Horrible pictures of dead caps

[trodas]

I would like to dedicate this thread to the most horrible pictures of dead caps you guys can find Just to scare these guys, who think that "this can never happen to them" Let's start with HP Pavilion machine pic. TMS 1500uF caps, machine shutting down in 30 sec after power on:

http://img406.imageshack.us/img406/4...tmscapsqx0.jpg

Frankly I would not dare to turn it on... Picture from Carcenomy user on TechPowerUp forums

[aghastpumpkin]

ARRRRRRRGGGGHHHHH Noooooooooooooooooooooooo! The poor wee souls! Burned alive! Tragic! Noooo! *cries*

[Wolf2000me]

http://www.badcaps.net/

A good read about this kind of thing. Luckily I don't have an example of my own

[Samuknow]

I have a couple of boards at home that have this problem. The one for the celery 1.8 shuts off right as it is booting into windows. Got the replacements, just have not changed them yet....

[trodas]

aghastpumpkin - burned alive? Well... what about PSU on fire thanks to Hec caps?

Klara's PC (Morphys friend, girl - but not girlfriend ), Eurocase PSU - Hec caps:

http://img99.imageshack.us/img99/712...action1cz6.jpg

http://img99.imageshack.us/img99/568...action2qu6.jpg

http://img141.imageshack.us/img141/9...action3ej0.jpg

[Áedán]

Sure it's the capacitors and not another failure mode in the PSU? Caps will die if you subject them outside their operating parameters.

[reconstrucdead]

When you got leaking caps on your MB then do your self a favour
also to tjek the caps in your PSU. When the MB run with one or more
bad caps,in some cases,it can course it to draw higher peakcurrents from the PSU and stressing some caps in it (and the other way around,not to forget!!). I have repaired many PC´s with badcaps on the MB
and in about 5 out of 10 times I had to replace one or more caps in the PSU also.

[trodas]

Áedán - entierly sure that this was the case of bad caps in PSU. The small Hec caps there are known to dry up and the smaller capacity they have, the stronger is the drive signal for mosfets... Where this lead to - you can see on my picture there:
Galerie
Say hello to bad caps

reconstrucdead - I could not agree more. PSUs are mostly the culprit. There is, also, except some very expensive server PSUs (eg. Zippy) no PSU w/o bad caps. And recapping only one does not help at all. You gotta replace all caps for respectable brands.


ATX Winner PSU, type PB300-A4S, courtesy by HoNY
http://img154.imageshack.us/img154/31/winnerpsu1op0.jpg
http://img228.imageshack.us/img228/3...nerpsu2zj1.jpg

CapXon caps rulez
http://img241.imageshack.us/img241/6...nerpsu3zi1.jpg

Bulging sligtly and leaking Vent cap, and some nonstandard inferior Fuhjyyu caps - 1650uF 16V? OMG!
Much to my surprise, the CapXon **** caps are used in Fortron PSUs:
SvÄ›t hardware - SrovnávacÃ* test 14 zdrojů - 1. Ä�ást - Fortron FSP400-60GLN

[cloasters]

Welcome to AOA Forums, reconstrucdead!

Didn't know about problem electrolytics inside PSU's. Are there replacement capacitors that aren't electrolytic?

[Gizmo]

Capacitors come in a lot of different types, ceramic, mylar, polypropylene, aluminum electrolytic, tantalum electrolytic (just to name some of the most popular kinds). The problem is that 'lytics are the only designs that give you large capacitance in a manageable size: e.g. a ceramic capacitor that could give you 100 µF of storage would be about the size of a dinner plate and would weigh several pounds.

Tantalum electrolytics are usually much more capable than aluminum electrolytics, but are also MUCH more expensive; It's not uncommon for a tantalum to cost as much or more than an aluminum that is an order of magnitude bigger. Consequently, tantalums are generally only used where nothing else will do.

Now, to get at what I suspect is driving your question.

Aluminum electrolytics are not inherently bad. As with many things, you have to use parts from a reputable supplier, AND you have to understand their limitations. Even when you use parts from a reputable supplier, you still have the possibility of getting bad parts occasionally, as with anything else.

Understanding the limitations of the parts is the bigger issue. Aluminum electrolytics dry out over time; that's just the nature of the beast. Operation at high temps will make them dry out faster, as will operation at elevated current levels, both of which tend to happen in power supplies, and to a lesser extent in motherboards, especially in the CPU vreg section. Good designs by competent engineers will tend to use parts that are somewhat overrated for the application in order to ensure a reasonable design life, but that costs more money. Hence, cheap designs tend to do three things that cause them to fail prematurely:

1. Source parts from the cheapest suppliers available: cost, not quality, becomes the overriding consideration.
2. Use parts that are of adequate capacity and voltage for the application when they are new, without allowing for derating as the parts age.
3. Use parts that are rated for 80C operating temps rather than 105C operating temps. Strictly speaking, an 80C part may be within tolerances for the application, but they also tend to break down more rapidly with operation at elevated temperatures. This gets back tonot allowing for derating of the parts as they age.

