Homemade High Amp Inverter Power Supply

[twenglish1]

I am interested in building an inverter type power supply that will be used to power a homemade plasma cutter, i always have been really into electronics and welding/metal working, so i figured id try to go about building my own equipment, i have build a welder in the past by rewinding the secondaries on microwave transformers, i want to try something new, i want to build a plasma cutter using an inverter powersupply, which is what i need help with.

Not sure if i want to use 120v mains or 240v mains yet, but the power supply will need to output about 200 volts (open circuit) at around 40 amps, will be later developed to have variable current output.

I understand the basics of how an inverter power supply works, i plan to use an AVR to provied the PWM signal to drive the IGBT module. i understand that that basically an inverter supply takes the mains, rectifies and filters them, feeds them through the IGBT module which is driven by the PWM of the microcontroller, and fed through a transformer, rectified and filtered again, i am just clueless on where to start with designing one to fit my needs, what components determine the output power? how do i design the transformer?this is where i need help

 

[Arouse1973]

An inverter takes DC voltage and converts it do AC. I don't think this is what you need is it?
Adam

 

[donkey]

an inverter uses low dc voltge to create higher AC voltage........ you saying you are using mains to output 200volts makes me think a transformer might be the best option....but having said that I really don't know what you want to do... unless you mean 2000 volts not 200
edit just saying 40 amps at that current... either 2000 or 200volts also makes me think I am not going to offer more advice to someone who appears to be a novice.... this is very very very dangerous and unless you know what you are doing, electrical shock may the nicest thing I can think that will happen to you.... seriously just don't play with this high voltage and really high current stuff unless you are fully trained

 

[BobK]

200V at 40 A is 8000 Watts. Normal house wiring cannot handle this. How do you expect to power it?

Bob

 

[donkey]

I was going to ask this but thought maybe there was a circuit I didn't know about (well there are billions of circuits I don't know about but thought there was one in relation to this)
I know in a standard house in AUS we can get maybe 15amps and that's for the stove..... very rare to see anything over that. so 15 times 240 is 3600watts... I know my dryer pulls 10 amps so assume that circuit is 10 amps ish.... but really never see much above that in houses.

 

[twenglish1]

I know more about this than you think I know, I know it will be possible I have a stick welder that outputs a max of 225 amps at 80v open circuit voltage, this machine runs off of a 240v 50 amp supply line, a plasma cutter works by ionizing a stream of air and it arcs through the air flow and through the metal being cut, I could go the transformer route which I wouldn't be asking questions about, I wanted to put my knowledge of programing microcontrollers to build an inverter power supply to supply the power to my plasma torch, and yes I know it is dangerous and will accept full responsibility for my own actions in building this, as I said earlier, around 200 volts dc at around 40 amps to start with, I have worked with and repaired welders and cutters and understand the risks involved with building a power supply like this

 

[BobK]

I don't think you mean inverter, perhaps you mean switch mode DC to DC converter?

Bob

 

[twenglish1]

I've heard them called inverters, and the companies that sell welders and cutters with these types of power supplies call them inverters, the supply will take the mains, rectify them, pulse them through the igbt module though a transformer and then rectify and filter again

 

[GonzoEngineer]

Finally, a post that could use my expertise!

Let's see, where do I begin..........

To start with, you estimate 8kW of output power. You are not going to get that with a single phase power source. You need 3 phases for that kind of power!

What you basically need is a Buck Regulator, which is a very simple design in concept, but there are a lot of pitfalls to be aware of. Pitfalls that can blow your IGBT to smithereens if you don't know what you are doing!

Granted, switching 40A with an IGBT is no big deal, and there are a plethora of PWM chips that will allow you to have both Voltage and Current feedback.

And basically, a Buck Regulator is just a switch, a diode, and an inductor. But you have to be very careful to reduce the inductances of the wiring, because the spikes that occur when you turn off the switch can be as much as 5 or even 10 times the voltage that you are switching.

You have to be very careful of the inductances of the wiring to eliminate those spikes, otherwise, you are going to fry your IGBT.

