Small DC Welder Arcs at Switch When Powered On

[Pancor]

I have a small DC inverter arc welder (https://www.decaweld.com/default.asp?cmd=getProd&cmdID=19306) that has an annoying problem I'd like to fix. One day after using the machine I noticed that the switch was stuck in the "on" position. Not thinking much of it, I replaced the switch. The first time I powered it on again, I saw a significant arc in the new power switch, which was now also welded in the "on" position. I have used the machine since out of necessity, just plugging/unplugging it from the wall to power on/off. When doing this, an arc occurs between plug and receptacle. While I have an incredibly basic knowledge of electronics and am comfortable replacing components, diagnosing this problem is well beyond my capabilities. I am hoping that someone with more skill and knowledge than myself can offer some assistance.

I have emailed the manufacturer for schematics, but they are unable to help me. From my understanding the machine uses dual IGBTs to output DC welding current. I have attached some photos and would like to say thanks in advance if anyone can offer assistance.

 

[tcmtech]

Many small inverter type welding power supplies have a simple inrush current limiter which is basically just a simple power resistor in series with the AC input side that once the main power capacitors are charged up gets bypassed by a relay in order to give the system full line power.

What you are describing sounds like the bypass relay has also gotten its contacts shorted together so when ever you flip the main power switch it too now gets hit with the full inrush current of the main power capacitor bank charging up which could easily be 10x what the machine normally draws at full power.

In one picture I see a black power relay that looks to be part of the primary power circuit. If it is I am suspecting that is where your problem is coming from. If you can pop the cover off and have a look to see if its contacts are welded together.

 

[Pancor]

First of all, thanks for your reply! The relay in my last picture is indeed on the input side of the welder. However, it doesn't appear to be the type you can pop the cover off of. The is the part on a seller's website : https://ca.mouser.com/ProductDetail/Fujitsu/FTR-K1AK012T/?qs=N2nzcguIp82IbFWJ3841qA==

Can I test the relay from the six leads on the bottom?

 

[tcmtech]

Yes you can. Just do a continuity test across the pins that would be carrying power if the relay was closed.

If working and the power is off they should read whatever the inrush current limiting resistors value is.

If they are zero ohms it's shorted together.

I used to work as a welder service tech and I saw this problem on occasion with the cheap low end inverter units.

Personally if that is the problem I would seriously consider doing a little bit of modification to refit a higher capacity relay in its place. Anything with the same coil voltage and somewhat close coil resistance will work. If it was me I would be stepping it up to a 30 amp rated unit.

 

[Pancor]

Awesome, I'm certain now that that relay is the culprit. Off to Sayal (**broken link removed**) tomorrow to see if they can help. I really appreciate it!

 

[OlPhart]

Howdy, some pixie dust for the fire... could MOVs or high voltage ceramic capacitors be put across the contacts to shunt the (what seem to be) back EMF spikes?

Perhaps a zero crossing solid state relay instead of a contact type? It would minimize the inrush current, but also require protection of the switch type.

Just ideas, doesn't mean they're good ones... <<<)))