250 to 5000 Watts PWM
DC/AC 220V Power Inverter
This is a heavy duty design of a Pulse Width Modulator DC/AC
inverter using the chip SG3524 . I've been using it as a backup to power up all
my house when outages occur since aprox. 6 years non stop.
Notes:
1> The schematic circuit design is for a 250 watt output, to increase the power of the circuit you have to add more of the Q7 and Q8 transistors in parallel, each pair you add will increase your power by 250 watts, eg. To get 750 watts of power from the inverter you need to add in parallel 2 of Q7 and 2 of Q8 to the original design.
1> The schematic circuit design is for a 250 watt output, to increase the power of the circuit you have to add more of the Q7 and Q8 transistors in parallel, each pair you add will increase your power by 250 watts, eg. To get 750 watts of power from the inverter you need to add in parallel 2 of Q7 and 2 of Q8 to the original design.
2> If
you increase the power transistors you have to enlarge the T2 transformer to
match the new needs, the circuit's transformer is rated 25 amps to handle 250
watts of 220v, for every 1 additional amp you need on the 220v side you have to
increase 10 amps on the 12v side, of course there are limits to the thickness
of the winding so if you need more than 750 watts i recommend that you use a
24VDC supply instead of 12 volts:
DC voltage and Transformer "T2" size
recommendation:
Power Supply Transformer
Winding
|
||||
Wattage
|
Battery
|
Primary coil
|
Secondary coil users define
|
|
750w
|
12VDC
|
24V "12-0-12
|
220V
|
|
1500w
|
24VDC
|
48V "24-0-24
|
220V
|
|
2250w
|
36VDC
|
72V "36-0-36
|
220V
|
|
3000w
|
48VDC
|
96V "48-0-48"
|
220V
|
|
3750w
|
60VDC
|
120V "60-0-60
|
220V
|
|
4500w
|
72VDC
|
144V "72-0-72
|
220V
|
|
5250w
|
84VDC
|
168V "84-0-84
|
220V
|
|
*The transformer should be
"center tapped" at the primary side.
**You can make the secondary 110v if needed but here 220v.
**You can make the secondary 110v if needed but here 220v.
Note18:
Test your transformer before doing this project. Disconnect the transformer
from anything it is attached to, connect the 220v Secondary side directly into
a 220vAC outlet and test the Primary side with your voltmeter, you should get
exactly the voltage necessary for this project as per the table above. If not
then don't waste your time building the project, it will not work.
****Do not supply the driver circuit with more than 24VDC
max. because the voltage regulator "7812" will burn.
3> R1 is to set the PWM duty cycle to 220v. Connect a voltmeter
to the AC output of your inverter and vary VR1 till the voltage reads 220V.
4> R2 is to set the frequency to 50 or 60
Hz (R2 range is between 40Hz to
75Hz), so guys that do not have a frequency meter are advised to blindly put
this variable resistor mid-way which should drop you in the range of 50~60 Hz.
If you want, you can substitute the variable resistor with a fixed resistor using the following formula: F = 1.3 / (RxC) in our case to get a 50Hz output we remove both the 100K and the variable 100K both from pin 6 and we put instead a 260K fixed resistor and we leave the 0.1uF (the 104 cap) as it is, this change should give out a fixed 50Hz as per the formula :1.3 / (260,000 ohm x 0.0000001 farad) = 50Hz. But in reality it will not exactly give 50Hz because the 260K resistor has a specific error value margin so does the capacitor, that's why i recommend a variable resistor so that accurate calibration can be achieved.
If you want, you can substitute the variable resistor with a fixed resistor using the following formula: F = 1.3 / (RxC) in our case to get a 50Hz output we remove both the 100K and the variable 100K both from pin 6 and we put instead a 260K fixed resistor and we leave the 0.1uF (the 104 cap) as it is, this change should give out a fixed 50Hz as per the formula :1.3 / (260,000 ohm x 0.0000001 farad) = 50Hz. But in reality it will not exactly give 50Hz because the 260K resistor has a specific error value margin so does the capacitor, that's why i recommend a variable resistor so that accurate calibration can be achieved.
5> Use
either tantalum or polyester film for the 104 caps, ceramic disc caps are heat
sensitive, they change value when hot and this in turn changes the frequency of
the inverter so they are not recommended.
6> Pin 10 of the SG3524 can be used to auto shut down the
inverter, once a positive voltage is given instead of negative to pin10, the
SG3524 will stop oscillating. This is useful for persons wanting to add some
cosmetic makeup to their inverter like "overload cut-off", "low
battery cut-off" or "overheating cut-off".
7> Wiring connections on the power stage side should be thick enough to handle the huge amps drain from the batteries. I marked them with thick black lines on the schema also I included a image so you see how thick those wires must be. (You can make the driving circuit section on a breadboard for testing purposes but NOT the power stage).
