To: mail11:;-at-msd26.ENET.dec-dot-com (-at-teslatech)
Subject: correction/ temperature
From: "I am the NRA." <pierson-at-msd26.ENET.dec-dot-com>
Date: Mon, 27 Nov 95 13:08:55 EST
>Received: from mail12.digital-dot-com (mail12.digital-dot-com [184.108.40.206]) by uucp-1.csn-dot-net (8.6.12/8.6.12) with SMTP id LAA09425 for <tesla-at-grendel.objinc-dot-com>; Mon, 27 Nov 1995 11:15:39 -0700
An ambiguity crept in.
> ----+ +----
Lessee wot i can come up with on calibrating this, ASSuming a sine wave:
D'arsonval (thats the descriptor for the usual "coil moves in fixed
magnetic field" type meter) fundamentally averages current flow thru
it. BUT in the nature of a sine wave, averaging it ends up low on
the "Effective" value. This is where the infamous sqrt 2 comes in.
(it can be derived, but its been decades since i've seen it done,
so i'll just take it as a given.)
What this means is that where a DC meter fed full wave rectified
AC will indicate the AVERAGE of the value (.5) the effective 
value is higher, the famous .707. Put another way, the peak is
1.414 the effective.
This is right-----^^^^^^^^^, "average"| from last post was wrong.
For a full wave rectified meter, the INDICATION must be multiplied
by 1.414 its value (1.0 volts means 1.414) to get the effective
value. (in usual purchased VOMs this is all buried in the calibration
of the ranges, so invisible to the user. BUT IT ASSumes a sine...)
Since Steve's proposed circuit bypasses half the current round the
meter, the meter will read half what it is supposed to. So 2*(sqrt 2)
times the indication is ALMOST there.
(I've run thru this step by each, if i'm wrong, sing out....)
A small (depending on the meter) effect is that the "bypass diode" will
leave 0.7V (coincidence, no relation to the .707 above) _backwards_
across the meter. The meter will average this in, and read "a smidge"
lower. Calculating the size of the "Smidge" needs knowing the
resistance of the meter & the normal currents, etc, etc.
Whats an Effective Value?
Same thing as an RMS value. 8)>>
It was found early in the transition from DC to AC that averaging did not get
it right. (and was obvious to the mathematically inclided, anyway...). The
"effective" value is that value of DC that is "effectively" the same as AC.
IF AC was triangular shape, the average WOULD be right. From the math of a
sine wave, this cranks out the 1.41/.707 conversion.
Temperature (new stuffe):
I see a note about thus and such caps dieing young, after running hot. Most
Electonic Components that run hot will run longer if cooled. Thermal transfer
is an area i have done some work in, the short version:
Simply putting a fan on a component will DRASTICALLY reduce its temperature. &
air is a dandy insulator (8)>>) so it does not get into trouble around HV. How
Much? Hard to say. ANY is better than "none". A muffin fan or two is cheap
insurance to protect caps, especially if bought cheap. (This can all be
calculated out, based on temp specs of the cap, etc. 200 linear feet per
minute can cut temp rises in half...)
Max voltage, I'll yield to expert commentary, except to say one small thing:
I doubt exceedingly that simply dividing USUAL dielectric withstand figures
into spark length is meaningful. My understanding is that lightning proceeds
in stepped leader strokes, with following strokes using preionized air, the net
result being bridging of gaps way over ~30KV/inch. My guess is that the same
thing happens with tesla coils. I THINK this is borne out by the observations
reported here recetnly of differences between "single shot" and continuous