms of
power:

12 volts at 50 amps = 12 x 50 = 600W

48 volts at 12.5 amps = 48 x 12.5 = 600W

However, in the case of working at 48 volts the intensity is much

lower, which allows us to use cables with a smaller section, and
everything is simplified and much cheaper. This is the reason why
electricity is transported at hundreds of thousands of volts, and not
at 220V

So... Why the inverter at 12 volts and not at 48 volts using four
batteries in series adding their voltages 12+12+12+12 = 48 volts?

Because most of us who have access to a battery have access to

only ONE: The one in our vehicle. So, for reasons of "public utility" I
chose the 12 volt voltage knowing that working with medium
powers - such as 600W - was going to bring the complications of
working with cables of considerable section.

It seems that there is not much difference between working with a

"normal" small section cable and one of 16 mm2, but you have to
get to it to really see that things get quite complicated:

With 16 mm2 cable everything tends to be filled with cable, they

take up a lot of space, they are difficult to manipulate, while you
strip a normal cable "with a stroke of scissors or pliers", stripping
one of these is more laborious. They also need a lot of heat (and
tin) to tin and solder them. The terminals and contacts have to be
equally large, which clashes with the small size of the mosfets they
power. In addition, the price increases considerably for the cable of
these sections. When you go to the store to get a "normal" cable,
you don't care much about the price, but the 16 mm2 cable costs 5
euros per meter and you already bother to measure so as not to
spend more.

But this is the price you have to pay for having a moderately
powerful inverter for a voltage as normal and available as 12
volts...

PROVIDE 220V ELECTRICITY TO A HOUSE

Especially if it is a house that does not rely on electricity for most of

the appliances. Large consumption appliances are excluded, such as
air conditioning, electric ovens...

But up to 600W many others can be used:

- Lightning. If low-consumption lighting is interesting, in this case it

is of the utmost importance. With LED lamps you can provide
lighting to a medium-sized house for perhaps less than 200W

- Refrigerators or refrigerators. The most modern high-efficiency

ones barely consume a little more than 100W, so there is plenty of
inverter...

- Audio, TV, computing: Since it can power a DVD, based on a

switching power supply, all of these equipment will also work. None
of them consume anywhere near 600W. So you can play your
favorite music, enjoy your PC (whether a tower or a laptop) or
watch TV

- Motors: If you need a drill, radial, or electric saw type tool, you
can use that inverter. Many of these tools consume below 600W. I
also include smoke extractors, fans...

A resource to have double the power (1,200W) without having to

make this larger inverter is... to make two. You can use one for
lighting, and the other for sockets. Or you can divide the installation
into two parts, and use an inverter in each part.

For those who want to make a more powerful inverter, for example
one of 3000 W, with which you can now install washing machines,
ovens, etc., I have good news: The size of the toroidal core
necessary for double power is not double but much more. minor.
When one nucleus has dimensions twice as large as another, its
power is sixteen times greater!!

Let's see an example:

The core that I have used in this work has these measurements:
External diameter: 13cm
Internal diameter: 7 cm
Height : 6cm

With these measurements we can determine the power of that

nucleus, which is equal to the "Square of the Cross Section Area of
the nucleus". Figure 34 will clarify this a little more. Be careful, you
must always use cm (and not mm) as a measurement for the
calculations to be correct:

Fig 34. Power in W of a toroidal core. It is equal to the square of its transverse surface.
and with these measurements we have 600W. How much power will
we have with a core twice as big? Double the power, 1200 W
instead of 600W?
No sir... If we apply the formula to a core with measurements twice
as large, we would obtain 9600W!! (16 times more)

So, to get 3000W, which is 5 times more than 600? What size
should I choose for the core?

Well, for example:

External diameter: 23cm

Internal diameter: 12 cm
Height : 10cm

As you can see, it's not a mastodon either...

If I were to make an inverter for that power of 3000W, clearly I

would play it safe and work at 120 volts direct current (ten batteries
in series), with which the intensity would be slightly less than 25
amps to achieve 3000W, which allows work with 3.5 mm diameter
wire in the toroidal winding and high current connections.

One complication is that 120V battery voltage is already a

dangerous voltage, especially in humid environments. A battery
park with these characteristics should already be provided with
protection against accidents.

You would also have to use higher performance mosfets, which of

course are not going to cost just €5 each like the IRF1405 of this
600W project, but if you want something, it costs you something...

I do not see any impediment to using this same circuit in a

hypothetical inverter at 3000W and 120VDC (provided that the
mosfets are suitable for 120 volts) and that the power supply for
the positives of IC1 and IC2 would not, of course, be

120 volts, but it would take a shunt from ONE battery to power it at
12 volts.
17. The video

Do you want to support this channel?

Become a patron of Terrazocultor and help him create more quality content.
Every contribution is welcome, whether big or small. -- Help me

My social networks:
Youtube: My YouTube channel, where all my videos are
Twitter: @Terrazocultor
Facebook: 100ciaencasa
Instagram: Photos, diagrams, drawings...