DC-DC SMPS M2-ATX bad caps, drawing large current, turning off - Repair

[Lope]

I've got an ATX DC-DC PSU M2ATX 160W. It worked well when I first bought it. I've not used it in a long time, and I've never used it for any substantial period of time. Mostly testing ideas. I don't recall it being faulty.

I wanted to use it now and I noticed 2 caps were crusty & burst. Both 10v 1500uF 105C gold. (see pic) Neither of the 2 caps I removed were a dead short. Both charged up as I held my ohm meter on them. All of the 3.3v and 5v caps (labelled purple & orange) are the same specs.

I've replaced both caps with (what I had available) 3300uF 10v low ESR.

I've got all jumpers removed, which means it's always on and functions as an ATX PSU. (basically it's always powered up to make 5vSB and whatever else and when u short green to ground it supplies the power on all the wires. (in theory)

When I apply 13.8v I get a really massive spark, and it's not just to charge caps, because if I keep tapping the input wire on my bench PSU I keep getting large sparks. (I've been tinkering for 20 years, I know the diff between big sparks and small ones, its pulling more than it should) If I reduce mt bench PSU to 9.5v the sparks substantially smaller but the result is the same.

ALL TESTS WITHOUT LOAD ON OUTPUT BOARD VISUALLY LOOKS 100% PERFECT, ZERO SMELL, RUNS SILENTLY.

GREEN WIRE OPEN: 5vSB comes up quickly (labelled in purple) and stay at 5v as long as I've got power applied, if I remove power it drops really fast.

GREEN WIRE GROUNDED 5vSB as above. Yellow 12V comes up quickly and stays at 12v as long as power is applied. If I remove power it drops fast. 3.3v comes up for 1-2 seconds if I tap the input power wire. But if I hold it, it drops quickly to 1~2v ish. Red 5v output does not come up. Stays at 0v.

I've not measured the current draw. But the sparks are pretty much the same size with green open or shorted to ground.

ANY GUESSES? My guess is there must be some heavy load or short on the 5v line. The problem with that theory is 5vSB sits perfectly at 5v :/ I'm guessing 5v gets switched to 5vSB with a mosfet when green is shorted to ground. But even with green open, the heavy load or short is still present.

Any advice will be appreciated.

 

[Lope]

Wow I feel like such a fool.

I discovered what I thought was a lot of current, and some kind of short was really just inrush current to the capacitors.

How I discovered this:
I've tried back-powering the PSU. The 3.3v and 5v lines both only draw a small current to charge the caps and the voltage on the caps drops slowly.

When I backpowered the 12v line, there was a large spark each time I applied power to it. At this point I thought "Ah, the problem must be on the 12v line!

But then I decided to finally measure how much current was being pulled when I backpowered each rail.

To my surprise the 12v rail drew a steady 20mA, the 5v rail drew 20mA, and the 3.3v rail drew 10mA (this is with a relatively cheap DMM that measures in increments of 10mA)

So then I decided to measure the power draw on the input of the PSU... 80mA. And while powered on (with a low input of 7.5v) it pulled about 130mA

Wow I feel like such an idiot. I should have taken measurements immediately.

I guess I'm not used to such low ESR caps, with a 20mA load on them. Once discharged they **** up current like a vacuum cleaner.