Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Capacitor Plagues - Premature Electrolytic Capacitor Failure

Status
Not open for further replies.

Laplacian

Electrical
Jul 15, 2002
246
0
0
US
Over the past several years, our industrial facility has experienced many electronics failures where the root cause has been capacitor failure due to electrolyte dry-out. The problem has been seen in power supplies for PLC's, protective relays, and process control equipment installed in the later 1990's.

I began investigating a recent incident, and suddenly remembered all the issues around motherboards and computer power supplies a few years ago. Our equipment is the same age as the failed computer equipment. I found an interesting wikipedia article that mentioned an IEEE Spectrum article from February 2003, titled "Bad Capacitors Mucking Up Motherboards".

This article only dealt with the personal computer side if the issue.

The goal of this posting (which will be cross posted into relevant forums) is to determine the extent of industrial failures and to estimate an approximate window of when the bad equipment was installed. This is due to the relative silence from the capacitor and equipment manufacturers on the issue.

The questions are as follows:

1. Has your facility experienced multiple failures with the root cause being bad capacitors causing voltage instability?

2. If yes to #1, what year was the device manufactured?

3. If yes to #1, what as the manufacturer (if it is appropriate to tell) or what was the equipment type.

Please also share any further information you may have on the subject. We are looking to determine the extent of this phenomena and our exposure to future events; one event can cost our industry several M$'s. Thanks in advance for any information.
 
Replies continue below

Recommended for you

Dunno about PLC's and such, but I have been replacing
caps in SMPS for years. Seems like the boards (of any
type) that show up most are the ones that a) the caps
are highly stressed (poor design, hi-temp environment,
marginal quality, etc...), or b) just plain cheap caps
(mostly on consumer-type gear, i.e., computer power
supplies, mainboards, peripherals.
I sort of imagine that it is the same situation in
any industry; cost-competitiveness and/or marginal
design. I have some 15 year-old IBM boards that I
routinely rob parts off to repair 3 and 4 year-old
equivalents made by someone else. Quality counts :)

Seems to me that caps used to be caps. Get close to
the right value, use a higher voltage if possible,
and you were in good shape. Nowadays, you worry
more about ESR, temp. rating, and physical size.
Not to mention the vendor. There ARE some awfully
poor quality components.
<als>

 
Electrolytic Caps are notorious for failing. I read somewhere (don't recall where now) 70%+ of all power supply failures are due to to electrolytic caps. Just look at the tolerance values written on them and you see why.
 
I am by no means a capacitor expert.

I do believe that the study of capacitors may involve a lot more than what has been discussed in this thread. There are many many types of capacitors and "electrolytic" still is a very broad category with many subdivisions.

An in-depth discussion of the types including some discussion of reliability, shelf life etc is available FREE here:


=====================================
Eng-tips forums: The best place on the web for engineering discussions.
 
I should clarify I was by no means criticizing any response. Just indicating there is probably a lot more to be gained by looking more specifically at the types of capacitors.

What to be gained....I don't know... I haven't read them all and don't know all those terms.

I should mention there is also an EPRI document 112175 with a tremendous amount of information on capactiros, quite a bit more than even the mil-handbook I think. It is available to members of EPRI.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.
 
Temperature is the big killer, most caps are only rated at 5000hrs at rated max temp which isn't very long in an industrial setting. You have a choice, keep them cool or routinly replace them.
 
And the thing which causes temperature rise is ripple current: keep away from the maximum ratings by oversizing the cap or (better) by paralleling several smaller caps. Use the 105[&#176;]C rated types, but design as though you're using the 85[&#176;]C types.

Good link Pete - thanks.

----------------------------------
image.php
I don't suffer from insanity. I enjoy it...
 
This is not completely off-topic, so bear with me...

My six year old air handler's motor stopped rotating. It's really a 230V three phase motor, fed by six FETs from rectified 230V single phase power, by means of a sensorless controller bolted on the end. One of the two biggest capacitors (with way bigger ufd ratings than I would have expected given the package size) in the controller showed signs of overheating, so I replaced both of them. That didn't fix the problem. Nobody sells a replacement controller. Nobody stocks a replacement motor/controller assembly, and the price I've been quoted for delivery at some uncertain date in the future greatly exceeds the price of an entire air handler, with the high efficiency load sensing motor, evaporator coil, controls, sheet metal, etc.

Luckily, the weather is nice here this time of year, so I'm taking a look at buying an industrial sensorless 1ph/3ph controller and adapting it to the existing, apparently undamaged, motor. I've downloaded the full manuals for a few similar controllers, and they all agree that capacitor replacement is a normal maintenance item, to be expected at five year intervals. That surprised me.



Mike Halloran
Pembroke Pines, FL, USA
 
electricpete,

The original intent of this thread was to determine if there have been other industrial users that have been affected by the "capacitor plague" issue in the late 1990's and early 2000's (that really crippled a few computer mother board manufacturers). If they had been, what devices were these capacitors installed in. We have several of one particular manufacturer's relay installed that were affected. These are electronic feeder and motor protection relays; both relay types shared a common power supply design that was sub-assembled by the same Southwest Asian electronics manufacturer.

I understand capacitor manufacturer claims of 5 year design life, but we have significantly older equipment (DC drives, heater controllers, large UPS systems, etc.) that have the same type of capacitor but live on for 10 - 20 years with no issues.

What good is a feeder protective relay if you have to worry about reliability every five years? Simply performing calibration checks may or may not expose any issues. Our company has been told by more than one relay manufacturer that thier relays are good for 20 years. If better reliability can't be achieved by today's current design of "value engineered" electronic products, then I believe we will keep the old electromechanical relay manufacturers in business for the really critical applications.
 
Status
Not open for further replies.
Back
Top