hi guyz,

i have built a small UPS about 800va (500w). the UPS works perfect. the inverter uses a large transformer to step up AC. i just wanted to know is there a way to make an inverter without this transformer? i mean the transformer is bulgy and heavy and it is wasting about 300watt of my power. how can i get this 300watts back using some technique or combination of MOS or IGBTs? i cant find a circuit which is designed for this purpose but i want to know the technique which may be helpful in making this transformerless.

any help will be appreciated.
Thanks to All.

 

No transformer will eat 300 watts while hadling 500 watts. Your calculation is wrong i suppose.

Anyways please check this appnote

http://www.st.com/internet/com/TECH...AL_LITERATURE/APPLICATION_NOTE/CD00253868.pdf

 

Transformers are one of the few man made devices that approach 99% efficiencies. If your transformer was absorbing that much wattage it would be hot, as in melting plastic hot. Is this the case?

I suspect you are dealing with inaccuracies with RMS measurements more than anything. Without schematics there isn't much we can suggest.

Transformerless inverters do exist. A UPS has an inverter as part of the design. They do this by using a large bank of batteries to create a large DC voltage. Somehow I don't think that is what you want.

 

I think the inverter that eats over half the input current is one of the lousy ones that were recently posted on this website including one posted on Bill Marsden's blog.

It has "protection diodes" that short circuit half the swing of the center-tapped transformer winding.

Modern inverters use a small inexpensive high frequency transformer with a ferrite core. They use sine-wave-modulated PWM. The output of the inverter is a pure sine-wave.

 

No transformer will eat 300 watts while hadling 500 watts. Your calculation is wrong i suppose.

Anyways please check this appnote

http://www.st.com/internet/com/TECH...AL_LITERATURE/APPLICATION_NOTE/CD00253868.pdf

bro i am attching the pics of the inverter i have designed. according to the designer 800va transformer is used for 500 watts. with the power factor of 0.625 if my calculations. so th

ink i am loosing 300 watts.

 

Well, you are not loosing 300W. The difference between the 800VA transformer and the proclaimed 500W output is a safety margin. This way the author makes sure that your transformer doesn´t melt when fully loaded.
You see that this inverter has no regulation or protection what so ever, which is kind of kludgy when he uses a microcontroller that could do much more than just generate some kind of a waveform.

Unless it is a square wave inverter, the 470ohm gate resistors are way too high and the single pair of transistors will melt when loaded.

 

I think the inverter that eats over half the input current is one of the lousy ones that were recently posted on this website including one posted on Bill Marsden's blog.

It has "protection diodes" that short circuit half the swing of the center-tapped transformer winding.

Modern inverters use a small inexpensive high frequency transformer with a ferrite core. They use sine-wave-modulated PWM. The output of the inverter is a pure sine-wave.

even with the perfect transformer with no losses it will be costly and bulgy. my main goal is to reduce usage of copper and inductors. this will make the size of the inverter very small as compared to inverters using transformer (Heat sink is another matter).

i can work on PWM but the major problem i am seeing in this project is how to step up voltage from 24vDC to 220vAC. i read somewhere that for this project i will have to step up the 24vDC to 500vDC and then turn it to 250vAC. any suggestions?

 

You didn't design it; you just built a project that someone else designed.
Their page is here: http://www.m0ukd.com/electronics/modified_sine_wave_inverter/

The transformer is specified with a higher VA than the expected load, because there is no output voltage regulation. Expect the output voltage to decrease significantly at higher load levels.

I think you have the wrong concept of what power factor is. A purely resistive load will have a power factor of 1. It's the ratio of true power to apparent power.
Link: http://www.allaboutcircuits.com/vol_2/chpt_11/3.html

 

Have you made the measurements to demonstrate this 300 W loss? As Bill mentions, that much power being lost in a transformer should make something smoke. The measurements are simple, but do require care as the waveforms is going to need an RMS measurement. If you used a simple DMM without RMS capability, you'll measure the wrong values (this is the voice of experience... ). It's best if you can use a scope, as it's then harder to be fooled.

