[Test-Equipment] HV probe resistor selection

[David]

On Sun, 24 Feb 2013 16:39:43 -0800, Dave C <davec2468 at yahoo.com>
wrote:

>On 23 February 2013, at 9:03 AM, David wrote:
>
>> The 25 megohm 3.8 watt resistors are just over 1 inch long.
>
>Using multiple resistors (3x 25 Mohm) means solder connections between. At max voltage (45 KV across all 3) means the first joint will be 35 KV. I'm concerned about flashover. Not sure about the absolute risk here but just comparing the 1 resistor to 3 and the relative risks. I think I'm more willing to go with resistors that won't give me a clean divide ration, and adding a pot to calibrate.
>
>Dave

Don't you mean 30 kilovolts (45KV * 2/3) at the first junction?

I wonder how conservative Ohmite's 15 kilovolt rating is and if it
takes into account stray capacitive coupling.  I would derate their
voltage rating significantly when using their resistors in a high
voltage probe where a test point has a more uncertainty then a fixed
design.

Derating the maximum voltage will also have the effect of distributing
the voltage over more and longer resistance elements which will yield
better accuracy because of a lower voltage coefficient of resistance.
200 volts could have the same effect as 1 degree of temperature change
with a 100 ppm/C resistor.  Good accuracy will limit you to a fraction
of the rated voltage per element.

The Fluke 80K-6 design is only intended for use up to 6 kilovolts.  If
you are worried about safety of instruments and yourself, maybe you
should be using an 80K-40 which has a much longer outside surface.