Power Inverter help


2007-09-22 10:23 am
I need a Help in developing class D sinewave inverter to convert 12VDC into 220VAC

pardon me, im still a beginner in class D amplifiers...

i really need to learn more on how to build the output stages of a class D amplifier, actually i am using the TL494 PWM chip.

my aim is to convert the 12V car battery into a 220VAC.

the TL494 should code the sinewave digitally through PWM and the output is step-up using a transformer.

please help me,,,this is what ive done:

first, i only use one error amplifier of the TL494 chip, inject a +3Vp 60 Hz sinwave into pin2 (inverting), which i thought will serve as a varying voltage reference of the PWM regulator, and feed the output of the TL494 back to pin1 (non-inverting).

i only use the TL494, i mean it is the only IC i used in the circuit, when i check the votage at the drain of the MOSFET (output), the pulses are wierd, and they arent look like pulses, as if its operating in the active region.

honestly, i really am not experienced in class D amplifiers, and i really want to learn

should the sinwave that i input into the TL494 be dual polarity?
would it work with only a single MOSFET?
should the TL494's output transistors work in push-pull configuration?
should the power MOSFETS be configured in complimentary pair?

i still got many questions, but i want to answer most of them through my experience,,,please guide me,,,tanks a lot!


2007-06-11 6:06 am
not sure about the d-amp , but if u looking for a inverter then this would be one of may ways to do it , pwm runs at several khz error amp gets 50/60 hz sine wave from combined sine/sqr wave gen , so volt out of secondary modulated with 50/60 hz, looks like a sine wave with many steps , diodes make it dc , small cap smooth it a bit , then h-brigde drive from combine sqr wave gen make it 50/60 hz ac,

i m working on a similar project with good success (not sine but sqr wave ac),
it’s a basic out line of the topology i m currently working on,

sqr wave is relatively simple to implement , and works great for general purpose applications ,


  • topo.pdf
    43.4 KB · Views: 403


2007-09-22 10:23 am
hi there!!!

im really glad someone replied for me,,,yep a square wave inverter would be very simple, but problems occur when the loads connected are inductive, thats the primary reason why i chose to design sinewave through pulse-width modulation, but thanks a lot really. currently i am using the tl494 PWM chip, my problem is how to design the modulating inputs, and how to make the power amplifier outputs using mosfet drivers and the lowpass filter to recover back the sinewave...

thanks a lot!,,,im looking forward to get to know more about you guys, and meet many electronic enthusiasts!!! more power to this forums


2007-06-11 6:06 am

here is a pdf full of theories could be a bit helpful, square wave is not that bad with small inductive load like florescent light with big choke coil or fan, only problem is they become a bit more noisy and draws a bit more current then before, don’t know about power stuff like motor , no plan of powering them up with a battery backup inverter,


2007-09-22 10:23 am
thanks a lot,,,

is there any way wherein i can minimize the heatsink? i mean how could i make it cooler,,,thats one of the reasons why im trying to get into switchmode circuits in order to minimize heat problems,,, ^^,

i really appreciate your help!,,,someday i could just repay you =)

there is also one big problem im facing, im really an electronics enthusiasts but sad to the fact that in our country, the available parts are limited huhuhu,,,actually i am constructing a portable solar power system, and im trying to design a compact 12VDC-to-220VAC PWM sinewave inverter, to make it more efficient and not bulky,,,


2007-09-22 10:23 am
helo areza

as for now,,,i have only been successfull in using the TL494,,,but with only a single power mosfet output,,,im still trying to understand how to brdge the mosfets,,,i cant even make the push-pull configuration work,,,

i tried experimenting with the SG3524 but i wonder why it wont work,

im also trying to let you see my schematic,,,im really busy designing the whole system...

here is my email address: [email protected]

thanks again! merry christmas pal!!! happy new year too! =)


2007-06-11 6:06 am
Is it a flyback converter , don’t know about fb converter , seems they need more data about core and need more math, then pull-push, also they need biger core then pull-push for the same output power, may be for 12v to 200v high power application pull-push has some advantage then fb converter,
Re: heyyyy,,,,,=(

lljohnll said:
i cant access this site eprints.iisc.ernet.in/archive/00006921/01/Analysis_Of_A_Class_Of_PWM.pdf"

Hi lljohnll,

You can't have the " character at the end of the URL.

Right-click on the link below and then select "Save Target As" from the pop-up menu, and save the PDF document to your hard drive. You will need Adobe Reader (free from http://www.adobe.com ), to display the PDF file. Here is a corrected link to the file:


(My 56K connection is behaving strangely, tonight [or maybe that link is very slow]. So I haven't actually gotten that PDF to download, completely, yet. But I was at least able to get to the main page, at http://eprints.iisc.ernet.in/ .)


There are lots of people here who could help you, IF they had enough information from you. You need to give as many details as you can. I think that we would really need to at least see your schematic, to be able to help you very much.

