Welding LVI supported ZVS inverter sim testing.

[kees54]

Hi There


I am busy with testing a sim of a design for welding, it does full bridge zvs softswitching
with a current driven coil who therefore vary induction, I did read this on a paper and did
starting it, a normal resonant version with a TL494 do also nice but only when resonant
coil and C is in resonance and Q is oke for wide zvs action.

Oke I do share some thoughts, it is also a learning test for making a smps for my audio hybrid
amp who need multiply voltages.

These is for a Mig, a constant voltage source, I need also a ltspice model for the uc3895 this one
is more easy to cope with. The uc3875 do sim strange with the slope resistor giving no fase shift
or just a little, except when make resistor much bigger and ramp cap also, the uc3879 what I
have now do really nice.

voltage feedback do work and also the current protection but need more work, it need to be fast
regulating with welding duty's.

Softswitch?



voltage feedback action.



softstart



current protection/feedback.



sim schematic



regards

kees

 

[kees54]

Hi There


I have play some with different way of feedback, however with welding what is a short circuit were
these supply are nof for originally I did some interesting things with comparators, it is I think
less sensitive for trouble, need of a slow loop is working with a nice hot arc reaction, who is
not really needed with a MIG welder.

Maybe some people here has also ideas, I do after some sims have a lot of pleasure in these
supplies, interesting also for my hybride amps.

Feedback to slow to much current difference.



overcurrent protection, (HICKUP), yes it is short
because I did make softstart cap smal for sim..



Making the loop faster, nice constant voltage but
yes it is sim, to fast loop make proberly instable.



Schematic of the inverter.




thanks in advance.

kees

 

[kees54]

I have sim a current/voltage feedback, maybe a way to combine MMA/MIG with a switch or such.

View attachment 5154

see different between RC time of feedback low pass, making it to fast it give trouble of unstability
in the loop.



RC time not oke.



But Maybe I have to build and see, the parts are on his way simming
is not 100% affcourse and the inverter need simple, not horn and bells.

regards

 

[blasphemy000]

I came across a nice application note for doing Phase-Shift ZVT with the UC3875. It seems pretty informative about the details and some of the troubles with setting one of these converters. The application note is a small(550W) converter, but I'm sure the ideas can be scaled up.

 

[kees54]

I came across a nice application note for doing Phase-Shift ZVT with the UC3875. It seems pretty informative about the details and some of the troubles with setting one of these converters. The application note is a small(550W) converter, but I'm sure the ideas can be scaled up.

Hi

Thanks for the information.

I have myself did find this only that was a short article, this is very good indeed.

Problem with a voltage source welder is the fastness of the feedback, making the inverter instable,
only a pic of a atmega controller can provide a fast regulation, I have also find some stuff in Russia
where is quasi resonant used with a tl494 of a uc3825 with frequenty as the regulator for current set.
However there is a special ic for this the vco driven UC 3861/67 becaus I need feedback, these ic,s
are fery fast, a new uc fase shift controlle has also variable dead time to prevent blow up the bridge.

There are so much different ways, I have seen so much zvs articles, so I thank you for these, it explanes
some more about the fase shifting. I do not now if it is suitable for a welder, because of short cicruit use
and spikes.

regards

 

[kees54]

Hi There


Keep a house warm with that 20 kw smps? I do not think so, because then it has a lot of loss.

Oke I have done some simulations for my MIG inverter ZVS version with lvi coil.

It does sim now after use a standard opamp model and put this on ltsspice directive.

.option cshunt=1e-15 gmin=1e-9 abstol=1e-9 reltol=0.001 chgtol=1e-9 and setting absolute current tolerance to 1e-005.


Now at last I can play with it.

the feedback circuit do work with it after some changes.

low amperage setting
View attachment 5272

High amperage setting.

 

[kees54]

Hi There


Keep a house warm with that 20 kw smps? I do not think so, because then it has a lot of loss.

Oke I have done some simulations for my MIG inverter ZVS version with lvi coil.

It does sim now after use a standard opamp model and put this on ltsspice directive.

.option cshunt=1e-15 gmin=1e-9 abstol=1e-9 reltol=0.001 chgtol=1e-9 and setting absolute current tolerance to 1e-005.


Now at last I can play with it.

the feedback circuit do work with it after some changes.

low amperage setting
View attachment 5272

High amperage setting.



ZVS/ZVC Soft switching LVI coil (I have now only one coil as resonant tank for sim because maybe I find others, better.)



schematic



regards

kees