The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: "Bill Sloman"
Subject: Re: generating calibration voltage
Date: Wed, 8 Jan 2003 00:46:23 +0100
Organization: Planet Internet
NNTP-Posting-Date: 7 Jan 2003 23:46:35 GMT
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
"Tom Torfs" wrote in message
> Hello All,
> I would like to generate a +/- 500uV (or something of that order of
> magnitude, exact value doesn't matter as long as it's stable)
> calibration voltage, 30Hz square wave referred to a common reference
> (not connected to ground). As input I have an LVCMOS 30Hz 0..3V
> (referenced to ground) square wave signal (I can provide the opposite
> phase as well if that would be useful). Supplies available are + and -
> 3V (from battery). The output will be looking into a very big input
> impedance (24 parallel 100 Gohm preamplifier inputs).
> The reference should see a large impedance (preferrably >= 10 Mohms)
> towards ground. I'm sure this must be feasible using some sort of
> opamp circuit. Maybe someone here has a nice suggestion for a general
> approach to this type of circuit from past experience?
Sounds like a job for a three-winding transformer. One winding is driven
with significant currents, and has nasty resistive voltage drops in the
windings. The second winding is grounded, and you monitor the voltage
generated across it with a high input impedance circuit to find out the
rate of change of flux being generated in the core. You can use direct
feedback from an op amp to keep this stable.
The third winding is your output and can be at any voltage you like. The
trick is to make sure that current through the primary doesn't "wind up"
positve or negative to the point where it saturates you core (or melts the
windings) because of small DC offsets in the monitoring circuit. This isn't
difficult if you think about it in advance.
With the right transformer, this can be very accurate indeed - to 1 part in
10^9 for a trifilar wound 1:1:1 transformer, or one part in 10^7 for rope
wound ratio transformer - which might be handy in getting from a stable
diriving voltage down to 500uV.
The version of this idea I worked up nearly thirty years ago with an RM6
core and a 748 op amp was as accurate as I could measure using a 6-decade
multimeter (after nulling out the 748's input offset voltage) and very
cheap, if slow (but faster than 30Hz).
You will probably neeed a soft-iron toroid and a fair number of turns for
the 30Hz transformer, but this is a well-established sort of technique, if
not much applied outside national standards laboratories.
See "Coaxial AC Bridges" by B P Kibble and G H Rayner, ISBN 0-85274-389-0
It's magic. You really only want pages 88 to 133 on transformers, but that
gives information that I've not been able to find anywhere else outside the
pages of the IEEE Transactions on Instrumentation and Measurement. The title
of the book is terrible misleading - by "coaxial" they mean that everything
is screened ...
Bill Sloman, Nijmegen
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup