Strange SCR power supply (welder) issue

From your other posts, it sounds like you have eliminated anything that may be a defect or cause the smoke to get out in the near future.

I went to PCTI's web site to see if they had manuals or schematics posted but did not see any.

Guesses: There may be a small offset in the current sense circuit.

There may be a minimum width conduction period imposed by something in the board design.

For my nickel, 12 amps doesn't sound like a bad minimum current to me, I think living with it is a good choice.

Good Luck, Bob aka wanderingmetalhead.

I have some documents, but not very detailed.

Well, not according to my measurements. I measured output from the current sensor(with my Harbor Freight multimeter), it is a pretty accurate reflection of the actual current.

Yeah.

Yes. What I will soon do is, I will switch control of the SCR firing board to the microcontroller, and thus I could compensate for this to match actual current with desired current.

I had a problem of output current not being exactly proportional to potentiometer setting, and I think that it is related. More specifically, if I adjust my board so that at potentiometer setting of X, I get 50 A, then 2X does not give me 100A, but instead gives me

I am now convinced that the output function current I, as a function of potentiometer output X, is

I = kX + 12A

where k is a constant, and not

I = kX, as I would rather prefer.

But I can live with it.

right

I do not think that it is the case. It would be great if it was, and your suggestion makes a lot of sense, but the output from the current sensor is very proportional to current, no offset. I measured i with my harbor freight multimeter.

I think that maybe I could add a tiny bit of voltage to the current sensing input to the SCR firing controller?

Keep in mind that the SCR's are controlling the reactor and not the main transformer.

Yes, they are controlling the main rectifier, not the transformer.

That's fine instrument there. Don't question it.

It's not bad for the $3 I paid for it, but a 9 VDC battery reads 18.9 VAC one way and zero the other. Well, hey, the average is pretty close! No precision rectifier, no capacitor, just a single diode in series. But maybe you have the deluxe version?

Paul