8KW Welder+Induction heater / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

8KW Welder+Induction heater

1 2 next

Move Thread

LAN_403

Linas

Thu May 01 2008, 12:39PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Hi
i am building powerful welder, and i get lot of Wima FKP1 caps, 1600V 2,224uF total, so i decided make and induction heater,
about welder: 88A 500V mosfet's (double FDH44N50) driving by PWM (TL494 push-pull output)
driver is very interesting, use 3 power supply's for each mosfet driving , (2 power supply's for high side, and one for low side) full-bridge topology.
how to using this circuit make resonant mode ? ( think use RT leg, witch was feedback transformer on work coil)

and how to make stable amperes, on output ? because i just using 4 leg to reg amperes on output

and i don't have feedback

Marko

Thu May 01 2008, 01:07PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I'm not too clear about what you exactly want from your post, but I can tell you this:

- for an arc welder, you absolutely need current mode control. If you just drive a forward converter transformer by fixed oscillator and bridge of mosfets, it is going to instantly explode as you short the output.

Acquiring some huge ferrite and powdered iron is other problem, I'd better have those first and base my design around them.

I hope you can expand your description somewhat before someone can tell you more,

Marko

Linas

Thu May 01 2008, 01:30PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Marko wrote ...

I'm not too clear about what you exactly want from your post, but I can tell you this:

- for an arc welder, you absolutely need current mode control. If you just drive a forward converter transformer by fixed oscillator and bridge of mosfets, it is going to instantly explode as you short the output.

Acquiring some huge ferrite and powdered iron is other problem, I'd better have those first and base my design around them.

I hope you can expand your description somewhat before someone can tell you more,

Marko

does someone know, how to make all in one small box ?what topology use and etc

Marko

Thu May 01 2008, 04:55PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Here are some pics of my 3.6kVA solid state arc welder.

I didn't want to disassemble it more than this but main parts can be seen.
One can conclude from the transformer and buck inductor that it is a forward converter (and I don't really know what other topology could be used here.)

For first time in my life I could see an IGBT module and the famous ISOTOP package - not something I'll see any time soon.

The transformer and the buck inductor. Both look rather small for what I would expect for this power level. Buck inductor only has 2 turns on it. I assume it's some sort of molybdenum-permalloy or similar powder core. Still it looks really small, just several of those PC-power supply material 26 toroids could probably replace it.

In the output stage the double rectifier diode is seen, and the current shunt. Note the poor looking shunt design with voltage sense screwed right together with high current terminals.

In the input stage a relay, apparently for inrush current limiting, is used.
3x 470uF 450V caps in parallel are used.
pink toroids of unknown material are used on ground conductors.

Another shot in attempt of showing more electronics - a small current transformer is visible. I'm not sure whether this CT, or output shunt, or both, are responsible for current control.

There is lots of electronics hidden behind the heatsink I couldn't reach, most in SMT.
There are 4 small relays there, I'm not sure what they do.

From the behind of the board I can see that the power module is one of those multi-modules with bridge rectifier, IGBT H-bridge, and probably gate drivers included all in one brick with lots of solder pins on the top. That one would be hard to replace if it blows.

Linas

Thu May 01 2008, 05:02PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

here is my old welder, madded chaged SSTC primary to transformer

Mates

Fri May 02 2008, 06:45AM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

Hi Linas,
Iâ€™ve been thinking about feedback for tl494 for quite some time. There should be possibility to substitute the internal oscillator of the tl494 for external one. The problem is you need a saw pulses. So Iâ€™ve been playing with an idea to use a PLL feedback followed by integrator (maybe simply a cap and resistor could be enough?) connected to the TL494 and drive it. Unfortunately my electronic skills are very limited, but would be nice if somebody could manage to build or at least simulate such circuitâ€¦

Steve Conner

Fri May 02 2008, 09:50AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Very interesting, Marko

I wouldn't be surprised if the filter choke core was just iron powder of some sort. After all, it's got forced air cooling, and isn't continuous rated.

Are those double diodes in parallel, or is one used as a freewheeling diode? Or is it the arrangement with a centre tapped secondary and push-pull drive, in which case it freewheels through both diodes in parallel? I guess it must be, since you mentioned an H-bridge.

It probably uses peak current control through the CT, and average current control from the shunt.

Linas

Fri May 02 2008, 09:55AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

well, TL494 is easy synchronize to my resonant frequency, but as you say, need saw signal,
and how to make better amperes control, by using CT ?????

lpfthings

Fri May 02 2008, 10:12AM

Registered Member #1361 Joined: Thu Feb 28 2008, 10:57AM
Location: Cairns, Australia
Posts: 305

lol Marco, my welder is VERY similar, if not identical to that one. I never knew it was solid state though...

Marko

Fri May 02 2008, 11:59AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Dr. Spoke wrote ...

Very interesting, Marko

You can see that there are 3 total conductors coming out of transformer, two are connected to diodes as a fullwave rectifier. In this configuration each diode also serves as freewheeling diode, so it's economic to do for high power levels.
As far as I know half-forward topology is used for very low power only.

I'm not completely sure it's a fullbridge, I just assumed from number of pins, power level and fact that all electrolytic caps are paralleled. Some sort of DC blocking caps are visible on the board, but barely.
It's very hard to see anything behind that transformer and caps, there are more traces on the top of the board, and I didn't want to disassemble the thing any further just to find that out.

Yeah, it may be a halfbridge connected without capacitive divider of electrolytic caps.

It probably uses peak current control through the CT, and average current control from the shunt.

Hm, so you mean, I have one error amplifier for peak current control, and other one that goes into it's negative input is average current feedback from shunt, just like it would be voltage feedback in SMPS?

Marko

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob