[Milsurplus] Re: [The WS No19] Diode Effect in Cap's.

[W7QHO at aol.com]

All,

Ran into something like this with a mica cap in my ATC (ART-13) some time 
ago.   The mica coupling cap, C112,   between the first and second multiplier 
stages was leaky and exhibited a DC reading of several hundred millivolts after 
the power was shut off.   Applying 250 volts across it for several hours 
eliminated the leakage and the battery effect, but It failed again several years 
later and finally had to be replaced.   (Would have done this the first time 
around but this component is some kind of buried down inside the multiplier 
chassis.)   The set had been stored for an extended time in a humid environment and 
I always suspected this was the cause of the phenomenon.   Have never seen 
this effect in any other cap, mica or otherwise.

Dennis D.   W7QHO
Glendale, CA

Original message dated 10/6/04 6:52:01 PM, gl4d21a at juno.com writes:
> 
> Gents:
> 
> There is a long history of old paper capacitors showing non-linearity.  I
> became familiar with it in the early '50s, working on TV sets.  One
> obvious characteristic was that these exhibited a DC voltage, even after
> being discharged multiple times.  The applications where high pulse
> voltage was applied seemed to be most likely to fail in this manner.  We
> always suspected that some triboelectric effect was taking place, but no
> careful scientific investigation was ever pursued.  Also, these were in a
> hot environment in a TV chassis, and unlikely to have absorbed
> significant moisture.  In any event, a capacitor in an equipment which
> had been turned off for several days and which still exhibited several
> hundred millivolts to a 20,000 ohm per volt multimeter was obviously in
> need of replacement.  I do not recall seeing the effect in mylar
> capacitors as much as the old (cheap) paper caps used in B & W TV sets of
> the 50s and 60s.
> 
> Hope this adds some useful information.
> 
> 73,
> George
> W5VPQ
>