ADVICE- What value is this capacitor ?

bilabonic

VIP Member
VIP Member
Joined
Apr 12, 2007
Messages
2,991
Reaction score
191
Agreed on C19, 1000 at 16v.

Board indicates C8 is a surface mount capacitor but can't see any markings.
 
Agreed on C19, 1000 at 16v.

Board indicates C8 is a surface mount capacitor but can't see any markings.

Yes it's a high temp 1000uF.

I have 3 faulty boards each with 4 of these capacitors.

I have not got a dedicated capacitor tester, but using a Multimeter i get -

Capacitors 1,2,3 and 4 in same location on each board -

1- 33.8ohm, 33.7ohm, 2.2Mohm and falling, 1.4MO and falling.

2- 55.5, 55.9, 2.4MO and falling, 1.9MO and falling.

3- 60.8, 61.1, 2.3MO and falling, 1.95MO and falling

What do you reckon guys ??
 
It's not the perfect test but falling values suggest capacitive action. I'd suggest the electrolytics were the most likely culprits.
 
Ceramic capacitors (C8, at the top) tend to go low resistance when they fail (although safety versions do exist which guarantee to fail open). It's difficult in-circuit but I'd expect a resistance measurement to go up as the capacitors charge but when charged there will be some leakage paths showing as a resistance. C19 is likely an electrolytic capacitor as aforementioned :).

At a guess these boards don't have any switching power supplies and can probably be powered up without any capacitors on (not permanently!).

I'd try removing them from one and briefly running it. If they are linear supply reservoirs you could try replacing them with something similar, if they are associated with switching power supplies their characteristics (ESR) might be a bit more critical.
 
Cheers guys, i know it's not the best test but it's the onl option i have atm, i have a capicitor tester coming and ESR, 2 were in seires and another 2 in series, with one in each series line showing low resistance.

The reason i went to these is that i compared them to a working board and hence the differing readings.

The GOOD news it works !!!!! You are correct it has no switching PSU and is powered via a pc or converted server psu via 9 PCI-E inputs.

Each card has 45 asics Finfets and a failing high powered server cooling fan failing (bearings) caused the fault.

I still need to get a decent way to test caps, with ESR and a decent chart to download to compare resistance values for value of the cap.

Cheers guys
 
Back
Top