Lashing up a HV Probe

[icon]

Hi

I want to check a 1700V supply on a scope I'm trying to fix. I have it in mind to make a chain of nine 10Meg (250V-rated) resistors. I could then rely on the 10Meg input impedance of the Fluke 87V to give me a 10:1 divider, or I could add another 10Meg resistor and tap off at that point to the meter and double(-ish) the figure shown on the meter (the 'bottom' of the chain going to ground, obviously)

Other than the obvious ("don't"), any thoughts either way?

Cheers
John

[Neilm]

Is this a DC signal or AC signal?

If it is AC then you may end up with some weird issues. Capacitances that are too small to worry about normally can turn up and cause all sorts of problems.

I would add the extra 10Meg resistor as it will mean that if the meter became disconnected accidentally, nothing would be damaged or become dangerously charged.

Yours

Neil

[SeanB]

Do not trust the quoted meter impedance. Use a 1M resistor and 10 series resistors instead, then you will only read roughlt 1/100 of the input voltage. The 9 series resistors must be cleaned after soldering with IPA and dried and baked to remove moisture, then encapsulated in high voltage silicone to prevent flashover.

[icon]

Hi

Thanks for the advice - I should have said, it's 1700V DC, though I might just check for 50Hz ripple on a scope. I will go for the 1/100 idea. Can metal film resistors withstand 100degC for a reasonable period of time without changing value significantly,  I wonder?

Cheers
John

[TriodeTiger]

Can metal film resistors withstand 100degC for a reasonable period of time without changing value significantly,  I wonder?

That's what ppm/c (or was it temperature coefficient) figures are there for!

Where are your 100C calculations coming from? Why does it need to be this hot? Seeing only 1.7kV on a 10meg 1/2W metal film resistor would be within its rating.

[icon]

Where are your 100C calculations coming from?

From the boiling point of water.

John