cap plague

Last year, I was asked to recap a an obsolete and non replaceable motherboard (would have required upgrading an old system with a completely new version for big bucks and new software). Several caps were bulged as shown in this article and the values were low, making the regulators unstable, but at least, none had yet vented.

Getting those leads out of a 6 layer, board without damage was a real challenge. But new, high quality caps did the trick and the board has been back in service for about a year without any more problems.

Reply to
John Popelish
Loading thread data ...

Yes, that is / was a BIG problem. I've recapped several motherboards over the years, as it was easier then replacing the motherboard with another model and having to spend hours installing software.

I also recapped all the servo amplifiers (x 7?) in one of the machines we have, as they all started leaking (120 some odd caps in total, some in very difficult areas to reach, as large components had to be removed / desoldered to get at them)

Another interesting failure I sometimes see is resistors with a large voltage differential going open! Sometimes the resistor is dissipating several watts, other times it's dissipating several uW. Removing the resistors coating shows that the resistive material had disappeared in a small area!

Reply to
Jeff L

"John Larkin"

** Had to perform some emergency surgery on my PC last Friday morning.

It refused to complete boot up and presented a totally black screen when switched off & re-tried.

Same scenario repeated exactly after 20 minutes of being off.

Suspecting a " heat "issue - I got out the vac and gave the insides a good clean, was pretty bad around the CPU heatsink & fan.

Tried again - same scenario.

Damn !!

Then, I noticed some VERY hot mosfets in the CPU regulators and some wonky looking electros next to them - one 1000uF, 6.3 volt Sanyo and two 1200 uF, 6.3 volt " G S C " brand.

Extracted some similar electros from another ( unused) motherboard and replaced the ones in my PC - these new caps tested extremely low ESR, about 30 - 40 milliohms.

Voila - the PC works fine again and those mosfets now run only warm.

Checked out the faulty electros and found they were all open circuit, the best showed an ESR of about 10 kohms !!!

How the heck could it function with O/C caps like that ??

Moral:

Check those hard working regulator electros from time to time and replace them all at the first sign of cap bulging or hot mosfets.

....... Phil

Reply to
Phil Allison

Happened to my VIA P4PB 400-FL (rev B2) motherboard after around 2 years of service. Had a bitch of a time finding a replacement MB that would run my video card (AllInWonder TV in etc). After trying many other vid cards in one of the MBs, research indicated that *NOBODY* specifies which protocol / spec the video card supports (most MBs seem to do that). Real bitchy. But i was lucky in that i found (new looking and sealed) VIA P4PB

400-FL (rev C2) MBs and got two. Bought some chip caps to parallel the same-size large caps that went bad on the original board. Despite the 2-year time frame and knowledge of the bad cap problem, VIA chose to use the *same* brand caps (GSC, one of the known offenders as well). Seems to me that a parallel chip cap will bypass those switching spikes and so keep that stress away from the poor aluminums - and so they should last a lot longer. I posted the fix in ViaArena, attempting to use ASCII graphics. Turns out their system ignores spaces between (?non-alfa?) characters, so i tried another one that looked like a space in the source,but got converted to a wierd character; so i gave up on that. They do not allow pictures so too damn bad; if anyone in ViaArena wants a picture they can either find me or attempt to complain to VIA.
Reply to
Robert Baer

Teleportation?

Reply to
Robert Baer

Better yet, on a new system (or after replacement), parallel themwith the largest physical chip cap that will "reach" the cap pins on the bottom of the PCB. Materially decreases switching (spike) currents thru the 'lytics.

Reply to
Robert Baer

"Jeff L"

** There are two cases of ( usually open cct) failure I see quite a lot:

  1. Metal and carbon film resistors ( 1/2W & 1 W) of from 22k to 220k with applied DC voltages of from 50 to 250 volts - as found in power amps, tube amps, scope vertical amps, SMPS kick starts etc.

  2. Metal and carbon film resistors of 1M ohm to 22 Mohms with applied voltages of 500 to 2000 volts - ie monitor focus dividers.

....... Phil

Reply to
Phil Allison

I believe it was Nichcon that had posted a warning on it web site about cheap copies of its electrolytics. Not sure its related but the two seem to go hand in hand.

Cheers

Reply to
Martin Riddle

Even Nichcon caps leak - I had to recap something like 127 of their 125 deg C caps after doing a repair (open resistor - mentioned in other post) on a servo amp and noticing almost all of them leaking in that unit. The caps give off a characteristic cat pee smell. The other 4 or 5 amps had the same caps leaking. The servo amp was approaching 10 years old at the time. About

6000 hours of use later, the amps are still working fine and will be soon be decommissioned with the machine.

to

message

Reply to
Jeff L

In the case of true high voltage circuits, this is common; it's corrosion. Not the usual salt-water sort, but St. Elmo's fire is also an ionic environment, and chemical attack is not picky about the difference.

The screen adjust resistors on old TVs and video monitors are notorious for failing in this way. Easy fault to diagnose and track down, too!

Reply to
whit3rd

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.