Is AC Voltage regualtion on inverter's DC rail an option?

Hello, I need to make a variac of a sort to controll aproximatly 2kW inductive loads voltage. Could it be possible to rectify, and use analog out to "dim" the dc rail using a very powerfull IGBT with one arduino that runs a PID regulator, and then invert the voltage back to AC using another arduino or a dedicated circuit(I already built the H-bridge using gate driving IC IR2136 in bootstrap mode used only single phase of it because I had it) ?

If so how would you go about placing the DC capacitors because the inrush at that power could be a problem for the "dimming" igbt, and both the converter, and the inverter should be fed a somewhat stable DC and not a choppy one?
This is the main problem I think.
Would the caps make the regulation slower(not by much I assume)?

If the inverter IGBTs can take 400V lowering the input DC should not be any problem for them?
If variable input would be the problem for the IR2136 vss somehow, may be even worse problem then the caps, how would one drive a variable input inverter then, gate transformers perhaps?

I dont need another solution like triacs or anything else at this point, so please stay on the topic. Buying aftermarket circuits is also not an option...

Its an idea so I dont have a schematic yet but I could draw it up...

Basically, the layout I have in mind is: full bridge rectifier - DC/DC converter ran by arduino(320-0V) - DC/AC inverter that outputs to say a welding transformer.

Why not make Arduino do it at the output inverter, because I don't know how to make one run LED with buttons, scan the probe, calculate, run a pid program, and vary duty cycle to vary output voltage all at the same time and fast.
It may be the only option for me, and should work? Low Efficiency is not a problem here, as the project is experimental.

I just need experienced opinions on this approach, information about this is scarce, but I red that this approach may even have advantages.

No part numbers, no links to existing equipment, no proposed schematics.

Semtex9:
information about this is scarce

Couldn't agree more! So far this reads like a proposal for a Darwin awards entry.

information about this is scarce,

Did you read all the information about the IR2136 and power conversion on the Infineon website?

Because you might have missed it.

Dangerous Equipment

Even brief accidental contact during operation might result in severe injury or death!

Only qualified personnel are allowed to handle this!

not by much I assume

I would not assume anything but double check everything with datasheets and expert advice.

Hello Semtex,
(I can't help associating you name with a possible negative outcome for this project :confused: )

What you want to do seems, to me at least, to be a very advanced electronics project. I am reasonably knowledgeable and experienced at electronics but I also know there are people here a lot better than me. If I were to try to do what you are trying to do I think it would require quite a bit of experimentation, probably result in some very black and charred components, parts of which might be embedded in the walls. I am concerned whether you have the knowledge and experience to do this safely, maybe you do but nothing in your post suggests that to me.

...to control approximately 2kW inductive loads voltage.

That alone worries me as for a reactive load I'd expect to see a specification in VA not Watts. You don't actually say what the load is and why you want to control it. Maybe there is a simpler option to achieve what you want. Maybe someone here can be more helpful, but my advice is to find a different project or find a different approach.

I guess if my project needed a variac, I would use a variac. If an Arduino needs to control the voltage, I would use a motor driven variac.

Paul

Jesus you guys are nit-picky, I have experience enough, its a proposal, and not a single answer given.
Thank you very much on your judgement, now I am fully aware of the situation and the risks involved, just as I was before I posted here, risks that are certainly not parts seared in walls (I see you don't know so much if you say that can happen, because the energy needed to put anything in armed concrete far exceeds my home's power supply or the projects capacity without protection engaging first). I am not a kid, this is a uni project. I will not kill myself or blow anything up despite my funny nick from early gaming days. So if you want to give some constructive comment, do it, if not then get off my thread, because I don't like people that can't be bothered to understand, talking about expert advice (that I AM ASKING for) and Darwin awards. It was a simple question, and I gave enough information for a normal person to answer. And I know what are reactive volt amperes, volt amperes, watts,and where they are used, you missed the point, and underestimated me, probably because of my style. In the datasheet the Vss is recommended 600V and thats all the information on high side voltage sense Klaus, It has internal level shifter for gates, but it dosen't say how low can it go on the high side, only how high. My question was can it go from 0-400. I don't have a variac, and using motors to turn variacs seems bronze ageish to me, and too slow for PID current regulation(I have a reason why I said current and not voltage), anyway thanks paul for you were the most reasonable here. Google anything about variable input inverters or gate driving, information is vague, and they didn't teach us at school. So all I have is to figure it all out myself. Instead of finding more excuses not to think, lets get back to the topic:

using a DC/DC converter to vary supply voltage to the variable input Inverter.
its like a very standard buck boost system, but in reverse to regulate rail voltage from 0-needed.
so I can feed it to inverter that will output to a 220V primary of the transformer, and then I pick up its output current on secondary with a Hall probe, so it can be regulated with Arduino that operates the said converter. Do I need to translate this from human that helps me think, into engineering language first?
Do you need a schematic or a block diagram(that I already did in a way), or parts manifesto, or for me to post on this forum for 2 years to get more status? or what, for you to understand this idea? All options of implementation are still open. I will draw it for you but I thought it is a simple concept to understand, and you can help me think.

specifically I asked you about DC rail capacitor placement, and the gate driving method for full H-bridge so it can take full variable input (I think gate transformers would be ok, but i would like to keep my IC if possible). If you know of any complications this system would have, and so on...

