[sdiy] Hand-matching capacitors for filter stages

[Mike Pepper]

> As for various capacitor dielectrics, I think what you're most likely
> going to find is differences between the two main dielectric families
> - ceramics and plastic film.  The differences between polystyrene and
> polypropylene will be minimal compared to ceramic and poly[ester|
> carbonate|styrene|propylene].
>
> If anyone is bothered to check through the archives you'll find this
> subject thrashed to death many many times.
>
 I agree somewhat, though there seems to be some justification for 1%
polystyrenes in high end audio filters

To avoid a certain amount of archive trawling, the famous Cyril Bateman
articles are here:

http://www.scribd.com/doc/2610442/Capacitor-Sound

The famous, (and famously argumentative), high speed circuit expert Ivor
Catt, (and when it comes to discrete chip logic design, this guy actually
DID write the book), totally disagrees with the capacitor model used by
Bateman et al, making the point that capacitors have no self inductance, but
that the industry preferred to sell "low inductance" caps, rather than
advise users to crop the leads as short as possible. Another very good
reason to use reverse board surface mounted mounted decoupling capacitors; I
notice Jurgen Haible is a fan of this techniue, on all his recent board
designs. I first used a similar technique with leadless disc ceramics
mounted in slots cut in the PCB, for RF and some early ECL logic work.
Catt's book was almost the bible for this type of work, though only amongst
engineers - physicists have branded him one of the heretics. I never got the
deep theory, but do I know that if I followed his guidelines, my circuits
ran better. You can put most of what I built on a small FPGA nowadays, and
it will run faster!

A minor reminiscence: ECL was very hot running, on my first ever run with
the stuff I did a trick I used with TTL, of putting my hand flat on the chip
side of a wire wrapped card, to see if anything was overheating - the board
stuck to my hand, it was that hot: for days I could read the Motorola logos
lightly branded into my palm. I later found that was the norm - it always
ran hot, which is probably the reason Seymour Cray described himself as just
as good a refrigeration engineer as an electronic one.

||\/||ike