# Measuring High Voltage DC while drawing low currents (nanoamps)

Discussion in 'General Electronics' started by Andrew, Aug 10, 2004.

1. ### AndrewGuest

I'm trying to measure DC voltages in the range of 100-1000V, however a
normal multimeter with a 10 megaohm impedance won't work because it
draws too much current (my device is very resistive and runs only
10-50 microamps). I have even tried a high voltage probe and it still
draws too much current. Does anyone know of a device (or how to make a
device) to measure high voltages (100-1000V) while drawing only
nanoamps of current?

I have also tried putting a 2.5 gigaohm resistor in series with the
multimeter and that doesn't give accurate voltage measurements.

Thank you in advance for your help and ideas.

Andrew

2. ### Brian WhatcottGuest

Hold onto that 2.5 Gohm resistor, and connect it in series with a
picoammeter. i = E/R = 1E3/2.5E9 = 4E-7 amps

Or, in plain English: this rig would read 400 picoamps. per volt

Brian W

3. ### Everett M. GreeneGuest

Isn't a resistor in the Gohm range the same as an open circuit?
I've measured non-zero current flowing through a supposedly
open mechanical switch.

Andrew may wish to measure the current through the Gohm resistor
and multimeter series combination and compute the voltage.

4. ### Brian WhatcottGuest

Well, your heart is in the right place:

the average DVM reads to 0.1 microampere,
so on a 1000 volt source it would read 0.4 microamp.
plus or minus 0.1

That's not great resolution!

Brian W

5. ### John LarkinGuest

Some of the old Fluke differential voltmeters used a null detector
working against a precise 1000 volt supply, essentially infinite input
impedance at null. They're a nuisance to operate (must be manually
nulled) but they're very precise and very cheap on ebay.

Or: charge a known value capacitor for a while, say 0.1 uF. Then
parallel it with another (discharged) cap of, say 100 uF. Both must be
good-quality film caps. Ignore the small spark when paralleled. Use a
high-impedance (fet opamp + dvm) circuit to measure the resulting
voltage, then multiply by 1001.

John

6. ### AaronGuest

I did as you suggested and the output is still a sinewave. I have the scope
set to DC Coupling 5 volts per division trigering on the upward slope at 5Ms
/div.

Is the scope bad?

7. ### Jerry GreenbergGuest

Trek Inc. Instruments

Electrostatic Volt meters. These will be accurate within a few
percent. This is for what you want. Be prepaired to lay out some cost
for any of these.

http://www.trekinc.com/products/esvm.asp

Jerry G.
========

(Andrew) wrote in message

8. ### Bob StephensGuest

You can try lifting the third prong on the scope's AC mains plug.

9. ### Glenn GundlachGuest

Since the scope is showing SOMETHING, it is clearly working enough to
show your sinewave. Bob Stephens suggests defeating the ground pin of
the AC cord. This can be unsafe for the unsuspecting though it can be
very useful. BTW, I carry a ground lifter in the pouch on top of my
Tek scope for this very thing.

I suggest checking the ground lead on the scope probe. They frequently
get broken and without the ground reference, your scope would display
what you describe.

gg

10. ### Robert BaerGuest

1) 1 1Gohm or a 10Gohm resistor is very definitely *NOT* an open
circuit. In certain cases, such values can act like a short circuit.
2) Perhaps the current you "measured" came from somewhere else; the
actual switch is OK. The body may be bad, or there may be alternate
paths that you did not recognize.
3) The lowest scale of any multimeter that i have seen is 33 microamps
full scale; with 50 microamps being "typical" of the decent quality
multimeters.
4) A 1Gohm resistor will load a 1000V supply at 1uA; a significant
percentage of the specified load (10-50uA).
5) Putting a 1K ohm resistor in series with the supply (and load),
will gove a voltage drop of 1mV at 1uA; within the capability of some
DVMs - and certainly not unduly reducing the available voltage for the
load. A 1Meg resistor would develop 1V at 1uA, which may be tolerable
(1% worst case for specified voltage range.

11. ### Robert BaerGuest

1) He mentioned a multimeter, not a DVM. A multimeter gives limited
2) The "average" 3 1/2 digit DVM can read down to 20nA full scale, as
long as one can tolerate the added 10Meg series resistance; the 200V
scale gives a 20uA full scale sensitivity (with an intolerable voltage
drop).
If a DVM is to be used, then use a shunt is recommended, so that the
IR drop is minimized to 200mV full scale.

12. ### Brian WhatcottGuest

Yes, he mentioned multimeter. Taking the Radio Shack digital
multimeter (Cat# 22-811) as an example, it can read to 0.1
microampere as I suggested. That gives very limited resolution with
a 2.5 Gohm series resistor, as I suggested.
What multimeter, or DVM did you have in mind?

13. ### Brian WhatcottGuest

You are thinking of a moving coil movement (called a D'Arsonval
meter to engineers of a certain vintage.) These days, multimeters are
digital displays, in general.

I take it you are thinking of a way of measuring the load current of
an HV supply. The question refers to measuring the terminal voltage
of a high impedance source, doesn't it?

Brian W.

14. ### Robert BaerGuest

The majority of DVMs have a 200mV scale and 10 meg input resistance.
Parallel with 10K shunt (which adds negligible error), and one has
(2E-1 volts)/(1E4 ohms) = (2E-5 amp) or 0.2 microamp FS sensitivity;
change the shunt as needed - just remember the error increases as the
shunt value increases (unless one compensates{ ie using 101K instead of
100K).

15. ### Robert BaerGuest

The term "multimeter" is generally reserved for analog (D'Arsonval)
meters, and the term "DVM" stands for "Digital Volt Meter".

16. ### Robert BaerGuest

In the special case where one knows the value of the load, then the
indirect method of measuring load current and calculating the supply
(load) voltage from that, is acceptable.

17. ### Bill Kaszeta / Photovoltaic ResourcesGuest

There is the indirect way. Use two voltmeters, and a reference power supply.
One volt meter measures the output of the adjustable power supply, the
other voltmeter measures the difference between the reference voltage
and the item being tested. The difference voltmeter can have a high value
resistor in series if you want. You adjust the reference power supply until
the difference voltmeter reads zero and draws zero current (thus not
loading the item being tested. Then use the second meter to read the
reference voltage. This is useful if the voltage to be monitored is not
changing. The procedure could be automated with operational amplifiers.

Bill Kaszeta
Photovoltaic Resources Int'l
Tempe Arizona USA

18. ### Brian WhatcottGuest

On Thu, 12 Aug 2004 18:49:12 GMT, Robert Baer
Hmmm....2E-5 amps is 20 microamps FSD.

Brian W

19. ### Brian WhatcottGuest

On Thu, 12 Aug 2004 18:52:58 GMT, Robert Baer
Hmmm...I would agree that perhaps 30 years ago, people thought of
multimeters as having moving coil movements.
Try a google search on "multimeter" now to get a feel for the current
understanding.

Brian W

20. ### Robert BaerGuest

Sorry; i goofed. Thanks for the correction.