Inverter welder stuck on max current. Possible opamp issue?

[Kureigu]

Hi all, yes its a POS Chinese CT312 3-in-1 welder...
However, when it was working, I felt that it was good enough for the money.

Prelude:-
So I had been struggling to get decent results it in any of its modes for some time, realising after a while that the output was vastly higher than what ever I had it set to. I'm fairly sure one should not be able to blow through 3-4mm steel in 2 seconds on 10 Amps... After a while of "just getting on with it", the unit finally decided to go into protection mode and cease to do anything. Some DMM probing and cable disconnecting ensued and I couldn't point a finger at any one specific component or set of. So I did what any reasonable person would do, and disabled the protection circuitry. To no ones surprise turning it back on gave me some parts to replace. Two 2SK2837 MOSFET's and two 6R8 resistors later had a welder back to the same state as previous; working, albeit at full current. So that brings us to today, trying to sort the current regulation issue which is probably what led to the demise of the above.

The Issue:-
Adjusting the current dial on the front of the machine seems to have little to no effect on the output. The potentiometer itself is fine, however I did replace it anyway. Following the route of the control signal from the bottom left plug (red wire) up and round to pin 2 (Inv. Input) of the CA3140 opamp sees expected results; the voltage swings with the the dial and voltage drops off after each resistor, eventually leading to a tiny 57-60mV range at the IC. Pin 3 (Non-Inv. Input) is tide to ground. Pin 6 (Output) sees a voltage of 120mV regardless of pot' position. I pulled the CA3140 off and got 395mV at the pad for pin 6, so it does seem to be working of sorts... However this is where i'm completely out of my depth. I don't know if the 57mV is actual input voltage or a result of being pulled down by the feedback circuitry, and presumably the 120mV should be changing as a result different input voltages? Or is my testing methodology of probing across pin 6 and ground simply incorrect?

I hope this was at least slightly clearer than mud, any help would be greatly appreciated. Ofcourse I'm willing to dive in with the DMM wherever required, but unfortunately I do not have a scope to test any frequencies. Thanks for reading.

- Craig.

[abraxa]

I suggest you attempt to create a schematic diagram from the knowledge you have gained so far. Without it, it's diffiicult to figure out what's happening.

[ptrfonck]

i had the same issue
and the adjustment on the amps it is not made with volts
it is made with current that come from the outside potentiometer and the shunt resistor
the opamp measure both currents and adjust the duty cycle on the pin 9 of the ic 3525
on mine the duty cycle was always on 45% even if the amps where on 0 or 200 amps
only when i strike an ark i was able to see the duty cycle on the ic 3525 change

[Gregg]

The board looks like it has been quite hot at R43 and possibly C39.  I would suspect caps in that vicinity and circuitry connected to those traces.  Dave did a video about SMD ceramic caps cracking / failing due to board flexing and thermal stresses.

Also a number of the solder joints look sketchy in your pictures and might need reflowing,  I would also test all control pots for proper operation throughout their travel; welders are known to live in less than optimal environments for electronics.