Can a single bad capacitor cause restarts?

R

RoGuE1230

Limp Gawd
Joined
Jun 25, 2006
Messages
464
I noticed a computer next to the trash a couple weeks ago and brought it home. It is this model:

http://www.newegg.com/Product/Product.aspx?Item=N82E16883227035

It restarted at random times, usually seconds after turning it on. I put it to the side and got back to it today, and now it doesn't seem to do that, but last a lot longer. I put in a Vista install disk, it will load up all up on that progress bar and then immediately restart. I tried loading Linux Mint, and it will give all kinds of errors. I don't think it is the RAM. It had 2x2GB Corsair XMS sticks. I tried them separately in all different slots.

So I took a closer look and I see one single bulged capacitor. How likely is that the one capacitor is the culprit? I'm considering replacing it.

Here is a photo of the only bulged cap, circled in yellow:

http://i.imgur.com/ZUmVgNd.jpg
 
Last edited:
I'd put money on the culprit being the PSU. They don't list a manufacturer for it on the product page so it's probably not a quality part to begin with. I wouldn't be surprised if the main capacitor in the PSU is bulging itself.

Also, from the image it appears the cap isn't in a "critical" place such as the power switching section so I doubt it's alone in causing the issue.

Definitely check out the PSU though.
 
I swapped out the PSU with a known good one. I noticed when I removed the PSU it is incredibly light. It almost feels like an empty box. Out of curiosity I popped it open and the two largest caps looks, fine. There is one small that is bulging. I also just removed the processor heatsink to get a good look at all of those capacitors and they look fine.

However...the problems still persists. As I'm typing this, it was running a memory test from a Linux Mint USB stick. It made one good pass on the memory, and after 6 minutes it just restarted itself. Right afterwards it restarted itself 4 times before even getting to the POST diagnostic screen.
 
Problems like this can be hard to track down but often times I see PSUs getting blamed for things more often than they should be. Although your problem sounds like something the PSU could be at fault for. On the memory I'd try adjusting the timings and verifying that the timings in BIOS are set to what the module specs call for. All too often MSI, GIGABYTE and ASUS (as well as others) will tune memory more aggressively to win benchmarks. Ensure that the voltage readings in the HW monitor in the BIOS are where they should be while your at it.
 
Until you replaced the obviously bad capacitor and test the other capacitors, you will be taking shots in the dark.....
And be sure you are hooking up the ATX12V connector. (4pin in corner of mb)


Kid
 
A bad capacitor could cause issues but I doubt it given where that one is.
 
Kid744 said:
Until you replaced the obviously bad capacitor and test the other capacitors, you will be taking shots in the dark.....
And be sure you are hooking up the ATX12V connector. (4pin in corner of mb)


Kid
Click to expand...

Yep, 4 pin was plugged in.

Dan_D said:
Problems like this can be hard to track down but often times I see PSUs getting blamed for things more often than they should be. Although your problem sounds like something the PSU could be at fault for. On the memory I'd try adjusting the timings and verifying that the timings in BIOS are set to what the module specs call for. All too often MSI, GIGABYTE and ASUS (as well as others) will tune memory more aggressively to win benchmarks. Ensure that the voltage readings in the HW monitor in the BIOS are where they should be while your at it.
Click to expand...

Good idea, I'll check that tomorrow.
 
I wasn't suggesting that the bulged capacitor was the only bad component. Between voltage regulators and caps, there is little else you can check with a meter except continuity of traces and voltages. So check the caps and go from there.

Kid
 
Replaced hundreds of bad motherboards in the last 5 years at my shop and a single bad cap can definitely cause that. That being said there are some machines that run with a whole row of bulging caps. Usually the cap failed for a reason though, and you do not know what other small components could be bad.

You could always try desoldering and putting in a new cap, just remember that you have to get things hot enough so that the solder runs down the hole to make contact inside the PCB and not just on the surface.
 
Gman1979 said:
I'd put money on the culprit being the PSU. They don't list a manufacturer for it on the product page so it's probably not a quality part to begin with. I wouldn't be surprised if the main capacitor in the PSU is bulging itself.

Also, from the image it appears the cap isn't in a "critical" place such as the power switching section so I doubt it's alone in causing the issue.
Click to expand...

