To: mail11:;-at-cimcad.enet.dec-dot-com (-at-teslatech)
Subject: Cap testing
From: I am the NRA <pierson-at-cimcad.enet.dec-dot-com>
Date: Fri, 27 Jan 95 13:52:21 EST
>Received: from inet-gw-1.pa.dec-dot-com by csn-dot-net with SMTP id AA07216 (5.65c/IDA-1.4.4 for <tesla-at-grendel.objinc-dot-com>); Fri, 27 Jan 1995 12:01:06 -0700
(Argggh. I seem to have ASSed.)
I think i shall have a chance to recheck that reference.
>From: "SROYS" <SROYS-at-radiology.ab.umd.edu>
>If you were using current through the cap to drive X and voltage across
>the cap to drive Y, wouldn't a "perfect" capacitor (i.e. - no inductance)
>give a circular trace since current leads voltage by 90 degrees?
After staring at it, i think so. (I did check & think before posting,
but seem to have assed anyway... )
This makes the technique less useful, since it requires finicky set up
to get the circle "circular". Still some utility. Gain mismatches
would leade to squish, either vertical or horizontal. Internal
losses would lead to "leaning" ellipses.
>Stray inductance would change the phase and distort the circle into an oval
Squished oval, i think, so masquerading as a gain error: effectively
the inductance "cancels" some portion of the capacitance.
The other stray, and where i think this would be most uself, anyway, is
stray resistance, be it "Real" (lead) or "effective" (loss in
dielctric). A "good" cap (salvaged) could be told from a lossy
(bypass only) cap.
>Only at resonance (or with a pure resistance) would the trace be a straight
Good point. The display SHOULD indicate an internal-to-cap resonance,
if the applied voltage were swept.
I think that was what i waas trying to say about the wire size, tho i
may not have used the right (coil builders terminology). Its not
obvious but there IS/should be/better be more current in the tank
than in the supply leads...