Old Miller Welder problem

[MotorBikeLes]

But, is the Murex the same saturated core control system? If so, do you have a circuit diagram?
Keeping options open!
Les.

[threeseven]

Reading back I can see that I was a little ambiguous, apologies.
The Murex is a more modern fully electronic machine, but the foot control is just a pot controlling power electronics.
So, in principle, I think the same technique could be used on the TIG in question, with a conventional pot on the foot unit controlling the current through a thyristor or triac, that the original high power rheostat did. So instead of the rheostat dissipating high power it would be the thyristor or triac.

[Radio Wrangler]

It's essentially a DC current which is being controlled, which makes it difficult to use thyristors or triacs.

Power bipolars, mosfets or IGBTs could be used in principle, but the magnitude of AC component which can sometimes appear, and the sixze of transients will make it a serious design task and you can expect to go through several sets of power semiconductors before a robust design is proven.

The modern machines are inverter-based and all the transient handling etc is already done. Their current is commanded by a low current voltage from a five-turn pot, and the function is diverted to the pedal when one is in use. To go to this system would mean replacing the innards of the Miller with an inverter in place of the heavy iron.

To use the existing transductor current amplifier system with a solid state arrangement in place of the pedal rheostat means that the solid state system will dissipate just as much heat as the rheostat did. It certainly could be done but it would be a significant amount of design work. Apart from the durability of the pedal, that Miller is a very durable machine and should have decades of life ahead of it.

Modern TIG machines have a pulse feature where the welding current switches to and fro between two values like a metronome. You get into the rhythm and move/feed in time to it. With old machines lacking the timers etc, you tend to pulse the pedal and sync the movements of your hands with your foot. The pedal gets worked very hard doing this. So does your foot!

I also considered that if someone has such a serious machine and is paying cylinder rental etc, then it has a job of work to do, so making spares and doing yet another re-wind may get him back on the air soonest.

David

[MotorBikeLes]

Right David, I guess that settles it, new resistor winding. I certainly favour the idea of thicker wire or a stepped resistor chain. All I need is a suitable ceramic former for more, thicker wire. The calculations are no problem.
Like me, Martin is in his dotage, so the welder is no longer used for other than occasional jobs, so there is no income justification for regular new (unobtanium) pedals, plus he never has any money after feeding his many cats!
I think the original question was worth asking.
Cheers, Les.

[McMurdo]

thyristors are fo course often used to control dc, you simply replace 2 of the 4 diodes in a bridge rectifier with 2 thyristors and incorporate a flywheel diode across the output if you expect any back EMF.

Potters wheels here in Stoke have DC motors and are controlled by a foot-controlled variac into a bridge rectifier, originally made by Rilton Electronics though no doubt others exist.

[MotorBikeLes]

Kevin, it seems things have moved on since my time there.
They were usually continuously run off a three phase squirrel cage motor, with knee operated clutch to the drive belt. I think it was usually a push in knob on the studio wheels. However, do you connections have the ability to produce a former, say 3" in diameter and say 10" long?
I could have got something like that knocked up quickly once, and fired in any one of one hundred kilns.
I have thought about using a firebrick for a Rayburn or similar, and filing notches in it to hold the resistance wire.
Of course I know most of the pottery factories have long since closed, which make people like me unwanted dinosaurs.
Les.

[Radio Wrangler]

Sorry, I didn't express it very well.

The control winding is currently fed (oops pun!) from rough, bridge-rectfied AC. And while a thyristor-based rectifier could be used to control the average value of the output, setting the average by switching on part way through a cycle is not what's needed here, because the welder would go to full output during part of a cycle, and to low output during the rest. This is not the same as driving the control winding with an entire half sinusoid suitably scaled for every half cycle of the mains.

This form of welding is quite sensitive to wave-shape. My machine can switch between sinewave and squarewave, vary the frequency over 20-200Hz and offers calibrated distortion of the mark-space ratio when I have it switched to AC mode. These things adjust the shape/depth of the weld-pool and adjust the aggressiveness of it blasting away oxide coating when welding aluminium on AC mode. You could look on it as a 9kW function generator.

To not change the welding characteristics of the machine, the controller needs to be linear.

David