[volt-nuts] Bohnenberger electrometer DANGER

[Andre]

Hi, re. capacitors it might be worth mentioning that the normal equation assumes charge and discharge through a constant current.
Don't forget that the equation includes a non linear term so you'll need to take that into account (Q=CV2 iirc) where Q is Coulombs, C is capacitance.
If this is done using something like an LM317T in CC mode or even a string of them (my idea) with anti-overload circuitry added externally then this
may well work.
Any series resistance will cause problems so you'd need quite a lot of regulators but there are ways to use JFETs selected by hand if you really wanted to
make a test setup.

________________________________________
From: volt-nuts <volt-nuts-bounces at febo.com> on behalf of Dr. David Kirkby <drkirkby at kirkbymicrowave.co.uk>
Sent: 07 March 2018 10:34
To: Discussion of precise voltage measurement
Subject: Re: [volt-nuts] Bohnenberger electrometer DANGER

On 7 March 2018 at 06:29, Andre <Andre at lanoe.net> wrote:

> Please be VERY VERY careful. To be honest its far safer to use CCFL
> drivers and rectify them with camera diodes in series and the absolute
> minimum capacitance for the job, shunted with a high value resistor.
>

The problem with 2.2 nF is it is difficult to know what voltage is on the
capacitor at any time, since with a 10 M ohm multimeter, the time constant
is 22 ms. I have a high voltage probe around somewhere, which probably has
a 100 M ohm input impedance, but that would still only give a time constant
of 220 ms, which is too short to measure easily.

I have 47 uF @ 650 V (or it might have been 550 V) capacitor on order, but
I might go to something a bit lower capacitance if the charge storage is
not required.

I would like to know where the energy comes from to move the leaf. I wonder
if any is taken from the capacitor. In the gold leaf electrometer, with no
internal supply, it is clear the energy much come from the charge on the
plates. But when there's an electric field, that might not be the case. I
was thinking of sticking a 50 uA FSD meter inside, to see if any current is
take from the capacitor, but I don't know if 50 uA would be sufficiently
sensitive to deflect.

Clearly having an electrometer here would be useful for these sorts of