old transformer tig welders and variable frequency inverter?

[coppercone2]

I have a tig inverter already, but I want to know this, if anyone knows, so I can get a feeling for the behavior of transformer power electronics.

If you hooked up a old transformer TIG welder (assuming you made sure the control system is functional) to something like a AC power source  (say like a chroma), could you adjust the frequency of operation?

Or say you plugged it into a airplane (400Hz).

Can I use a transformer designed for 60 Hz at 400 Hz? Yes, a transformer that is designed for 60Hz operation will work at 400Hz, but may have a loss in efficiency due to eddy currents and skin-effect increase in resistance.

Has this been done before? It sounds like you would need to de-rate, but I wonder what is possible here. I wonder if you could go even alot higher then that if you had the source (say KHz). If you start with a high power one could you get some unique performance @ low power levels? A chroma can do like 2KHz. I think that may be useful for sheet metal.

[MadTux]

Depends on the machine, simple transformer box likely would work @ 400Hz and you could probably somewhat weld with it.

More complex machine that is designed/uses 50Hz for regulation (SCR pre-regulation, active SCR rectifier....) probably doesn't work
And AC/3phase motors for water pump, cooling fan will go crazy on 400Hz.

Running old transformer box on 50Hz with rectifier and Mosfet DC/AC converter afterwards likely works better.
Mosfet DC/AC converters on TIG isn't complex stuff, square wave signal, 2 isolated power supplies for low/high side driver, bipolar driver stage and some optoisolators to contol them, all thats needed. Then a big inductor to isolate it from HF igniton pulses, HF coupling transoformer, TIG torch.

Primary inverter usually more complex and more difficult to repair and more fireworks, if something goes bad.

Pulses for sheet metal usually is DC pulses, high frequency AC might be nice for aluminium.
DC pulse stuff is usually done on primary inverter side, by modulating inverter duty cycle.
Even old Fronius MW450 from 1992 can DC pulse @ 2kHz, seems nothing too new.

[coppercone2]

do you have any literature or information about AC at the high frequency of multi KHz?

Everything I search comes up for ultrasonic plastic welding.

It kind of seems like it would turn into a plasma welder type thing at that point, since they say the higher frequency focuses it down. But just making a new weld head is easier then trying to make higher electrical frequency.

And yes I mean to imply that the auxiliary systems would be wired to the proper frequency, such as motor devices. I meant if it was carefully modified just for the power electronics stuff.

[MadTux]

No idea, how it is, to weld with kHz AC TIG.
With DC pulses @ a few 100Hz to kHz, you get noticeable better focus and it somewhat like stirs the molten metal, so filler sticks much better to base metal. With no pulse, you sometimes have problem, that the arc/filler would not melt/stick to one side of the metal in a _| joint, with kHz DC pulse, molten metal connects to both sides much more evenly.

But as I said, it should not be too difficult to build own Mosfet DC/AC stage and simply try it.
Perhaps without smoothing/HF backstrike blocking inductor afterwards, so touch start only.
Likely also the reason, why no multi kHz AC, because it would be blocked by the inductor needed for HF start, at a certain frequency, so touch start only, then.

[coppercone2]

im sure you can get around that some how, just put a seperate power supply no? or just put a big piezo crystal in the foot pedal. If its a separate supply thats HV you can put a big resistor to protect it from the relatively low voltage of the welding arc maybe, and just raise the starter voltage to compensate?

I kinda just want to know how much it would focus it down.

I have 4 spare IGBT (two of two types but their all high power, 300-400A), I wonder  if you can use the welder as a DC source and then use the IGBT externally to generate the wave. But the problem is I would need to make gate driver circuits and stuff. I think the ones I have said something like 60KHz in the data sheet.

I see they have gate driver boards for sale on ebay though. That looks attainable. I don't want to make a custom PCB though for a experiment. I made a ISO5500 board but it lacked the power IIRC, you needed a buffer...

You think messing with the down stream of a miller dynasty with a igbt bridge could screw it up?