Please stop giving advice on this sort of problem if this is your level of knowledge of electronics. "I'd put money on the culprit being some random thing I just said rather than the actual problem your picture shows." Riiight. Okay. So why don't you put your money on it and buy him a new PSU, then? I think you owe OP some money. I can see wanting to help and answering the question and saying "It could also be the power supply," but you took your ignorance and doubled down on it with the over-confident BS.

Yes, a single bad capacitor ANYWHERE in the system - on ANY component - can cause significant instability. Also, capacitors are not always visibly bad. Most of the time, if one capacitor is bad, other capacitors will also be bad even if not visibly such.
 
Last edited:
QuiteSufficient said:
Replaced hundreds of bad motherboards in the last 5 years at my shop and a single bad cap can definitely cause that. That being said there are some machines that run with a whole row of bulging caps. .
Click to expand...

Since the thread hasn't been active for almost 1 mth. I have a similar question:

it just so happen that I have a motherboard w/ 1 single capacitor that is slightly raise, virtually the same as the raised capacitor on the photo above. may 0.5 mm higher.

The capacitor is located next to the PCI E card of an old Vista PC. The PC functions fine. ANd I have just removed 2 junk cards that was connected near the left and right of that capacitor (1 is a old dial up modem, 1 is a old TV card)

My question is, from the photo like the one posted above

http://i.imgur.com/ZUmVgNd.jpg

how many mth. would it be before capacitor of this nature totally bulge out w/ liquid coming out?

Or what is the chance that it stays slightly bulge for years?
 
I had a couple of ECS brand motherboards become unstable because of a single bulging capacitor, and it always in the voltage regulator for the DIMMs.

I'm guessing the circled capacitor isn't causing problems because it's a bypass capacitor, meaning it filters fairly steady DC, and there's nothing plugged into the slots adjacent to it that need their power feeds filtered. Generally bad caps cause problems when they're in a voltage regulator circuit, meaning near a coil (may be a donut with thick sold wire wrapped around it or a grey or black cube), such as around the CPU or DIMM slots, but often also near one of the PCI or AGP slots (between the battery and the diagonally mounted heatsink), Having said that, maybe another capacitor of the same brand (3 pie cuts in top indicate United Chemicon/Nippon Chemicon) and model is defective. Chemicon made models KZG and KZJ that were known to fail eventually, even if they didn't bulge on top (because they leaked out their bottoms through the rubber plugs there), while KZE was considered reliable. Normally if one brand and model fails, it's best to replace every capacitor of the same model, but that can be expensive, so maybe you should start by replacing just the ones in switching voltage regulator circuits and maybe the hottest chips, like that north bridge. New capacitors should match the characteristics of the originals when they were new, and with motherboards you generally have to use caps with very low ESR ratings. It's safest to replace just a few caps and test before replacing more because that makes it easier to track down any solder bridges and find out if the caps are incompatible.

That motherboard also has several Panasonic capacitors ("T" cut on tops), which are considered much more reliable than Chemicon KZG and KZJ. BadCaps.net has guides about replacing motherboard caps, both in its FAQ section and its forums. If you haven't soldered/desoldered motherboards, first practice on a junked board with 4+ layers of copper and cylindrical capacitors on it.
 
larrymoencurly said:
That motherboard also has several Panasonic capacitors ("T" cut on tops), which are considered much more reliable than Chemicon KZG and KZJ.
Click to expand...

http://www.badcaps.net/images/caps/bp6/index.html

you mean all those different cut on the top of each capacitor, they mean something?

Is there a table to list the break down?

and why is it that w/ some brand PC, such as dell, there is always a felt-tip marker (black or blue) to mark the surface of the capacitor? Is that to check if there is any plague? As they are new at the time, there shouldn't any need for that, is there?
 
RoGuE123 did you ever figure out what was causing the reboots?

Have you tried testing things not in the chassis? I had issues with a machine, replaced the motherboard for a friend . Machine kept rebooting, took the new motherboard back to the supplier and it worked perfectly. Turned out it was a faulty switch in the chassis causing the problem. I re tested the original motherboard and it had gone bad, but no visible faults like bad capacitors.

I have experienced clients complaining a machine is unstable and reboots and first thing I would check is the motherboard and often there would be multiple swollen leaking capacitors. I haven't come across these problems in the last few years.

Usually these are machines which are very old and need to be replaced.
 
You must log in or register to reply here.
Back
Top