SUCCESS! TIG DC --> Advanced SquareWave INVERTER is working (prototype)

http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/
Here you see my first working prototype of the IGBT based TIG
inverter. Again, this is an add on inverter for making Advanced Square Wave AC from DC supplied by the welding machine.
Some highlights (see schematic page for component info:
I used Semikron SKHI 23/12 boards for driving the gates
I used four Toshiba 200A IGBTs, each of them is a half bridge, so I have two paralleled full bridges with 400 A capacity.
A regulated current limited DC power supply was used for testing
No snubber circuit yet
On the oscilloscope you see:
top -- voltage across the AC rail. You see some nice ripple caused by the lack of snubber circuitry.
bottom -- voltage at the gate of one of the half bridges
My plans include building a robust snubber circuit (I already have all components for a RC snubber), as well as fabricating a sensible enclosure that would fit inside my welder.
As I intended, no deadtime is seen here (the bridge conducts all the time, there are no transient "dead" intervals but there are transient shorts).
I feel that this is a nice milestone in my project. I feel that I have, in principle, proven that I can get to the completion of the project. All I now need is not to fuck anything up. To that effect, I should make a nice snubber circuit and err on the side of overbuilding that snubber. I should manufactore an enclosure that provides proper mounting, vibration protection and cooling. After that, I should make sure that front panel controls are properly installed and are protected from EMI.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Mon, 21 Nov 2005 00:27:59 GMT, Ignoramus28489

oops, make it
http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/01-Prototype-1/
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus28489 wrote:

And someone said this guy was a troll? Well done Iggy - it looks great!
Chris
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Mon, 21 Nov 2005 00:33:24 +0000 (UTC), Christopher Tidy

Thanks... It is rare, I suppose, that people would assemble inverter circuits for the purposes of trolling...
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus28489 wrote:

I think that would be limited to academic trolls on $100k research budgets. Now there's a thought. Perhaps Cliff is a sociology postgrad conducting some kind of experiment :-). You never know...a guy got a grant to pursue art through shoplifting in England recently!
Chris
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus28489 wrote:

Is that a bullet lying on your bench near the center of this photo?
http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/01-Prototype-1/dscf0002.jpg
I'm trying to figure out how that is involved in the development process. :)
Dave www.davewilson.cc
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

No, it is a cartridge. :)

I am sorry, I would have cleaned up better.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 21 Nov 2005 06:01:50 -0800, the renowned snipped-for-privacy@evcom.net wrote:

Could be a safety hazard if it accidentally got some juice flowing through it. I'm impressively messy at times, but when stuff gets powered up, it usually is very, very neat (especially expensive and/or power stuff). Igor has the bulk of it relatively neatly placed on the wood board, but there is enough clutter to be worrisome.
Best regards, Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
snipped-for-privacy@interlog.com Info for manufacturers: http://www.trexon.com
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Mon, 21 Nov 2005 09:40:12 -0500, Spehro Pefhany

Well, nowhere I have over 0.5 A flowing right now.
Anyway, regarding layout and such. The wooden board is used temporarily so that stuff does not fall off the desk.
You may see 4" or so tall brass headers on top of the heatsink.
I am planning to mount a aluminum or steel flat piece on these headers, attach a fiberglass insulation board on top, and mount the driver circuit on that fiberglass board. The twisted wires leading to IGBTs will be going to the sides of the flat piece and then into the driver circuit. I will shorten then as well as I can.
All that I will enclose into some enclosure so that if the IGBTs blow up, at least the rest of the welder will stay intact. I will also make sure that this thing is cooled with two fans, one intake fan and one exhaust fan.
My design approach is to go through a series of successive prototypes, not to make a perfectly complete piece right away. I use the rsame technique in my computer programming. It really works because mistakes are caught early on.
Hence you see wooden boards, possibly alligator clips etc.
My mom gave me an earful last night. 20-25 years ago she worked at a Soviet research center that was figuring out stolen (by spies and such) US technology and how to use it in industry or defense.
So, she told me that it is a dangerous project for an amateur and that attention to safety sohould go beyond merely reasonable levels. I think that she is right. I would rather overbuild where enclosures and such are concerned.
Some of safety highlights will be:
1. A solid enclosure that would stop flying molten IGBTs 2. A snubber circuit that is overbuilt, featuring big caps, big resistors and big paralleled varistors. 3. Tiny snubbers on every IGBT 4. Normally closed temperature controlled switch to turn the circuit off if temperature of the heatsink goes above 70 degrees C (about 158 F)
My Semikron 23 driver boards also feature IGBT overload protection, they measure collector/emitter voltage and softly shut down the IGBT if they sense an overload. The normal C/E voltage is about 2.5 or so volts for 1,200 V IGBTs.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
    [ ... ]

