PDA

View Full Version : Best Aluminum Electrolytics?



Chuckster_in_Jax
February 29th, 2012, 03:36 PM
I have a Gateway2000 486 machine with a bad motherboard. Upon testing, I found several shorted Aluminum Electrolytic capacitors.
I also have 2 motherboards go out in the last year. Both of the boards were less than 6 years old. One of them has bulging top vents on caps next to the CPU.

Some Google searches indicated that capacitors that are manufactured in Japan are of much higher quality that ones manufactured in China or Taiwan. Check out the table of Japanese manufacturers on page 3 and the Chinese manufacturers on page 4.

http://www.hardwaresecrets.com/article/How-to-Identify-Japanese-Electrolytic-Capacitors/595/1

What manufacturer(s) are known for high reliability capacitors other than the ones on the table in the linked article?
Ones to stay away from?

Tr3vor
February 29th, 2012, 03:38 PM
I've heard that caps from the early to mid 2000's were prone to failure and they are better now, but I don't know for sure.
maybe in the mid-late 90s too, something like that.

Chuck(G)
February 29th, 2012, 05:20 PM
I stick with the Japanese brands; Nichicon, Panasonic, Elna, etc. I stay away from Rubycon because there are just too many counterfeits of that brand out there.

Sprague and Illinois are also brands that, although not strictly Japanese (mostly Taiwanese) have reputable firms behind them (e.g., Vishay=Sprague), so I wouldn't have too many qualms about using them.

pearce_jj
February 29th, 2012, 11:25 PM
For motherboard use the caps should also be low-ESR types I think.

Ole Juul
March 1st, 2012, 01:57 AM
For motherboard use the caps should also be low-ESR types I think.

Why would that be? Perhaps in a low power critical application like a cell phone, but a mother board?

pearce_jj
March 1st, 2012, 02:08 AM
It's just what I've read, for example Badcaps.net FAQ (see 2.5) (http://www.badcaps.net/forum/showthread.php?t=425).

Chuckster_in_Jax
March 1st, 2012, 11:28 AM
It's just what I've read, for example Badcaps.net FAQ (see 2.5) (http://www.badcaps.net/forum/showthread.php?t=425).

I signed up with this group yesterday and spent some time going through their posts. There is a lot of good info here with a repository for members to report experiences with higher than usual failure rates with certain manufacturers.

Another interesting read is about "Capacitor Life" on Chemi-Con's website. Goes through a lot of math and circuit analysis, but summarizes the main causes of capacitor failure. Temperature (ambient temperature and internal heating due to ripple current) is the most critical to the life of aluminum electrolytic capacitors.

http://www.chemi-con.com/u7002/life_ms.php

I wasn't aware of the negative effects of ripple current on motherboard capacitors. Cheap power supplies and ones that are failing produce high ripple on their output that result in excessive heat in the decoupling capacitors on motherboards. This leads to premature failure.
So after I replace the failing capacitors, I need to put an oscilloscope on the power supply and check for excessive ripple.

Chuck(G)
March 1st, 2012, 01:32 PM
It's not just ripple; it's high-frequency ripple that does them in. A filter cap in a 60Hz analog supply will last a long time, but put the same cap in a 40KHz SMPSU and it'll be toast in short order.

Some motherboards use "buck" regulators for low-voltage components and the associated caps are targets for ripple-caused heating as well.

Ole Juul
March 1st, 2012, 01:50 PM
It's not just ripple; it's high-frequency ripple that does them in. A filter cap in a 60Hz analog supply will last a long time, but put the same cap in a 40KHz SMPSU and it'll be toast in short order.

Some motherboards use "buck" regulators for low-voltage components and the associated caps are targets for ripple-caused heating as well.

I suspect that the main problem with motherboards is that the industry runs on low margin. To my way of thinking, good design would minimize the ripple, but that would cost another couple of bucks. Also, it is standard practice in RF to also use a small value ceramic or similar in parallel so as to avoid the high frequency leak. Of course, there's another buck or two.