Help needed for Homebuilt MIG

[dandumit]

Hi,
I'm trying to build an MIG welder.
I have made a wire feeder: http://www.dioda.ro/coppermine/displayimagepopup/pid=26/fullsize=1.html

As transformer I'm using a heavy duty mna transformer wich is delivering 16.8V to 22V at output after rectifier.
The section of final winding of this transformer is about 9.5X4 mm square.

Like shielding gas I'm using pure argon.
The hole machine doesn't have electrolytic capacitors and doesn't have either current limiter-choke .

Unfortunately it doesn't weld at all. It makes a sound like a machine gun no matter how big speed I would use. The has wire has no penetration on material, all the wire is melted and makes "bubles" of melt material on top of material to be welded.

Please, give me some suggestions ! I have lost a lot if time with this project ! Could current limiter inductor improve so much the general performance ?

Daniel

 

[shenion]

You need an output choke to even out the current. Otherwise the wire will just spatter when it hits the metal.

Also make sure the gun is the positive terminal.

Pure argon will produce tall welds.

 

[dandumit]

Thanks for info - that's exactly what's happening ! The wire it is spattering and it doesn't touch the material continuously.

I have seen some schematics of cheap welders without this final inductor. I suppose that their transformer has an air gap to prevent saturation and limit the current.

I will add the current limiter inductor. Do you know how this should be made ? I mean What inductance has to have, no of turns etc.

Also the electrolytic capacitors may have such a drastic improvement ?

Regards,
Daniel

 

[shenion]

Yes, the cheap MIGs without the inductor are just cheap MIGs that produce crappy welds. I guess they relay on the low-gauge wire to limit the current a bit. Will only work well in a small range of settings. Poor design; actually no design.

Here is a reactor I made for my CP-250:

It is a piece of 1.5" iron pipe as a core with some spare doubled 3/4" box section around it. I knew it would probably saturate on higher currents, but was just a test to see how much inductance was needed. It has 30 turns of #6 wire wound around it. Worked pretty well, but needed a bit more. With a solid core and heavy duty box around it, 30 turns would have been about right.

I could calculate the inductance, would take several hours (much easier with solid core, this was pipe.)

I ended up using one of the inductors in the CP-250. It was 3-phase and I converted it to single (was also a spray-transfer only machine; hence no reactor inductor.) So one of the coils was unused. It has 28 turns around a bigger core.

If I was to make another one, I'd probably use 1/4"x2" solid bar for the frame and either a huge bolt to 1 1/2" to 2" solid bar for the core. But this is a 300+ amp machine, you could do with less for something smaller.

The inductance is not to limit the current. It is to even out the current and maintain the arc. When the wire hits, it is a dead short. Without the inductance, the current will spike and vaporize a section of wire. You end up with very short pulses of high currents. Average current is low.

With the inductance, the current builds up slowly (in the order of a few milliseconds). Then the wire melts back and a gap is produced. The inductor then collapses and maintains the current in the arc; without it, the current drops way down during the gap.

With the inductance added to the Miller, the current was increased dramatically. Also the weld was spatterless and had a soft buzz instead of the sound like grinding gears before.

I did add 60,000uf of caps to the rectifier. It increased the current below 50A considerably. The main reason I added it was to try to maintain the arc between the rectified pulses.

You should have some caps across the rectifier. Most have 0.5uf caps across each diode (look like old caps from distributer points.) That will complete the circuit during the time the inductor is maintaining the arc. It will also reduce the noise radiated out the input side considerably. That is something else left out on cheap MIGs, there posts about causing interference and lights flickering on cheap MIGs.

 

[dandumit]

Meanwhile I have found some similar threads :
http://www.mig-welding.co.uk/forum/showthread.php?t=1933
http://www.mig-welding.co.uk/forum/showthread.php?t=1999

but with no information about construction of limiter inductor.

 

[shenion]

By the way, for constant current welders (stick and TIG) there are two inductors, one is before the rectifier. That is usually a large variable inductance (or variable saturation coil) that sets the current. It will still have the output side inductor to maintain the arc.

That is why AC stick welders are cheap to make. They don't need the extra inductors. Real low end AC stick welders have nothing but the transformer with usually just a variable shunt across the core..

There is an "Arc stabilizer" coil described in this ancient article. Is about the same size as mine.

 

[dandumit]

@shenion - thank you for your detailed answers !
Last days I toked an transformer from an old tv set and I have wounded it with 2 CuEm of 3 mm diameter.
the results seems to be better - it isn't spatter so much but unfortunately it doesn't weld continuously at all.
I have tried various speed of wire but it is interrupting a lot - I mean that when the wire touches the piece it is melted on a long distance (few milimeters) and so on. All the wire it is melted on some small bubbles over the material to be welded.
Also a strange thing, the material to be welded it isn't penetrated at all. The melted wire it is easily toked away from ground where it was supposed to stitch.

Any suggestions are greatly appreciated !

 

[shenion]

I bet the transformer core is saturating. Something in a TV will be low power.

When I was working on the Miller, I suspected it needed a reactor coil (before I read that the newer model had one.) I took an old crankshaft and wrapped about 12 turns of the ground lead around it. The results were a bit better than you described. It still needed more inductance. The welds just sat on the surface but it did at least fuse to the parent metal.

A TV transformer would be too small. The picture of my coil shown in a previous post does not show scale. The coil is 8" square. The wire is probably close to 5mm. Then again, I was designing for 200A so the wire is bigger.

My small Clarke 130 had a core that was 5" square and over an inch thick. Looks like it had about 40 turns of wire on it.

BTW, what is 2 CuEM?

What output voltage are you running it? It may be a bit high.

I'd suggest finding some heavy steel. The center core on my test coil was 5" of 1.75" heavy iron pipe. I would have liked to have used solid bar stock for the frame.

Look at the "ancient article" link I posted above for their reactor. It looks like a good size.

The amount of inductance does not seem to be critical. There seems to be a wide range of a good value. The Lincoln I had had variable inductance, did not make a lot of difference on MIG after you reached a minimum value.

At some point adding more inductance will start limiting the current, but that would be considerably more.

I looked at your wire feeder. looks like part of an old microwave case The wire looks silver. Shoudl be copper coated; what type of wire is it? 0.8mm?