[ARC5] capacitor leakage

[Jim Mandaville]

Stew Beckley writes:

>Now for the question I haven't seen answered: What IS the acceptable
>leakage current per cap in those old triple-05's? I suppose there are two
>values worth quoting: a)The "spec" value for that particular cap (if
>anyone has the parchment on this), and b) an empirically derived value for
>proper and safe operation in a command Tx or Rx.

I thought I would be able to contribute something on this from the manual
for my Heathkit IT-28 capacitor checker.  But I can't find the manual for
this (and perhaps never had one).  By the way, when I restored this very
useful instrument (not  to be confused with some other Heath models with
less capability) I found that the various voltages applied to capacitors
for leakage checking were far out of specs due to resistors in the divider
chain changing values; had to replace several resistors.  Anyway, the
"magic eye" tube used in this instrument to show "unacceptable" leakage at
rated voltages has different bias settings for different capacitor types,
and the designer must have had some data on what was acceptable for each
type.  This might have been in the manual for the IT-28, so anyone with the
manual might have a look.  Or (if I were more ambitious) I could push a
paper cap to the "unacceptable" leakage point and check that with a
microammeter).

By the way, when I put some of my ARC-5 ancient caps on the bridge of the
IT-28 they generally showed MORE
capacitance than their labels, often twice as much (I know from new caps
that the IT-28  bridge is accurate).  Most of the caps showed " Heathkit
unacceptable"  leakage when I got up to about the 100 v level.

Jim