So i have an old DC welder which just stopped working altogether. Symptoms, besides no arc, is that it turns on, but if you dial potentiometer to lower than 70 amps, welder starts to turn itself off and on (without trying to weld), and the lower you go with the potentiometer, the faster it does like that. If i dial more than 70 amps, it works(doesn't turn off), but there is no output.

So i've taken this thing apart (here is the pic of the board https://imgur.com/a/ArolJ ), and first thing i measured is diodes - and all of the secondary rectifier diodes are shorted - so i wanted to ask, is it possible that all the diodes have blown? Or i'm somehow measuring them wrong?
The primary rectifier when measured is good.

If it helps, there is a manual with the schematics and working principles:
http://valvolodin.narod.ru/schems1/kemppi_minarc_150-vrd_151_ver-11_sm.pdf

 

Yes, the diodes may all go together. When you tested them, did you disconnect the transformer?
The below from your pdf.

 

I actually didn't, partly because it's soldered and for a couple of days i wont have a soldering machine, and partly because i've thought that they want to disconnect transformer just because you can damage it's connector when handling the board. Or are they somehow shorted if the transformer still connected when i measure them?

And that is what's interesting - they say that if diodes are shorted, the welder shouldn't even start. But somehow it starts(though it turns itself on and off, depending on the potentiometer settings), so i wondering maybe i'm measuring them wrong..

 

Well, either way, if they read shorted with the transformer connected then the next step is to disconnect it. If they are still shorted then you've got the fault, if not then you need to look elsewhere.

 

Okay thanks, the next step will be to disconnect the transformer then. I'll update when i test them.

By the way, what could be the reasons of blown diodes? Can they all just go bad at the same time? Or something else can fail and damage them?

 

Or one fails and takes the others with it.

 

Update: turns out that diodes are connected in parallel, so it looked like all of them were shorted, when in fact only one was. Replaced it, and now welder welds.