So, in general, don't blame electrolytics for being bad; they do what they are designed to do very well. Instead, blame the manufacturers who insist on using parts that barely meet spec when new in an effort to shave costs.

There's a reason quality PSUs cost more than bargain-basement units.

[cloasters]

Thank you for the explanation, Gizmo! Has the problem in electrolyics engendered by the incomplete, stolen formula for the electrolyte been laid to rest yet?

[Gizmo]

I surely hope so. That has been many, MANY moons ago, and the company that was making those parts went out of business (deservedly so). The only place I can imagine you could still get hold of those particular parts would be in surplus, and I would think they would be old enough by now that infant mortality would be killing them before they made it out of the factory.

[skool h8r]

Ah, don't you just LOVE solid-state caps ^^.

[Gizmo]

Solid-state capacitors are still electrolytic in nature. All that has changed is that instead of using aluminum foil, they use a polymer AFAIK.

Granted, that change makes for a longer lasting device, but it's still basically the same technology as is used with tantalum capacitors, just not as expensive as tantalum.

[trodas]

Gizmo - whoa, whoa, whoa Carefull out there. Technology make a great leap forward in ceramics.

ceramic capacitor that could give you 100 µF of storage would be about the size of a dinner plate and would weigh several pounds

This is a 100uF Taiyo Yuden Y5V 100uF 6.3V cap:
http://trodas.wz.cz/kondiky/cap_images/1173609032.jpg
The size is standard 1210, eg. 3,2 mm x 3,3 mm in size
You can purchase them in Digi-key, for example. A bit expensive, tough:
Digi-Key - 587-1397-1-ND (Taiyo Yuden - JMK325F107ZM-T)
But available. I have few for extreme overclocking - place them into the CPU socket

So, easy there, fella

Consequently, tantalums are generally only used where nothing else will do.

It is also almost impossible to make a very big tantalum capacitor. While Panasonic FC elytes line end up with 15 000uF, the very same company can't produce bigger that 220uF tantal caps.
Such size is useless in Vcore regulation, for example. Not to mention the step price... So, there are more that price reasons why not use tantaum caps. Besides, good elytes like Samxon GA line or Nicicon HZ line are hard to beat performance. Not to mention price.

And it is all not that easy. For example polymers aren't any rescue at all. If I oversimplify things, then polymers shine in hi-frequency regulation only and elytes "beat the crap out of polymers" at all commonly used regulators frequency. Especially in PSUs.
Hence if the Vcore regulator use a very high frequency - polymers is the choice for obvious reasons. If the frequency is common (eg. low), then elytes are much better choice by all cases and in results and lifetime as well. Please note that guaranted lifetime of polymers are only 1 000 to 2 000 hours. For example Panny FM d12.5 cap has 7 000 hours lifetime and Samxon RS elytes even more.

So to make long story short - for vast majority of regulators are elytes best and polymers simply is not good choice at all. There is no other hi-frequency that Vcore regulators currently used even in hi-end mobos. So yes, I say that mobos full of polymers are bad design. And I have proof of this:

Asus P5K-E - exploded Fujitsu polymer 270uF 16V on Vcore imput, just right after PSU...

http://i4.photobucket.com/albums/y13..._270uF_16V.jpg

Why? Well, because user NewBeetle from Xtremesystems forums did overclock with OCZ PSU and only a 520W one, and filled with Teapo caps...
http://www.xtremesystems.org/forums/showth...ad.php?t=151968

The imput Vcore caps should be ALWAYS elytes for best result - because they filter the ripple out of the low frequency PSU regulators...!

Now about the lifetime.
Caps have a life span. However it is like 2 000 to 7 000h per type (some even more) at the rated maximal temperature 105°C and with maximal ripple. Now for each 10°C down from the rated 105°C (for the 105°C caps) does the guaranteed lifetime DOUBLE (still with maximal ripple!) Hence in the absolutely WORST case:
105°C - 2 000h
95°C - 4 000h
85°C - 8 000h
75°C - 16 000h
65°C - 32 000h
55°C - 64 000h
45°C - 128 000h
Now bear in mind, that year has only 8 544h Hence well-cooled cap has to stay alive at least 14,9 years with every parameter in limits. Caps failing in good environment in like year or two are most common. Some excellent designs with Teapos (mine ABit ST6R) make 5 years. But that is still far from what they should last.
Hence they are bad caps.

Some bad caps fail and leak even w/o actually being used. These are even inferior brands of bad caps and caps like GSC, G-Luxon and Fuhjyyu are among them. From the better side of bad caps are OST, Teapo and few others.

Yet still they are bad caps and ANY hardware that use them (especially PSUs!) will fail and in the process, will make YOUR mainboard fail. Overclocking and folding accelerate the process, but it is inevitable.