Also, do you have the manufacturers' datasheets for the devices you are using (i.e. the TL494 and the MOSFET)? And have you searched for and downloaded any related Application Notes, from the manufacturers' websites?

What particular MOSFET are you using? And what does the Vgs look like, that your circuit is applying between the gate and source of the MOSFET?

The way to minimize the heatsink, for a MOSFET, is usually to make sure that it turns ALL the way on, very quickly, and then all the way off, very quickly. Then, if it has a very low Rds(on), even with a lot of current flowing through it, the power, i.e. (i^2)R, will be low [where i^2 = current squared]. It will only dissipate a lot of power (and thus get hot) if it spends too much time in the active region, where its R is much higher than Rds(on). So, if your MOSFET is getting hot, and its Rds(on) spec is low, then maybe you are not hitting the gate hard enough and/or fast enough, with your Vgs voltage.
1. To Convert 12V into 220VAC is is huge step up. Ideally you should consider a higher input voltage such as 36 (~42Volt) using three 12V batteries in series. The smaller the difference between the input voltage and the output voltages improves efficiency and reduces thermal dissipation.

2. You need to consider a full bridge topology instead of a Flyback (One Transistor) or even Push-Pull. This has been discussed in several early threads of this forum.

3. You need to use a PWM to generate a pseudo since wave that modulates the duty cycle. I could tell from your original post if you were using this method, or just using the PWM Controler to output a 60 hz sine wave to your switching transistors. Below is an example of a PWM generated Sine wave:


4. You need t make sure that your transistors are fully saturated (and quickly too) to minimize switching losses. If your switching transistors aren't fully turned on (saturated), they will create excessive amounts of resistance, which then is converted into heat. In most cases its appropriate to use MOSFET drivers to turn on and off transistors instead of driving them directly from the PWM controller. You'll likely also need MOSFET Drivers when driving Hi-Side Transistors to get the Gate Voltage above the Drain voltage (necessary to turn them on).

5. remember that if your target is 220VAC that about 620V Peak-to-Peak. You'll likely need to set up your transformer to output at ~310V, and then flip polarity every half cycle to simulate 620V P2P.

6. You have to be very careful about transformer saturation. If you go over the edge you likely experience a bang and a small cloud of smoke orignating from your switching transistors.

there is also one big problem im facing, im really an electronics enthusiasts but sad to the fact that in our country, the available parts are limited huhuhu,,,actually i am constructing a portable solar power system, and im trying to design a compact 12VDC-to-220VAC PWM sinewave inverter, to make it more efficient and not bulky

What your asking is a pretty tall order. What your looking for is going to take considerable engineering time and lots of money. To be honest your much better off buying an Solar/Off-Grid Inverter rather than attempt to build your own, hoping to save a few bucks. In reality, its going to take many many times the cost of an off-the-shelf product than to develop your own.

It sounds like your looking for this:

xantrex has a wide range of portable power products that probably will suit your needs. I don't think you will be able to design and build a reliable inverter cheaper than an off-the-self solution.


2007-03-19 8:41 pm
xantrex has a wide range of portable power products that probably will suit your needs. I don't think you will be able to design and build a reliable inverter cheaper than an off-the-self solution.
12v to 120v inverters are very cheap, but 12v to 240v inverters are expensive and far less common. (In fact, since 240v devices are almost always large loads, 240v output inverters almost always accept an input voltage higher than 12v. A 3kW, 240v unit I installed for a server room had an input of 48v and even then, the battery cables were quite thick and heavy.) For a common application like 12v to 120v, it would be cheaper to just buy an inverter. For something less common, custom design or modifying a commercial product might be the way to go.
I have actually modified a common 12v to 120v inverter to output lightly filtered 170v squarewave on one outlet and I'm planning a mod to output 170v DC on the other outlet. That is done to maximize efficiency of digital power supplies and allow compatibility for power supplies that wait for zero crossings before starting up. The DC output greatly reduces EMI.
Low power 12V DC to 230V AC inverters are widely available in Europe, where all low power loads like TV sets, DVD players, video game consoles and the like work with 230V.

A 230V AC inverter involves stepping up 12V DC to 400V DC with a conventional (regulated or unregulated) push-pull transformer stage, with full wave secondary rectification and 600V diodes. Several transformers with independent primary MOSFETs and series secondaries are recommended for perfect current sharing and low voltage stress on transformers at higher power.

Then those 400V should be chopped with a full bridge made with 500V or 600V MOSFET and/or IGBT. This 400V chopper may be of "modified sine wave" style (simple direct switching) or it may be of class D style with LC filters for true sinewave output. Obviously the latter case requires some kind of modulator while a NE556 dual timer seems to be enough for producing a poor man's "modified sine". There are low-cost PIC devices like PIC16F684 and PIC16F690 that suit very well this modulator application because they include specific PWM and full-bridge control hardware. Otherwise, an analog modulator like the ones found in audio class D amplifiers (either triangle-wave based or self oscillating) should be employed. In any case, recoverable current limiting (with no shutdown) must be implemented to get proper reliability.