Re: HV Measurement - The Divider Problem (fwd)

[<sroys@xxxxxxxxxxxxxxxxx>]

---------- Forwarded message ----------
Date: Thu, 23 Sep 2004 07:35:18 +0930
From: Matthew Smith <matt@xxxxxxxxxx>
To: High Voltage list <hvlist@xxxxxxxxxx>
Subject: Re: HV Measurement - The Divider Problem (fwd)

Charles Pike wrote:
<blockquote>
According to the manual on my Wavetek 27XT the input impedance is 10
megohms ( This is typical of most DMMs).  Therefore when you connect the
meter across a 33 meg resistor its effective value becomes:  R =
(33*10)/(33+10) =  7.67 meg.  The division ratio is then 39.7 :1 instead
of 10 to one.  Sorry about the fried meter, I just cooked one myself.
</blockquote>

Aargh!  Yep, you've hit the nail on the head.  Thanks to Bert as well for 
spotting what I missed.

I've (foolishly) regarded meters of "having such a high input impedance I don't 
need to worry about it".  In future, I will!

The upshot of this is that:
a) My divider is fine.
b) I have a 37XR on the way which has the added advantage of an inductance range 
up to 40H.  I approximate transformer ratios using SQRT(Lp/Ls) and have found a 
lot of MOTs seemed to give me "OL" (out of range) for Ls with the 20H range of 
the 27XT.
c) I have proven conclusively that DMMs are fragile beasts and don't like 3.4kV 
put on their 1kV range!

The irony is that I actually built the divider to extend the range of my 4kV 
electrostatic meter.  (Made in the early 1940's and, by all indications, still 
just as good and accurate as it was sixty odd years ago.)  I guess that 
electrostatic devices have an input impedance as "near to infinity as makes no 
odds" (to quote the Hitchhiker's Guide).

Cheers

M


-- 
Matthew Smith
Kadina Business Consultancy
South Australia