Furthermore, they did not "corrode" at all. What is happening is, that cap is made full of electrolyte. This is water with additives. Good additives prevent the water from breakdown to Oxygen and Hydrogen when current is being drawn. Bad ones - cheapo ones - did not prevent it much, so inside the cap is building a pressure. Hence bad caps start bulging and eventually it end up in explosion, because Hydrogen and Oxygen mix is extremely explosive.
Check out what happen to Space Shuttle Challenger. Yea, that was it.
Some caps just dry out, some explode at the end. Teapos usually dry out (no leaking, bulging!) and G-Luxons usually explode. That is why there are these vents for - so the force of the explosion is low and directed directly up, not damaging PCB.
Once cap explode near your face and you lose your hearing for a while, you will know that caps can be very dangerous things and it is not wise to confuse corrosion and explosion

Asia-X caps in OCZ PSU:
http://img352.imageshack.us/img352/7...soczpsuot2.jpg
User spyder from BadCaps forum.


PS. yes, polymers with lower temperatures extend their guaranted lifetime as well and they even claim that they do extend it MORE that 2x, but I did not fully believe this marketing informations.
And many polymers out there are plain and simply - fakes. As with elytes. So, it is a jungle out there. And there are just like 5 good caps brand and the rest are known to be bad caps... So, it is not like Gizmo says "That has been many, MANY moons ago, and the company that was making those parts went out of business (deservedly so)."

That was a lie, even I do wish it is true, it is not. There are like hundred (if not more) companies making bad caps and selling them under different names and brands. Just look at this discovery:

In short is says that the old-known terrible GSC caps are still out there, because: GSC = Evercon = Sacon = Capsun.
This is backed with solid evidence you can see for yourself in many and many images and caps types database, and screenshots...

It is ALSO proven by the fact, that they DO fail like the inferior GSC crap did years and years ago:

Sacon FZ caps on Magic-pro - it is a JetWay OEM:
http://i4.photobucket.com/albums/y13...on_FZ_caps.jpg
Don't let them fool you. They are NOT polymers, these are sleevless electrolyte caps, and a bad ones, I might add. JetWay do this all the time. Remember their mobos full of GSC crap-caps?

So, sorry guys for rather long post full of bad news. I think you deserve the truth, even it hurts badly. "Fun" fact: there is no PSU with good caps now in the market, except a server-grade PSUs like Zippy made.

But do not trust me, ask others with the same experiences like mine

[Gizmo]

Originally Posted by trodas

Gizmo - whoa, whoa, whoa Carefull out there. Technology make a great leap forward in ceramics.

Hmm.......I admit it has been a while since I was in the game. The last I knew, surface-mounts in that size stopped about about .1 uF, but that was 15 years ago.....

Now bear in mind, that year has only 8 544h Hence well-cooled cap has to stay alive at least 14,9 years with every parameter in limits. Caps failing in good environment in like year or two are most common. Some excellent designs with Teapos (mine ABit ST6R) make 5 years. But that is still far from what they should last.
Hence they are bad caps.

I hope we are getting tripped up by language here.

A 'bad cap' is one that fails prematurely, when operated within it's normal parameters. Now, again, it's been a while since I was in the game, but I can't imagine the likes of Rubycon, Nichicon, or United Chemicon putting out parts that fail prematurely, even today. I submit that the vast, VAST majority of 'bad caps' are in fact not being operated within their design parameters, and THAT is why they are failing. I base that opinion on several years spent doing failure analysis and design of switch-mode power supplies. Granted, that experience was 15 years ago, but the basic principles of switching technology remain unchanged; you use a switching transistor to drive current through an inductor, charging a field. When you turn off the transistor, the field of the inductor collapses and the resulting current flow is used to charge a capacitor. By varying the frequency and/or duration of the switching pulse from the transistor, you control the voltage impressed upon the capacitor.

In addition, my own observations of various mobos and PSUs that I have disassembled has shown that the components that tend to fail are in fact being operated at or very near to their design limits. As you yourself have already noted, even quality parts are going to fail after about 8,000 or less (i.e. 1 year) when operated near their limits, and parts can be operated near those limits without the user even realizing it. Case in point: my Abit NF7-S mobo, running water cooling on the CPU, was running the VREG at about 100C because there was no air flow across that part of the mobo. Had I continued operating it that way, I would have had a dead board in less than a year (actually, I'd probably have written it off as a crap mobo when it started getting unstable in about 4 months). Instead, I put a fan on that section, got the operating temp down to about 45C, and I ran it for 4 years, until it was finally destroyed by a fire I had in my house.

In short, parts generally don't fail prematurely (with the exception of a small percentage of 'infant mortality', which a good manufacturing process will weed out), unless they are being operated outside their design limits. When you start seeing a rash of such failures, you have to look at why the design is pushing the parts past their limit, not just assume the part is 'bad'.