You will have to have a flyback diode across the IGBT, so that these spikes can be routed back to the input capacitor, and that capacitor has to be of very low inductance to absorb those spikes. You will also want that capacitance on a laminated bus to minimize the inductance.

Another consideration is your switching frequency, and how it affects the inductance.
You will want to use Litze wire between the switch, diode and inductor.

You want a PWM circuit that is controlled has feedback for both Voltage and Current.

This is not a project to be taken lightly, and is a lot more complicated that rewinding Microwave Transformers.

I am at work right now, but if you have any further questions, feel free to PM me, and I will get back to you this weekend.

By the way, I have a Buck Regulator design that uses 3 Phase 480VAC, and delivers more than 80 kW at 400VDC.

 

[twenglish1]

I might have to scrap this project as I have no access to 3 phase power, how do the commercially bought machines that use switchmode powersupplies run off of 240v mains? I even used a plasma cutter before that ran off a 120v 25 amp circuit, I believe that machine was 30amp output and was a switchmode machine

 

[(*steve*)]

I even used a plasma cutter before that ran off a 120v 25 amp circuit, I believe that machine was 30amp output and was a switchmode machine

I have no experience with plasma cutters nor the sort of equipment that GonzoEngineer is clearly intimately familiar with, however I do know how to multiply

If you want 200V at 40A then that is 8000 Watts. If you can draw 15A at 240V then that is 3600 watts.

It is possible that 200V is the open circuit voltage and 30A the short circuit current, and that at any point in the operating curve the product of the voltage and current is less than the maximum you can draw from your mains.

 

[twenglish1]

On the circuit for my welder I am capable of drawing 12,000 watts, its a 50 amp circuit at 240v, and that makes sense didn't think of that, this power supply will need to be constant current, which will allow the voltage to fluctuate, open circuit voltage will need to be at least 200 volts, now that I think, more would be better, to make lighting the arc easier

 

[GonzoEngineer]

Whoa!....lets back up a minute....... I thought you just needed 200VDC @ 40A.

Now things are starting to make sense, but I need some numbers before I can give you some good advice here. (And a Buck Regulator is not going
to do the job.)

1. What is the open circuit (striking voltage) needed to trigger your plasma in free air?
I mostly deal with triggering plasma in a Vacuum/Gas environment, and that usually requires at least 1000V!

2. What is the actual power required once the plasma is conducting.

I can see already that you are going to need a transformer, and switching it with IGBT's is the best way in my opinion.

You will need an H-Bridge to drive the primary, and a PWM circuit that delivers pulses in groups of two equal pulse widths.

If you switch the primary one direction, the next pulse in the opposite direction needs to be the same pulse width, otherwise the core is going to get mighty ****ed off.

It can be done, but usually requires a microcontroller, as I haven't found a PWM chip that does the job well. You are going to want to switch at the highest frequency possible, to keep the transformer and the inductors down to a reasonable size.

Let me know the actually striking voltage, and power required, and then I can suggest some components that will get you started.

 

[twenglish1]

The open circuit voltage should probably be around 200 or higher and for a start let's say 40 amps, I might get into programming variable current into it later on, and it is ignighted by a high voltage start circuit that is triggered over top of that 200+ volts, that pulses through the torch and ionized the compressed air stream which then allows the high current to flow as a jet of plasma, out the torch and through the metal being cut

EDIT: I should probably have more than 200vdc I was doing some more research and it said the typical open circuit voltage is 240 - 400vdc, with a 2 megahertz pulse of 5,000 - 10,000 volts to ignite the arc

This website explains better than I can how they work: http://www.axisplasma.com/what_is_plasma_cutting.html

 

[twenglish1]

I came up with a basic schematic after doing some research, am I getting the right idea? I left out the h bridge drive circuit, and probably other required resistors and inductors

 

[twenglish1]

Can anybody please help me!?

 

[(*steve*)]

Can anybody please help me!?

Possibly only Gonzo Engineer, but he's not always around.

Seriously high power stuff like this needs lots of knowledge and experience. If anyone can help you, it's him.

 

[twenglish1]

I hope he comes around soon, I've tried posting on other forums and doing extensive searching and I cant find any information anywhere, just the same basic information I already know about it, and more people telling me I can't do it