Capacitor lifetime formula

Guest
Hi,

I have recently been servicing some equipment that has had failures of
a 1000uF 40v 85 Deg C capacitor, the equipment has been running for
about 5 years and are used in the smoothing circuit of a linear power
supply. .

I have measured the ripple current (with a true RMS meter) of about
280ma which is well within limits and can only assume that the CAN
temperature of the capacitor of around 55 deg C is too high. The
capacitor is between a transformer and the heat sink of a regulator.

My question is how do I calculate the life of the capacitor with the
temperature of 55 Deg C? I have been using the 10 Deg increase in
temperature halves the life of the capacitor rule (1.82 years in this
case) but is there a better way of showing this? I have to go to my
client and tell them they have a problem but I need some better
formula to show them (or maybe I do not?).

Any help would be greatly appreciated.
 
I have recently been servicing some equipment that has had failures of
a 1000uF 40v 85 Deg C capacitor, the equipment has been running for
about 5 years and are used in the smoothing circuit of a linear power
supply. .
Is there any chance that this is one of the Taiwanese-made capacitors
which have been failing in large numbers due to the use of a defective
electrolyte formula? What brand it is?

--
Dave Platt <dplatt@radagast.org> AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
 
On Tue, 23 Mar 2004 16:14:28 -0000, dplatt@radagast.org (Dave Platt)
wrote:

I have recently been servicing some equipment that has had failures of
a 1000uF 40v 85 Deg C capacitor, the equipment has been running for
about 5 years and are used in the smoothing circuit of a linear power
supply. .

Is there any chance that this is one of the Taiwanese-made capacitors
which have been failing in large numbers due to the use of a defective
electrolyte formula? What brand it is?
nover, but I thinkits an enviromental problem, not the manufacturers
 
In article <rvi060p4p6ke8132qikua2i12ic6c8as6l@4ax.com>, q@q.com wrote:

Hi,

I have recently been servicing some equipment that has had failures of
a 1000uF 40v 85 Deg C capacitor, the equipment has been running for
about 5 years and are used in the smoothing circuit of a linear power
supply. .

I have measured the ripple current (with a true RMS meter) of about
280ma which is well within limits and can only assume that the CAN
temperature of the capacitor of around 55 deg C is too high. The
capacitor is between a transformer and the heat sink of a regulator.

My question is how do I calculate the life of the capacitor with the
temperature of 55 Deg C? I have been using the 10 Deg increase in
temperature halves the life of the capacitor rule (1.82 years in this
case) but is there a better way of showing this? I have to go to my
client and tell them they have a problem but I need some better
formula to show them (or maybe I do not?).

Any help would be greatly appreciated.
I did a google search on "arrhenius +reliability +capacitor" and found
loads of information. Most of it relies on using failure data from
accelerated testing to determine life. You might be able to find a chart
from your capacitor manufacturer. Good luck.

Al

--
There's never enough time to do it right the first time.......
 
My question is how do I calculate the life of the capacitor with the
temperature of 55 Deg C? I have been using the 10 Deg increase in
temperature halves the life of the capacitor rule (1.82 years in this
case) but is there a better way of showing this? I have to go to my
client and tell them they have a problem but I need some better
formula to show them (or maybe I do not?).

Any help would be greatly appreciated.

You have the right starting point, but do you know the ratings of the cap
that you're using?

Jim Williams gives a more accurate formula in one of his books, IIRC 5-7
terms that have to be taken into account, including the cap's rated ripple
current vs the circuit's ripple current, the frequency of the ripple, cap
impedance etc..
 
On Tue, 23 Mar 2004 16:22:23 GMT, q@q.com wrote:

On Tue, 23 Mar 2004 16:14:28 -0000, dplatt@radagast.org (Dave Platt)
wrote:

I have recently been servicing some equipment that has had failures of
a 1000uF 40v 85 Deg C capacitor, the equipment has been running for
about 5 years and are used in the smoothing circuit of a linear power
supply. .

Is there any chance that this is one of the Taiwanese-made capacitors
which have been failing in large numbers due to the use of a defective
electrolyte formula? What brand it is?

nover, but I thinkits an enviromental problem, not the manufacturers

Suggest you do some research on the cap at
http://www.nover.com/home.asp

A measurement of RMS ripple doesn't show the real ripple amplitude,
you should have a look at it with a cro. If the cap is one of Nover's
general purpose types and used as the bulk filter in a dc supply as
you seem to suggest, it is not likely to have a long service life if
it is running at 55C constantly.

In my experience caps used in this position should be long life and/or
high ripple 105C types, and the physically larger the better.
Generally speaking, I think a larger can means a greater volume of
electrolyte, larger surface area to disipate heat, and hence longer
life. These days I am convinced all cap manufacturers try to compete
with each other to see who can make the largest cap in the smallest
can, and in real life this does not make for long term reliability
especially in power supplies. Running a physically small cap at
relatively high temperature over long periods doesn't take long to dry
out the small volume of electrolyte in the can, despite what the
manufacturers claim.

I have found that when I replace one of these modern small can types
with a 20 year old good quality Philips, Panasonic or Rifa 85C type,
the repaired item never comes back. This assumes there is room to
mount the physically larger older device.

From what you say about the location of the cap it has both high
ripple and high temperature from the adjacent heatsink to deal with.
If the cap is closer than 12mm from the heatsink then this is too
close in my opinion. The temperature of the heatsink will have a large
bearing on it though. If I am using 85C caps I don't like them to run
at more than 35C max if I want a fair service life.
 
I went to a Sprague seminar, many years ago. They said keep them
cool, and don't run them at more than half their rated voltage.
A very low ESR would prevent heating, as well.
Dan

On Tue, 23 Mar 2004 14:37:39 GMT, q@q.com wrote:

Hi,

I have recently been servicing some equipment that has had failures of
a 1000uF 40v 85 Deg C capacitor, the equipment has been running for
about 5 years and are used in the smoothing circuit of a linear power
supply. .

I have measured the ripple current (with a true RMS meter) of about
280ma which is well within limits and can only assume that the CAN
temperature of the capacitor of around 55 deg C is too high. The
capacitor is between a transformer and the heat sink of a regulator.

My question is how do I calculate the life of the capacitor with the
temperature of 55 Deg C? I have been using the 10 Deg increase in
temperature halves the life of the capacitor rule (1.82 years in this
case) but is there a better way of showing this? I have to go to my
client and tell them they have a problem but I need some better
formula to show them (or maybe I do not?).

Any help would be greatly appreciated.
Colorado Springs, CO
My advice may be worth what you paid for it.
 

Welcome to EDABoard.com

Sponsor

Back
Top