Has nobody ever done, or try this? ? ? ? Then either tell me why, or think for a minute, or don't comment, else you might embarrass yourselves, like me apparently. But it says project guidance, my project is a concept at this point and I need some guidance because its very complicated, even at this point. If I knew everything about electronics I would not ask. How exactly am I not reasonable here? Can't we have an open-minded, good intentions conversation?

It sure will require experimentation. I charred components in projects before, and will probably char more while doing it myself with or without you guys, it happens but you can maybe help me prevent some of it. I just thought this might interest someone that knows a lot on inverters and reactive power, that can actually help me, and wont just criticize everything just to appear smarter, that I can also accredit in the damn paper. I will do all the work, and code, no complete solution is required, just opinions on the things I actually asked, twice now. Thank you
I will post more once I get to the next stage alone it seems.

OK Sem, if I can call you that, calm down.

The comments made seem to be in your best interests and people are worried about your level of expertise and the possible danger to which you might be exposing yourself.

Real world concepts begin with a block diagram, followed by a schematic.

Paul

Semtex9:
too slow for PID current regulation anyway.

So you want something that can change the output voltage continuously and very fast as well (while PID control is actually most commonly used for systems where the output reacts slowly to a change in input). That's new information. Don't you think that may be important enough to include in your project description?

Quite sure I have seen such a circuit before on this forum - that was indeed about welding control, a year or two ago. Too long ago to remember much detail, can't find it now. I remember the circuit was based on a 220V AC to 12V (or 24V?) AC transformer; rectifier; high frequency inverter (this is the point where the Arduino came in, controlling the output current), second transformer for the actual welding current. So what you want to do can probably be done, but a definitive answer can't be given as you don't seem to be willing (or able) to clearly explain the concept you have in your head.

Anyway, your posts so far don't give me the feeling you have the experience to design and build this kind of project. Also, if it's a uni project as you say, why don't you start by discussing this with your professors instead of posting bits and pieces of what is definitely a pretty advanced project on a what is very much a beginner's forum? That are the people that should be able to help you with it, and not just the "don't kill yourself" bit.

Ok, I'm sorry for the misunderstanding, English is also not native to me. I have my reasons for not consulting professors on this. I could have spent more time, precisely explaining everything, then I might have gotten the same answers...

I first tried building a triac ac power regulator with a current feedback from the welder, worked fine on resistive load, but when feeding the welder it went wrong, nothing burned but we scrapped it without much thought and went for the inverter. Perhaps I could have solved it by placing a balanced inductor before the triac, or by changing the firing angle. Anyway whats done is done

Then I thought that I can controll the duty cycle of the inverter to achieve variable voltage generation, but it proved very hard with uno that needs to do manny more things.

So i thought that I could make a simple spwm variable input inverter and do the regulation on the dc side. DC is fairly easy to controll with Arduino, but this presented new problems. I have 7, 400V, 750uF capacitors that I wanted to use some of them to stabilize the dc rail. If I wanted to change rail voltage with an IGBT in series using arduinos pwm analog write, then I would get pulsating voltage to inverters input, I would need to place capacitors after the converter to avoid this, but the inrush at igbt on, would probably become problematic especially if I had caps after the rectifier too, that can provide huge current instantly. My specific question was around this, do I need to have stable input to the converter, will this be a problem. Also, since this is my first time designing gate drivers for a full-H bridge, I never thought about variable input for an inverter, that shouldn't be a big problem to achieve, but can my chip do it, or I swap it out...

you think I can avoid the PID regulator? I wanted to keep the welding current constant, by controlling the primary voltage. Suppose I could dismiss PID since the changes are very fast, but isn't it more precise/fast enough?

Paul_KD7HB:
Real world concepts begin with a block diagram, followed by a schematic.

Paul

Semtex9:
Its an idea so I don't have a schematic yet but I could draw it up...

Basically, the layout I have in mind is: full bridge rectifier - DC/DC converter ran by arduino(320-0V) - DC/AC inverter that outputs to say a welding transformer.

The problem mostly revolves around this part of the block diagram, only its not in rectangles because I thought that this would be enough. My bad...

This would be the layout, I can provide more information, but all I wanted to know is how to stabilize the DC, and what to use for driving the H-bridge if I want variable input inverter basically, if I leave it like this, the capacitors would pop the fuse, or possibly burn out the IGBT switch of the converter if I placed more capacitors before the converter, and make them charge over ptc, before engaging a work relay bypass like in this schematic to separate the momentary drain from the grid. Maybe I just need to use IGBTs in paralel to withstand higher current, but I would rather avoid it, maybe I don't need as much capacitance to achieve enough stabilization for the inverter, how would you go about it?
I wanted to use the same IGBTs for chopping DC and in the inverter HGTG30N60A4, because I have them readily available. The R1 in the schematic is used to drain the caps after disengaging so I don't kill myself, does the job in arround 10 minutes, could lower the resistance a bit later, 380V transil is used to protect from overvoltage(experimental) there is an input fuse, a network filter and some more protections before the rectifier that is highly over-specified, I had it lying around, but its not the issue here.

I can go and calculate everything but it would take a lot of time, that's why I'm rather overcompensating with components for now, But in the end this could potentially be used for a more powerful welder then I intended or in motor drive. Maybe this whole DC regulation approach is seriously flawed, that's why I am here for general advice

Untitled2.jpg