7> Wiring connections on the power stage side should be thick enough to handle the huge amps drain from the batteries. I marked them with thick black lines on the schema also I included a image so you see how thick those wires must be. (You can make the driving circuit section on a breadboard for testing purposes but NOT the power stage).
8> The design does not include a battery charger since each
person will be building a custom version of the inverter with specific power
needs. If you are ordering a custom made transformer you can ask them to take
out for you an additional output wire on the primary side to give 14v (between
point 0 and this new wire) and use it to charge a 12v battery, of course this
needs a separate circuit to control charging auto cut-off. But anyway this is
not advisable because it will shorten the life of the transformer itself since
using it as a charger will toast the enamel coating layer of the copper wires
over time. Anyway .. YES can be done to reduce cost.
9> A cooling fan will be needed to reduce heat off the heat sinks and transformer, i recommend getting a 220v fan and connecting it to the output T2 transformer, when you power up the circuit the fan will start this will always give you a simple way to know that 220v is present and everything is OK.. You can use a computer's old power supply fan if you like.
9> A cooling fan will be needed to reduce heat off the heat sinks and transformer, i recommend getting a 220v fan and connecting it to the output T2 transformer, when you power up the circuit the fan will start this will always give you a simple way to know that 220v is present and everything is OK.. You can use a computer's old power supply fan if you like.
Note that the fan must suck air out from the inverter case
and NOT blow inside, so install it the correct way or it will be useless. Also
note how I fixed both the heat sinks and where the fan is, in a way that the
fan sucks hot air from like a channel between the 2 heat-sinks.
10> 2
circuit breakers are recommended instead of fuses, one on the DC side and one
on the AC side, depending on your design Ex: for a 24vDC ( 1500 watts design )
put a 60Amp breaker on the DC side and a 6Amp on the AC side.
For every 1amp of 220vAC you will be draining like 8 to 10 Amps from the 12v battery, make your calculations !
For every 1amp of 220vAC you will be draining like 8 to 10 Amps from the 12v battery, make your calculations !
11> The
2 Heat sinks should be big enough to cool the transistors, they are separate
and should NOT touch each other.
12>Important: If you're building a big design that uses more than 24VDC as
power source, make sure not to supply the driver circuit with more than 24v
maximum. (EX: If you have 4 batteries 4x12 = 48v , connect the v+ supply of the
driver circuit to the second battery's (+) terminal with a thin 1 mm wire which
is more than enough. (This supplies the driver circuit with +24v while supplies
the power transformer with +48v)
13> "Optional"
: Deep Cycle batteries are your best choice, consider them for best results
14> Be
cautious when building this circuit it involves high voltage which is lethal,
any part you touch when the circuit is ON could give you a nasty painful jolt,
specially the heat-sinks, never touch them when the circuit is on to see if the
transistors are hot !! I ate it several times :)
15> The
optional "Low voltage warning" is already embedded in the PCB layout,
you can disregard it and not install it's components if you do not need it. It
does not affect the functionality of the main circuit, it just sounds a buzzer.
16> The Motorola 2N6277 is a durable heavy duty power
transistor, it is used in many US tanks for it's reliability but unfortunately
it is a very hard to find part, instead you can substitute each 2N6277 with 2 x
2N3773 or any equivalent, and yes equivalents work too.
17> I've
included an optional "Battery level indicator" circuit diagram that
has 4 LEDs, you can see it installed on the front panel of my inverter pic, it
is functioning great and shows precisely how much juice the batteries still
have. I have included a small relay that is powered by the last LED to auto
shutoff the inverter once last LED is off.
18> Also
included an optional "Overload circuit", it is very easy to build and
can be calibrated to the desired overload current threshold cutoff point
through the potentiometer VR1. R1 is rated 5watts for inverters up to 1000
watts. For bigger versions of the inverter like 1000 to 3000 watts inverters,
replace R1 (1 ohm, 5watts) with (1 ohm, 17watts) which should handle loads up
to 10 VA. Make sure you install a proper relay to handle big current drains.
19> Please guys take your time to read and understand my
notes, browse and read the posts and questions asked by others because there
are many useful information listed in replies. The main reason for me not
answering your question is because it has already been asked before and
answered upon.
20> It would be nice and inspiring for others if you
take some photos and show us how you built your version, any additions to the
circuit are mostly welcomed to be listed here, we can all benefit from them.
21> Please click on the "I've made it"
button/icon if you did build the circuit so I know how many people benefit from
this design.
22> Testing the circuit on a breadboard with
crocodile clips or thin wires WILL NOT WORK ! You'll get wrong
voltage readings. Don't come back crying that you're getting a 150v output or
so.
No comments:
Post a Comment