Duh! Moment of the day

[mendip_discovery]

Today I was calibrating one of those wand-like High Voltage probes[1] and I used the 34401A as the measurement device for the output of the probe.

For a while, I thought this probe was rubbish 20 mV out at just 200 V[2] then when I tried 800 V and it gave me a not too dissimilar reading to 600 V I thought, I know what I might have done.

10 MΩ vs the 1 GΩ on the input will mess with the maths of the voltage divider in the HV probe.


[1] https://www.farnell.com/datasheets/1992834.pdf
[2] The customer only uses it in low ranges, and I can't supply more than 1kV anyway.

[bdunham7]

[2] The customer only uses it in low ranges, and I can't supply more than 1kV anyway.

I always manual range to 100V when using these, otherwise there are multiple issues that can get you.  IDK how you can calibrate it with 1kV since voltage nonlinearity is the big issue with these.  And if your customer only uses it at low ranges, I hope his DMM doesn't have 11M input impedance in those ranges.... 

[mendip_discovery]

Yeah its one of those off things, I think they just prefer the wand like probe for getting near to voltages.

I have it on my list of things to make is a HV source and a voltage divider up to 10 kV as more than that starts to get into the dangerous stuff that I really don't want to play with. I think I might keep an eye out for an old Flash Tester I could use for the source.

[Kleinstein]

With a 1:1000 divider and 1 Gohm for the input the output resistance for a simple divider would be around 1 Mohm. This would imply some 1.1 M internal to be supported by 10 M from the DMM.
So I don't see why it should make that much difference if the meter is 10 M, 11 M or possibly > 1 Gohm.  The high Z meter would give some 10% too high, but not off by that much and especially not too low.

[guenthert]

If you don't have an electrometer with sufficient accuracy, you could use the 34401A against (series opposition) a well known second source of sufficient stability and voltage (potentiometer method).  The 34401A (LO terminal) is good for potentials up to 500V above ground.  Just be careful 
That allows you to use a low range with high impedance (and good resolution).

But does the high voltage probe really expect a high impedance meter or does it specify a (terminating) input impedance?

[Kleinstein]

The linked probe specs show that the probe is made for 10 M meter impedance. So most modern meters should work, even with a high Z meter the result should only be some 10% high.