Audio Empire wrote:
> On Tue, 9 Mar 2010 17:55:50 -0800, Dick Pierce wrote
> (in article <(E-Mail Removed)>):
>>Let me illustrate a specific and demonstrable case where swapping
>>and ostensibly "better" part is garaunteed to result in worse
>>performance.
>>
>>One of my clients had me design a network for their speaker
>>that used the Vifa 1" ring-dome radiator tweeter. It does not
>>use ferrofluid in the gap, with the net result that its
>>impedance peak at resonance is fairly sizable, on the order of
>>25 ohms or so at about 650 Hz. This was enough that it would
>>screw up any plausible 2nd- or 3rd- order network and without
>>impedance compensation, it is impossible to achieve the design
>>target for the crossover response.
>>
>>So, an onjugate network, consisting of a series LRC was designed
>>to go in parallel with the tweeter. The L and the C value depend
>>upon the effective mass and compliance of the driver, and the
>>R should be about equal to the DC resistance of the voice coil,
>>about 3.4 ohms. Since the L and R are in series, our good friend
>>Thevenin states that it makes no difference whether it's L followed
>>by R, R followed by L, 10 1/10Rs interspersed with 10 1/10Ls or,
>>in fact, an L whose DC resistance is 3.4 ohms. It just so happens
>>that I found an of the shelf air-core inductor wound of the right
>>value whose DC resistance was within about 3% of the value I needed,
>>meaning I could use 1 part instead of 2 AND because it was wound
>>with small wire, it was substantially less expensive to implement
>>it using a single high-resistance coil instead of a low-resistance
>>coil and a high-resistance resistor.
>>
>>However, a number of people, when they opened up the speaker, saw
>>all these wonderful poly caps and this shitty little inductor. The
>>first step was to rip out the inductor and replace it with a hefty
>>low-resistance inductor.
>>
>>Doing so completely screwed up the crossover, because it was
>>suposed to be a 3.4 ohm resistor in series with a 4 mH inductor,
>>NOT a 0.15 ohm resistor in seriers with a 4 mH inductor.
>>
>>In a similar vein, a popular "tweak" of the old KEF RS104-AB
>>network was to rip out all the electrolytic caps and replace
>>them with polypropylenes and, yes, it DID change the sound,
>>but NOT necessarily for the better, because the ENTIRE crossover
>>was designed with knowledge of how EACH part in it worked.
>>
>>Simply substituting parts is NOT necessarily a good way to go, no
>>matter what someone might like about the result. Without the
>>detailed knowledge about WHY a designer did it that way, how
>>could one possibly know WHY a part should or should not be
>>substituted at random.
>
> I don't really understand why you are confusing the issue with, what I
> consider, extreme examples of a hypothetical problem.

A "hypothetical" example? Sir, I gave two examples that were used
in production runs of at least three models from two manufacturers
whose total product runs were in the tens of thousands of units.

Hardly "hypothetical".

> Most speakers have used Mylar caps in their crossovers (where
> feasible) for decades. While it is true that polypropylene
> caps have somewhat lower ESR than do Mylars, it's nowhere
> near a big enough delta to make any real difference in this
> application.

I left my entire post intact. Where in my post did I ever
say anything about mylar caps? (hint: nowhere).

> Now, I could see it if we were swapping polypropylenes for electrolytics or
> tantalums. but not for Mylars.

Now, read my post again, especially the part that says:

"a popular 'tweak' of the old KEF RS104-AB network was
to rip out all the electrolytic caps and replace them
with polypropylenes and, yes, it DID change the sound,
but NOT necessarily for the better, because the ENTIRE
crossover was designed with knowledge of how EACH part
in it worked."

see if you can show us where I did not at all mention a
difference between electrolytics and various film capacitors.
(hint: it's there)

> Now if you know differently, I'd love to see some data that
> shows where the crossover characteristics of a speaker system
> have been changed significantly by swapping polypropylene caps
> for the same value of Mylar caps, as to make a fundamental and
> profound enough change to audibly impact the speaker's
> performance.

Since I never once made that assertion, I don't understand
why it's up to me to supply data to support or refute it.

The point of my post was to privide at least two example of
why substitution WITHOUT the detailed knowledge behind why
a particular component was used DOES NOT necessarily lead
to "better." Different, yes, but different is not necessarily
better.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+

"Audio Empire" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)

> I don't really understand why you are confusing the issue
> with, what I consider, extreme examples of a hypothetical
> problem.

Like like a wind-up to a dismissal of an opinon from an expert witness.


> Most speakers have used Mylar caps in their
> crossovers (where feasible) for decades.

Right, where feasible, which means that there are tons of speakers still
being made with electrolytics in their crossovers.

> While it is true
> that polypropylene caps have somewhat lower ESR than do
> Mylars, it's nowhere near a big enough delta to make any
> real difference in this application. Now, I could see it
> if we were swapping polypropylenes for electrolytics or
> tantalums. but not for Mylars.
>
> Now if you know differently, I'd love to see some data
> that shows where the crossover characteristics of a
> speaker system have been changed significantly by
> swapping polypropylene caps for the same value of Mylar
> caps, as to make a fundamental and profound enough change
> to audibly impact the speaker's performance.