>>
http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/01-Prototype-1/dscf0002.jpg
    [ ... ]

    Looking at the image zoomed in, it *appears* that the primer is indented, so it may be an expended one with a replacement bullet seated in it to serve as a setup gauge for reloading dies.
    Still not sure why it is there, however. :-)
    Enjoy,         DoN.
--
Email: < snipped-for-privacy@d-and-d.com> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 21 Nov 2005 06:01:50 -0800, snipped-for-privacy@evcom.net wrote:

7.62x54R actually. Its a known length..so it makes a good reference item, though its usually laid horizontally in the frame to make comparison measurements easier.
The round also looks like a 30-40 Krag, or 6.5 Dutch, because of foreshortening effect, but based on Iggies proclivities..Id have to say the Russian round is more likely, and given the red primer sealent..most likely East German..though it could be Slovak. I couldnt blow up the photo large enough to read the head stamp code and tell you which factory it came from
Gunner
"Pax Americana is a philosophy. Hardly an empire. Making sure other people play nice and dont kill each other (and us) off in job lots is hardly empire building, particularly when you give them self determination under "play nice" rules.
Think of it as having your older brother knock the shit out of you for torturing the cat." Gunner
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

correct, it is 7.62x54. These rounds (50 year old surplus) are kind of shitty, often do not fire. They are good for punching holes in railroad tie plates though.
i

--


Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@evcom.net wrote:

http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/01-Prototype-1/dscf0002.jpg
It's used for troubleshooting. Ed

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus28489 wrote:

Well, this is sure progress. But, you haven't run actual welding currents through this yet, I gather. When the currents in the IGBTs reach even 50% of rating, that's when all the effects of the wiring layout are going to start showing up. Since this is inherently a current-limited application with the TIG power source, you do have the benefit that momentary shorting of the welder should not cause massive shoot-through currents.
Good luck taking it to the next step!
Jon
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Yes. Thanks. There is progress.
It was conducting too much due to slow turnoff (due to my selection of 100 ohm turnoff resistors).
I experimented and essentially removed turnoff resistors (added alligator clips across them).
After that, I just measured its cross conduction due to transient shorts.
OCV Current
20V 0.035A 30V 0.047A 38V 0.050A
Would you think that it is sensible?

Yep. That's why I hope to have some cross conduction and transient shorts. Ideally (but not in reality) that would fully obviate the need for a snubber circuit.
Since in reality all kinds of things can go wrong, I will have the mother of all snubber circuits: a RC snubber between the main DC rails, with extra 150V varistors, plus a little cap across every IGBT witha 390V varistor. Just to cover my ass.

Thanks... And thanks for support and actual helpful suggestions.
I find electronic design to be somewhat akin to computer programming, I go through a succession of prototypes, there are subsystems, same logic is used in debuggig problems etc.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Mon, 21 Nov 2005 00:27:59 GMT, Ignoramus28489

Dont forget to make your enclosure explosion proof. Might I suggest a fly off top panel?
<G>
Gunner
"Pax Americana is a philosophy. Hardly an empire. Making sure other people play nice and dont kill each other (and us) off in job lots is hardly empire building, particularly when you give them self determination under "play nice" rules.
Think of it as having your older brother knock the shit out of you for torturing the cat." Gunner
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

snip--Somehow I came across an IR2183(4) half bridge driver. Did you mention that chip in one of your posts, or consider same? Jim
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I did indeed mention it. In fact I got two IR22141SS chips for free as samples from IRF. But I did not use them.
I ended up using Semikron 23 boards, two boards, one for every half bridge. They are quite straightforward to use and feel sturdier and with more features.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus1487 wrote...

One of my objections to the IR chips, if I have one, is their rather slow delay times, e.g. about 200ns for the IR2183, IR2184 and 440ns for the IR22141, etc. And their rather long deadtime, e.g. 400ns min for the programmable IR21834 driver. While such long delay times may not affect your IGBT application, I've found considerable benefit from having more flexibility and being able to tightly control the driver's deadtime, even down to the 50ns level. Furthermore, it's valuable to have tight control over the deadtime as a function of load current, as offered by some TI Unitrode chips.
I've forgotten, how do Semikron 23 boards stack up in this regard?
--
Thanks,
- Win
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus22022 wrote...

Don't forget diodes in series with gate resistors to separately control on and off delays. Like this popular scheme for rapid turnoff, to get "break before make."
. ---+----/\/\-----+--- gate . | | . '--|<|--/\/\--'
--
Thanks,
- Win
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.