Extend the total voltage input on a oscilloscope / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Extend the total voltage input on a oscilloscope

previous 1 2

Move Thread

LAN_403

Sulaiman

Mon Apr 02 2007, 08:03PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I think you should rethink the resistor values so that the input resistance is as high as possible
e.g. for x100 use 99Mohm in series with the 1MOhm 'scope input resistance
Use a high-voltage 0.5pF capacitor in parallel with th 99MOhm resistor,
and a 50pF variable capacitor across the 'scope input.

Usually we want a 'scope probe to load the circuit being monitored as little as possible
Imagine the impedance of a 347pF capacitor at 10MHz (45.866 Ohm)
Use several low voltage resistors in series to make similar high-voltage high-resistance.
(e.g. I have 45 x 2.2MOhm in series for 99MOhm.)
This also allows several cheaper low-voltage capacitors to be used
(e.g. 45 x 10pF)
and produces a resistor long enough to ensure no flashover.

Fusion

Tue Apr 03 2007, 09:56AM

Registered Member #354 Joined: Thu Mar 30 2006, 10:12AM
Location:
Posts: 55

I simulated it and was good also (scaled down waveform rises and falls within 1ns with measured signal). But 1Mohm catch more noise from environment.

I have seen the scope probe specs and said 10Mohm and 9.5pF (the older was 1Mohm 4.5pF), but I tried measure it and is infinite, so I do not know exactly the impedance. Anyway is better measure 400V with a multimeter and compare.

Sulaiman

Tue Apr 03 2007, 02:28PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

My HP High-Voltage-Probe has a 'gain' of x1000 with 100 MOhm and 40 kV maximum input
roughly it has two stages, 100:1 and 10:1
giving a source impedance as seen by the 'scope of about 100 kOhm
and several adjustable potentiometers and trimer capacitors for gain and frequency response adjustment
I haven't got or traced the actual circuit so I'm not sure of the exact circuit.

Fusion

Tue Apr 03 2007, 04:06PM

Registered Member #354 Joined: Thu Mar 30 2006, 10:12AM
Location:
Posts: 55

I have modified the parallel capacitor at simulation from 0.096pF to 0.056 and signal felt to half!

So I think your probe must be too sensible to flat metal surfaces close to test point.

I think I have the solution: if you set a parallel capacitor with scope probe input of 100nF and then with 990Meg of 1nF and then I have a good response. Now the scope probe is less sensible to other capacitances:

Sulaiman

Tue Apr 03 2007, 10:14PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

My homemade probe is actually for dc, it IS a little sensitive to nearby objects

I use my Tektronics probe for hv-hf
The majority of the 100 MOhm resistor is shielded inside an earthed metal tube
only the first 1 inch or so is exposed.

P.S. the probe originally had CFC gas filling, I now use white spirit vapour - seems ok

Steve Conner

Sun Apr 08 2007, 10:34AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Sulaiman, what model of Tek HV probe do you have? Does it mind being filled with white spirit?

Sulaiman

Sun Apr 08 2007, 10:45AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Steve - It's a P6015 - very old.
I don't know if it 'minds', but I've had white spirit in there for years with no leakage.
It's not full, about 10cc of liquid only, the rest is vapor.
I store it at all angles and so far the seal seems good
and the voltage standoff is definitely better than just air.

For others - the old voltage probes used to have CFC gas/liquid for insulation/arc-suppression.

Steve Conner

Sun Apr 08 2007, 09:05PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Wow, the P6015 is a serious high voltage probe. 40kV and 75MHz, IIRC

Terry Fritz has one, and once he ran out of freon, he filled it completely with mineral oil and readjusted all the trimmers on the compensation box to get it working again. I'd like one, but they seem very expensive even for an old beat-up example.

PS: Terry said that his arced over at about 22kV when filled with air, IIRC. How much better do you get than that with white spirit vapour? Aren't you worried about air getting in it and making the vapour catch fire if it arcs?

If white spirit vapour is a good insulator, why aren't we using it in our HV projects? I have seen old textbooks recommend carbon tetrachloride vapour, probably because it's not flammable, but that's probably not too good for you...

Sulaiman

Sun Apr 08 2007, 10:36PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

My hv probe had been lying in it's box unused for years as I'd run out of the CFC and didn't buy any replacement,
I wanted to 'scope the output of an ignition coil I was playing with, so I dug out the old probe
it arc'd-over somewhere above 20 kV. Thought I'd ruined the probe but it was ok.
The only likely fluid that I had to hand was white spirit - so, hoping for no ill effects, I used it.
Worked ok up to the 30 kV region with no problems, so I left it like that.
(the other idea I had was butane lighter-refill but I thought it too flammable)

I was impatient so I'm sure there's still air in the probe - so far no explosions!

I may try butane one day (if you don't hear from me again you'll know what happened)
but I like the idea of completely filling with mineral oil - may try that also.

It is a lovely probe - good even at low voltages due to 100 MOhm//3pF input.

For low-frequencies I use a string of 45 x 2.2 MOhm metal-film resistors (VR37-type) with 1 MOhm load
(no special reason for the quantity/value - I just happen to have 1000 pieces form an eBay purchase)
For low frequencies and up to 10 kV I use 3 x 33 MOhm resistors (VR37-type)
Incidentally the VR37are commercially rated for 3500V, but only 250V for safety-critical - so my probe isn't safety-critical

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob