Isn't that where "if it ain't bust, don't fix it" came from?
I've never had a capacitor fail in over 30 years, that includes running cars
for six years and 70k miles, and until I replaced the distributor on the
roadster last year hadn't changed one either.
Hunt's 4th Law of MG ownership "Many break-downs occur soon after a car has
been worked on; 'new' parts can be faulty when you receive them; 'new' parts
will sometimes fail soon after fitting; 'new' parts almost certainly won't
last as long as the originals"
PaulH.
----- Original Message -----
From: Simon Matthews <simon_atwork@hotmail.com>
To: <mgs@autox.team.net>
Sent: Thursday, January 20, 2000 1:51 AM
Subject: Changing condensers/capacitors
> I do not recommend changing points capacitors and would like to solicit
> comments to see if people agree/disagree with me.
>
> Explanation:
> Most electronic devices exhibit a 'bathtub' failure characteristic.
> THis can be explained in 3 phases:
> 1. Infant Mortality. Devices exhibit a high failure rate initially.
> Probably due to manufacturing defects which do not show up on test.
> THis initial rate declines over time until one reaches the second
> stage.
> 2. Normal operating life. Random failures. Components exhibit a
> low level of random failures. This is measured as the MTBF (mean
> time between failures). This low rate continues for the component's
> operating life until it reaches the final stage:
> 3. End of life. The components wear out and failure rates increase again
as
> the devices age.
>
> Now the key question is how long are these stages? I have never changed a
> capacitor and have never had one fail. Conversely, I have read of many
> people on this list who put in new capacitors and had them fail very
> quickly. However, when do end-of-life failures occur?
> One posting suggested that they can last at least 30 or so years. Can
anyone
> provide any data points on how long capacitors which get through the
infant
> mortality stage actually last?
>
> I suggest that, if you replace capacitors on a routine basis, you will
> constantly run the risk of an early failure, whereas, if you leave it
there
> for 10-20 years, your risk of failure will be less.
>
> Arguments, anyone? Facts and data welcome.
>
> Simon
>
|