# Inverter Circuit Help

Discussion in 'General Electronics Discussion' started by ikimpoy, Jan 14, 2014.

1. ### ikimpoy

7
0
Jan 14, 2014
Can someone help me with our problem in our inverter circuit. i got the circuit from this link.
http://www.instructables.com/id/250-to-5000-watts-PWM-DCAC-220V-Power-Inverter/
Because im making a 2000w inveter circuit for air-condition unit.

*I used 1 amp 24-0-24/220v on T1
* 10 amp 24-0-24/220v on T2
* i used parallel 13 IRPF240 Mosfets in power stage a total of 26 mosfets

Problem.
1.) Once i check to Vac from my inverter it only produces about 170+ Vac instead of 220
(*i already turn the trimmer as he suggested but nothing happens).
2.) once i try it in a electric drill about 650w it cannot turn it on.. it works in 50w electric fan though but still slow.. how can i achieve the required power i needed? is there something wrong with my setup?

pls help me.. i need to finish this for my thesis..

2. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,481
2,830
Jan 21, 2010
You might note that 170 * sqrt(2) = 240.

This means your circuit is operating correctly.

You are supplying 24V p-p on the 24 V winding and getting 240V p-p on the 240V winding.

Since you're producing a PWM signal corresponding to a sine wave (presumably) the RMS voltage of your output will be 170V.

The simplest solution is to get an 18V ct transformer.

If you want 650W, you'll need a transformer rated for about 18V 40A.

Unless you're using a 24V supply, I think you mean 24V CT (12-0-12), not 24-0-24

Your thesis? And you're getting designs from instructabes? I hope you're not doing electrical engineering.

3. ### ikimpoy

7
0
Jan 14, 2014

so you mean i need a 18V 40A ct transformer on T2 to power the 650W drill?
what if i need atleast 1500W?

im using 24-0-24 10 amps on T2. and supplying it with 24v.

nope im computer engineering student, im integrating android application on inverter circuit to control the frequency via phone. and im not that good at circuits so i really need help...

4. ### duke37

5,364
769
Jan 9, 2011
The transformer will be designed for a certain frequency. It will object, perhaps violently, to running on an incorrect frequency.

5. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,481
2,830
Jan 21, 2010
No, you need an 18V CT transformer to get the right voltage. In fact, you really need a little lower, 16V CT would be better, but is less likely to be available.

You need that because 18VAC actually swings from about -25.4V to +25.4V. In a special method to measure the average voltage (RMS) it becomes 18VAC RMS.

You need a larger transformer (higher current rating).

But you'd need to supply it with 34V to get the output voltage you desire.

Note that losses in the transformer mean you'll need an even higher voltage, but most appliances don't mind too much running from slightly lower voltage (as opposed to slightly higher voltages).

I would have asked you to hang your head in shame if you were an EE student.

This is high power electronics. It isn't going to be as simple as you may think. Expect failures in components and weird issues.

6. ### ikimpoy

7
0
Jan 14, 2014
I know i am really close on what should be the output of my inverter, please help me clarify things. i attach some photos of the circuit so that you can see what should i replace..

my aim is to have 220VAC with power rating of atleast 2000W.
you already said i should change the transformer on StepUp side,
but what value of transformer should i use to have 2000w?

Should i also change the first transformer?

one more thing.. is the 18v 40A ct transformer big? what i know is while increasing amps will also increase the size of the transformer.

i dont want to waste money so i want to make sure. because 18v 40A transformer is about 70 \$ is electronic shop is a little expensive..

T2 Transformer

Mosfet Side

PWM stage

T1 Transformer

Whole Setup

7. ### ikimpoy

7
0
Jan 14, 2014
but i already test it and it works.. my only problem now is the power rating.

8. ### BobK

7,682
1,688
Jan 5, 2010
To get 2000W at 18V you would need a transformer capable of 111A. 120A for a little margin.

Bob

9. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,481
2,830
Jan 21, 2010
Have you tested it under load (say with a 15W light globe?)

What voltage do you currently get?

With a 48V centre tapped transformer I expect you'll get around 35% of the rated primary voltage -- 84V. (what is the primary voltage?)

With 24V CT, you'll get about 71% (170V)

With 18V CT you'd get about 95% (227V)

This all assumes a 240V rated primary.

Note that this circuit can kill you. I'm not impressed by your layout. Be very careful and preferably do not work alone (it would be useful if the person with you knows how to isolate your circuit and perform CPR).

10. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,481
2,830
Jan 21, 2010
In many ways that's going to be a problem.

Your driving of the gates of these mosfets is such that they will switch more slowly than might be desired. Be aware that this may cause them to get quite hot.

Since heating is proportional to the current squared, this effect can be unnoticeable at low powers and suddenly (well, perhaps "more quickly than you might be expecting") becomes catastrophic as you raise the power.

In addition to that, you're wiring is not suitable for carrying huge currents, and you might find that you need to reduce the inductance of the wiring too (although the inductance of the transformer will probably swamp it out)

A significant portion of your losses may be current flowing through the body diode of the mosfets. Fortunately this will rise only a little faster than linearly with current.

11. ### ikimpoy

7
0
Jan 14, 2014
Yes under a 40w lamp. and it works quite fine.
I got about 170V then after the i plugged-in the lamp it drops to 140-150v.

we try to supply it with 24V and 48V still it has the same output which is around 170v.

Thanks for the concerns. i got my groupmates with me. and sorry for the layout, we are kinda not that good yet.

-----------------------------------------------------------------------------
Should we change the value of our T2 Transformer?

So far those mosfets dont get really hot because of our load is not that high.

should i use solid wires for the mosfet power stage? what size?

12. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,481
2,830
Jan 21, 2010
Have you examined the output waveform?

Is it sinusoidal? Does it change with load? What does it look like unloaded?

5,165
1,087
Dec 18, 2013
Coming into this a bit late guys but isn't this just loading of the reversed transformer on the output. If you go up in voltage you come down in current by your turns ratio. Ns/Np=Ip/Is

14. ### ikimpoy

7
0
Jan 14, 2014
Yes, i had a sine wave output..

i also check the output from the PWM it produces a square wave output.

I cant check it now, cuz i dont have a oscilloscope in my home.
but should the wave change if i add load? what should be my expected output?

by the way im modifying the PWM stage because i forgot to put the feedback to the PWM.. should this affect my output?

15. ### ikimpoy

7
0
Jan 14, 2014
You mean when i need higher voltage, i must have a lower amp transformer?
but i need a high current transformer to power high loads about 2000W?

and that 100A transformer is too big for the inverter model..
as i see on what the designer on instructable he uses not that big transformer.. as i noticed.

16. ### mahkohb

2
0
Jan 15, 2014
i got the same problem too.. i hope you help us.

17. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,481
2,830
Jan 21, 2010
The first thing you guys need to do is show me some waveforms from the secondary (240V) side with no load, a very light load, and finally with a heavier load.

My strong suspicion (based on nothing but mathematics and physics and long experience, and from looking at the schematic) is that the original designer didn't understand the difference between peak and RMS voltage and that the transformer is wrong.

But hey, I'm on the internet. I could be a dog for all you know.

Yes, it could be loading on the transformer, but I hope I'm not the only one to realise that the ration between the loaded output voltage and the desired output voltage seems to be the reciprocal of the square root of two.

edit: oh, and don't kill yourselves taking these readings on a scope. And post actual pictures. I'll also need to know more details like the type of probes used and the vertical sensitivity setting (usually measured in volts per division)

Last edited: Jan 20, 2014