[Boatanchors] Disc caps vs paper caps

[Richard Knoppow]

     Ian did not say what kind of dielectric was in the 
caps.  Hi-K dielectric ceramic caps _do_ have rather high 
temperature coefficients and also voltage coefficients but 
Class-1 types do not and are very stable.  While silvered 
mica caps are considered the gold standard for stability 
ceramic caps with Class-1 dielectric probably equal them. 
The only drawback to Class-1 dielectric is its low 
dielectric constant but that is important only for 
sub-miniature parts and not for a cap which is to replace a 
paper cap.
     FWIW, while some types of ceramic dielectric are 
piezoelectric (maybe all, I have to research this) ALL 
capacitors are subject to mechanical deformation upon 
application of voltage and will work as capacitor 
loudspeakers and microphones, at least in theory.  I have 
been in the switching yard of a very large electrical 
substation when a group of power factor correcting 
capacitors were switched in; it sounded like a bomb going 
off. This is probably of absolutely no relevance to normal 
audio or radio frequency usage.


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com


----- Original Message ----- 
From: "J. Forster" <jfor at quikus.com>
To: "Ian Wilson" <ianmwilson73 at gmail.com>
Cc: "Boatanchors at mailman.qth.net" 
<boatanchors at mailman.qth.net>; "Mark Foltarz" 
<foltarz at rocketmail.com>; "Glen Zook" <gzook at yahoo.com>
Sent: Thursday, September 27, 2012 11:02 AM
Subject: Re: [Boatanchors] Disc caps vs paper caps


> Prove it.
>
> -John
>
> ================
>
>
>
>> I have noticed enormous amounts of drift in Colpitts-type 
>> crystal
>> oscillators using some
>> types of ceramic capacitors in the loop. This resulted in 
>> significant
>> chirp
>> on each key down.
>>
>> I don't know whether this is due to some form of 
>> self-heating, but the
>> capacitance changes
>> were so large that I can readily imagine that if you 
>> scaled the values up,
>> and used these
>> capacitors in an audio coupling network, that you should 
>> be able to detect
>> the nonlinear
>> change of time constant with input readily enough using 
>> both normal ears
>> or
>> some sort
>> of audio spectrum analyzer.
>>
>> 73, ian K3IMW
>>