"Arny Krueger" wrote in message
. ..
"ScottW" wrote in message
...
"Trevor Wilson" wrote in message
...
**Actually, I suspect you do. Electrlytic capacitors begin wearing out,
as soon as the equipment is turned on. Many caps destined for domestic
products have a very short (published) life span. Certainly less than
5,000 hours.
Couple of points.
1) These short lives are at very high temps. 80c.
No domestic product has temps that high...they'd have to
provide thermal shield to meet U/L safety requirements.
So the cap life doubles with every 10c decrease in temp.
Get down to a quite reasonable 40c and you're looking
at 40,000 hours...
Agreed.
2) The primary wearout failure is loss electrolyte causing a
a drop in capacitance with failure usually spec'd at 80% of
rated value. Most electrolytics are applied as filters which
aren't terribly value sensitive.
Electrolytics in a crossovers value sensitive applications would
be very poor design.
Agreed.
If I was presented with a 20 year old piece of equipment, with a large
number of caps, some of which had failed, I would advise replacement of
most/all electros as a matter of course.The reason for this is that I do
not like to see my clients return with the same equipment, for similar
faults, within a year or two of service. I'd much prefer that my repair
last AT LEAST as long as the equipment had lasted to date. Preferably
longer. As a rule, I always use superior quality replacement caps, where
possible.
Agreed up to the point where a ludicrous number of parts are involved,
like 150 caps in a simple stereo active crossover.
**You keep parrotting this statement, like it makes sense. Neither you, nor
I have seen the product in question. Let's examine what is involved (worst
case scenario):
Linn probably use OP amps in the unit. It is, as I understand it, a three
way crossover. If it uses balanced circuitry, the complexity and the number
of caps could be prodigious. Each OP amp will have a pair of decoupling
caps. The power supply would likely contain 6 or 8 caps, at a bare minimum.
Etc, etc. 150 caps is within the realm of possibility.
The still-excessive failure rate may help explain Linn's continual
redesign of its active crossovers from discrete components to circuit
boards with plug-in modules to plug-in modules placed directly in the
power amp.
Failure of 150 parts in an electronic crossover, even after 20 years of
24/7 use is characteristic of junk-as-built.
**Nope. It is quite common. Ten years is about what we expect from a
permanent power supply in an old VCR.
Irrelevant because VCRs aren't comparable to electronic crossovers.
**Sure they are. They use electrolytic caps and are permanently powered.
Every
VCR I've scrapped had worn-out heads.
**I suggest you examine some 20 year old VCRs. Don't forget: Stephen's
crossover is 20 years old.
Cap failure is a normal event. Particularly for products which operate
at elevated temperatures. Even quality brands like Sony and Panasonic
are so afflicted. Quite a large number of engineers appear to treat
caps as if they are 'blcak box' passive components, which are
impervious to heat. They're wrong, of course.
Come on Trevor, this is a high end electronic crossover, not a power amp
or piece of consumer mid-fi.
**So? I've measured transistors in a preamp operating at 75oC. And there was
no ventilation holes in the case either. Of course, that was an issue of
very abd design. Still, I've seen soem pretty dumb design choices in mass
market products. In fact, on Tuesday, I worked on a cheap Sony CD player,
where a powewr supply cap was stupidly placed adjacent to a hot running
regulator. It was stuffed.
Interesting that 'same price for better specs' translates to
'fantastic
markup' for you.
Stephen, can't read can you? Or is it that context means nothing
to you?
The above is irrelevant to what I said, and more strikingly, it is
irrelevant to what you said.
It's a paraphrase of what you said in reply to what I said. I agree
that
it didn't make sense in context.
You can buy motor oil cheaper in the store than at the
mechanic's, too.
Of course, but irrelevant.
Exactly relevant to the cost of replacement parts "obtained
competently"
by the tech vs the cost to the consumer installed as a repair.
That wasn't what I was talking about. I was talking about the
incremental
cost of a higher-spec part which is what you started out talking
about.
Yes, I said I was able to get the higher-spec'd parts installed for
the
supposed cost of a Linn renewal to original specs. Then you said the
repairman got a 'fantastic markup' which doesn't make sense,
It makes as much sense as you care to understand.
As things stand, the whole story is strange.
The number of parts (150 capacitors!) that were replaced was way out of
proportion to reality for either a passive or active crossover.
**Not necessarily. I've not sighted the crossover in question, but it is
entirely possible that Linn have used a large number of decoupling caps
in their design.
All things are possible, but 150 decoupling caps in an simpe 2-channel
electronic crossover is ludicrous.
**Have you examined the crossover? I haven't. I can conceive of the
possibilty that 150 caps were used. It is an analogue, three way crossover
and it is 20 years old, don't forget.
I've worked on a few crossovers and some use rather prodigious numbers
of eletros in their design (usually as decoupling caps).
I guess some people do crazy things. For example, look at the power supply
caps in a ME power amp!
**It seems your memory is faulty. I will explain why so many caps are used.
* ME amps use no Global NFB. As a consequence, their quality is heavily
dependent on power supply quality.
* Capacitors possess several paramaters which are of interest to designers:
* Capacitance.
* Ripple current rating.
* ESR (Equivalent Series Resistance)
* ESL (Equivalent Series Inductance)
Using a large number of capacitors acheives several things for the
manufacturer:
* Due to the packing efficiency of a large number of small caps,
more capacitance can be squeezed into a given chassis size, than several
large (round) caps. The effect, if you like, is to have a large, almost
rectangular block of caps.
* Ripple current is increased massively, thus enhancing the life
span of the caps.
* ESR is reduced, due to the large number of paralleled caps.
* ESL is reduced due to the large number of paralleled caps.
* A side benefit has been the increased longevity of the caps, due to the
both the paralleled nature of the caps and the fan cooling used in later
models.
Why is a dc block needed in an active crossover?
No, these are decoupling caps, for the power supply. A large number of
electrolytic decoupling caps are not needed - the decoupling caps that
show up in any volume at all are ceramics or film caps, and those
typically have much longer lives than electrolytics.
**Nope. Most domestic products use electros as decoupling elements.
Sometimes, these will be bypassed with small value film caps as well.
Anyway...electrolytics not being symmetrical with polarity are
an awful choice for an ac coupling/decoupling application.
Not true.
**Agreed.
See this guide.
http://www.electrocube.com/support/bullet3.asp
I don't see where this document supports your claims about electrolytics
being bad for coupling, Scott.
**Also agreed.
Trevor Wilson