 

<snip>
Unless it is a square wave inverter, the 470ohm gate resistors are way too high and the single pair of transistors will melt when loaded.

It's a square wave inverter. The PIC used has a low current source/sink capability; +/-20mA. The 470 Ohm resistors might be sufficient protection to keep the uC from getting fried if a MOSFET failed shorted from drain to gate (a common failure mode).

Sinking 36A current at a 25% duty cycle, the MOSFETs would dissipate about 11.8 Watts each. An improved gate driver circuit would help somewhat, but very little due to the low frequencies involved.

 

Well, you are not loosing 300W. The difference between the 800VA transformer and the proclaimed 500W output is a safety margin. This way the author makes sure that your transformer doesn´t melt when fully loaded.
You see that this inverter has no regulation or protection what so ever, which is kind of kludgy when he uses a microcontroller that could do much more than just generate some kind of a waveform.

Unless it is a square wave inverter, the 470ohm gate resistors are way too high and the single pair of transistors will melt when loaded.

bro i want to avail these 300 watts whatever be the causes. because i want to terminate transformer so i can make an inverter purely ELECTRONIC device. not electronic+electrical.
well the author says that the main losses will be within the transformer itself so he didnt say he is using margin of anykind.

here is the link

 

You didn't design it; you just built a project that someone else designed.
Their page is here: http://www.m0ukd.com/electronics/modified_sine_wave_inverter/

The transformer is specified with a higher VA than the expected load, because there is no output voltage regulation. Expect the output voltage to decrease significantly at higher load levels.

I think you have the wrong concept of what power factor is. A purely resistive load will have a power factor of 1. It's the ratio of true power to apparent power.
Link: http://www.allaboutcircuits.com/vol_2/chpt_11/3.html

yeah i meant to say that exactly that i have build the inverter designer is someone else who shared the circuit. i know bro that my concepts are a bit freakish since i havent played with transformers that much . thats is why i am visiting my teachers to make my concepts right.

but as i said early i want to remove the transformer.because i have heard that without transformer i will have about 95-97% efficiency and this will reduce the cost of transformer as well as reducing the size of transformer and less heat generation.

 

It's been said 100 times, building a decent UPS will cost you more than buying one. The one you link to is barely better than a simple gate based inverter and is no good for use as a practical device.

And you need a transformer.

 

bro i have implemented this ckt for myself and it is working for 50watts ( as i had v small transformer for that).it will cost me about 25-35$ including transformer worth 18$ of 800va.... but you know i just want to do something new and build up my concepts.

 

I just had an idea of how it could be possible to make one withouta bulky transformer, using two alternating smps. Tomorrow I will think about it and maybe start drawing some schematic.

 

SMPS (switching mode power supplies) still have big coils, especially for the wattage the OP wants. This will be close enough to having a transformer to have very little difference.

I get the feeling people who want to eliminate transformers really don't understand them and their advantages. They are used for some excellent reasons, the same reasons you don't see designs without them. They perform an indispensable function.

Still, I wouldn't mind seeing the design from kubeek.

 

I have almost 100 wall-wart AC-DC power supplies. Most are 500mA but a few are 1A and are pretty big and heavy.
I recently bought some surplus name-brand switching power supplies that are 5V at 2A and they are tiny and weigh almost nothing. They use a tiny ferrite core high frequency transformer.

 

There is a difference between 10 watt power supplies (5V @ 2A) and a 100 W unit.

 

Most comercial SMPS here that are 12V to 240V ac use a SMPS using a HF transformer creating about 375V DC. This is then choped up with H bridge mosfets to give sort of squarewave 240V AC. This way the transformer is relatively small.

 

Bill,
Among the stuff I sent you was a rather large Magnetics Inc. 584T 7743-A7 Kool-Mu toroid; it's black, and 47mm/1.85" in diameter (OD). You also have several "flavors" of switching regulators that were in the same package, along with some rather efficient MOSFETs - and switching regulator ICs.

There were a number of other toroids in the package, but that one might be a good candidate for experimentation on a project like this, if you have the interest.