What inductance for mig.

[Dr pepper]

Ok as some will know I'm working on a onboard mig/tig weld set for my landy.
The power source is a big alty at 200a, I was thinking what sort of inductance will I need for doing car body stuff, 0.8 and 1.5mm.
I have done a lash up test with a rewound microwave transformer wound with 16mm2 copper cable, the most I can get on is 8 turns, it measures 330uH, I thought I saw somewhere that 1mH is a good value for mig welding.
The other issue with the microwave tranny is saturation, and thats hard to measure, but with the volts and no of turns is probably ok.
I have a multiweld 180 and could pull it apart and meter the output inductor, but its a git to get in bits.
Just looking at shenions earlier posts about the set he made, the core was 3/4" and there was 30 turns of 6awg wire on it, to my calculations taking the permeability of mild steel at 2000 then it'd be very approximately 100uH, so maybe my micro tranny will be ok.
Ta.

 

[shenion]

300uh is good. Mine was about 150uh and I wanted more. My Miller has about 200uh right now, it is adjustable from 60 to 280uh. Does about the same above 150.

Would like to measure a big MIG though. I bet it is more than 300. Maybe another microwave core will do.

The reactor I got from surplus center has about 10uh. Has about 20 turns of wire, must of uses some crappy steel for the core. It works for MIG but has a bit of spatter and current is low.

I'm finding the lower the current/voltage, the more inductance is needed.

 

[Dr pepper]

Thanks for that, so I'm in the ball park then, excellent.
I just did some more calcs.
Presuming you used mild steel for the core and that it was thick walled or solid.
The magnetic flux density at 24v would be 124,100 Oersted or 12.4 Teslas, the plain mild steel h curve goes into saturation at about 2 Teslas, so your inductor will be well saturated.
A microwave tranny with 8 turns has a much lower magnetic density, about 3.2 Teslas at the same 24v which will begin to saturate soft iron laminates, then I spose at 24v/200a you'd be well into spray transfer where it doesnt matter, however I'll stick with the micro tranny, that lash up worked good, without any calculations.
I think maybe the way forwards then is to use 2 trannys, my original 8 turn by 16mm2 wire version for higher currents, and wind another larger inductance choke using thinner wire (and therefore be able to get more turns on to get the inductance) for use at lower currents, and bring the output of each to a different socket on the front panel.

 

[Robotstar5]

I think maybe the way forwards then is to use 2 trannys, my original 8 turn by 16mm2 wire version for higher currents, and wind another larger inductance choke using thinner wire (and therefore be able to get more turns on to get the inductance) for use at lower currents, and bring the output of each to a different socket on the front panel.

Thats just what the "big boys" do, professional weldsets normally have 2 sometimes 3 work return sockets for different inductances

 

[Dr pepper]

There was cloos and esab weld sets where I used to work and they had 2 return sockets, never had a esab apart as they never broke, but the cloos ones from memory only had one choke with 2 taps.

 

[Robotstar5]

Found a pic of a Kemppi with 3 work returns



and the Saturn beside it has 2

 

[Dr pepper]

Thats a big set for a 230, whats the power switch 230 and 415?

 

[Robotstar5]

Its a 3 phase only machine with seperate wire feeder, the switch selects 11-19 & 20-33V ranges which you can then adjust with the fine & coarse switches below, you do the fine tuning and wire speed on the feeder above.

 

[Dr pepper]

yep makes sense, I was thinking there is not that many settings on the course and fine switch, but of course thats doubled by the main power selector.
I've never used a kemppi, are they any good.

 

[Robotstar5]

I've never used a kemppi, are they any good.

I've never "driven" one manually, but have integrated quite a few with robots when they were "set up and forget" - pretty reliable.
The Kemppi in the pic was one of about 15 assorted welders of various branded makes being used in production which we replaced with a "fleet" of Murex sets to standardise and make spares / repairs easier (for me!)

 

[Dr pepper]

We had a dozen or so cloos robots where I used to work, and a couple of panansonic doo dahs sort of artificial robots, the weld sets were cloos and panasonic, there was a load of manual bays too which had butters sets, but the later ones were cea an I wasnt all that impressed with them maintenance wise, the butters machines were good to work on and good to use.
I now work in a papermill and theres no weld sets at all, 'cept for the one here thats mine, its a esab compact 300 amp, and was part of my leaving present at my last job, its legal, they said take it put it in the skip or the back of your car, it went in my car.

 

[Dr pepper]

Back to the original thread, and probably a statement for shenion.
My micro tranny idea seems to be succesfull, and a couple of trannys and sockets would work, but I was thinking if I wind another coil on the micro tranny choke of thinner wire then I'd be able to vary its inductance using the saturation effect, and have a pot for inductance.
The good thing about this is that the tranny isnt bothered about the current through it saturation wise, just the volts/freq and cross sectional area effect the saturation, and with a stabilised voltage suppy like the regulated alty should work well, I'd need either a coil with a lot of turns, or construct an oscilltor using a 555 simlar to my hf start circuit that uses a car ignition coil and drive the winding at a higher frequency to force the xformer near saturation.

 

[shenion]

The lincoln R3S-325 i used to have had variable inductance. Basically a large secondary connected to a variable voltage clamp. It simulated the core saturating.

One of the Millers is the same, just has a rheostat across the secondary.

My Airco uses variable saturation coils to set the current. A few heavy turns then a large secondary. These are on the DC side of the rectifier. They add a DC bias current in on direction to increase the range so the weld current starts near the negative saturation then can go to the fill positive saturation. Current citrol is done by variang the bias from negative to positive current.

 

[Dr pepper]

The voltage clamp is interesting, must have had big igbt's or fets for that.
There was a few butters where I used to work, they had a seperate choke on the output, and these were infinately adkustable with variable dc provided via I think a rheostat, or it could have been a variac if it was a variac then it must've had its own rectifryer.
I dont really know if my idea will work, but the bh curve for soft iron laminates suggests that it probably will, the curve approaches 4th order at the top edge so the controls gonna be sensitive at the saturation point.

 

[shenion]

The voltage clamp was a simple transistor and voltage divider.

 

[Dr pepper]

Interesting.
I just found this on the web, very interesting.

http://aaawelder.com/reactor.html

I was thinking after your earlier comments that I'd need 2 mot's to regulate both the pos and neg cycles, and connect the primary dc saturation coils in anti parallel to cancel the transformer effect from driving ac back into the dc control circuit, however this guys had a brill idea about winding the coil on the outside of both sides of the mot in opposing directions cancelling the transformer effect from bothering the dc drive side and doing the whole thing in one tranny.
I have a couple of dead switchmode supplies which rectify mains and smooth it into dc, if I build a simple power regulator maybe using the switching supply's own chopper transistor and appling pwm to the dc primary saturation coil at fairly high freq to make the inductor 'see' more or less a dc current (the primary saturation coil I'm refering too can just be the mot's original mains primary).

Circuit bashing and thinking continues, I like this kind of thing.