Recommendations for 100x oscilloscope probe

[tomchr]

Folks,

I need to measure supply ripple on a 400+ V supply. I think that'll let the smoke out of my 10x o'scope probes...

What's your recommendation for a 100x probe? Are the Pomona probes any good?

100x, 1 kV rated would be minimum requirements. It would be nice if it supported the Tektronix probe gain detect (or whatever it's called) so the o'scope automatically multiplies the V/div by 100.

~Tom

[Svein_B]

I think typical Tek 10x probes are rated for 500 or 600V, so you might not see any smoke with your intended DUT.

If you have to use 100x probes, the AC signals will also be attenuated 100x so you may not be able to observe low magnitude ripple.

SveinB

[Elvee]

I think most "serious" X10 probes are rated for 600V total DC + peak.
If you want to be on the safe side, here is a successful design of a X100 probe:
HV Probe compensation

[AndrewT]

Elvee,
what are the voltage ratings of the caps used in your 100x probe?
How do we find/make 0.15pF?

[Elvee]

Originally Posted by AndrewT

Elvee,
what are the voltage ratings of the caps used in your 100x probe?
How do we find/make 0.15pF?

Yeah, I should have made those details clear.
You don't have to implement the 0.15pF; in fact they are unwanted, they represent the capacitances of the resistor string to the grounded shield and they have to be compensated.
The resistor string is made of 4x 22Meg/5% VR25 resistors selected so that their total is 90Meg (the individual value is unimportant, only the total):
http://www.electrodacus.com/store/D1090015/doc/spec.pdf
The main compensating capacitor C5 is made from two lengths of the center conductor of a coaxial cable rated at 4KV, held together with a heat-shrinked sleeve.
C7 and R6 are actual components, their value may need to be tweaked to get a perfect square wave response.
R5 is the input resistance of the oscilloscope, paralleled with a 10Meg resistor installed in the probe.
Its purpose is to fine tune the DC calibration of the probe, if you don't succeed in having precisely a 90Meg input, and to avoid the build up of high voltages if the output is left unconnected.
C4 is the total of the cable's capacitance and oscilloscope input, plus an adjustment capacitor to get a flat square response (C5 is unadjustable once assembled).
The probe is good up to 5KV.

[AndrewT]

Where does the 0volt reference clip on?
to the metal tube? at the probe end?

Presumably the metal tube has external insulation for operator safety?

[tomchr]

Originally Posted by Svein_B

I think typical Tek 10x probes are rated for 500 or 600V, so you might not see any smoke with your intended DUT.

If you have to use 100x probes, the AC signals will also be attenuated 100x so you may not be able to observe low magnitude ripple.

The Tek P6139 10x probes I have are marked 300 V RMS max. I'm expecting 10's of volts ripple, so it shouldn't be a big deal to resolve on the scope even after 100x attenuation.

~Tom

[jcx]

why not just a blocking Cap and gnd resistor?

or a 2X R divider can use pretty high R if you're looking with a 10x 'scope probe

[Elvee]

Originally Posted by AndrewT

Where does the 0volt reference clip on?
to the metal tube? at the probe end?

It doesn't really matter, unless you work with VHF signals, which is not the primary purpose of this probe.
In my case, I found more convenient to attach it at the coax cable entry, but you could solder it anywhere on the probe's body.

Presumably the metal tube has external insulation for operator safety

Mine is insulated, more for comfort and grip than safety: the metal connects to the chassis of the oscilloscope via the cable, and the oscilloscope itself is earthed via its power cable, it is thus safe, even without insulation.

[Mooly]

I suspect most 10/1 divider probes are OK on 400 volt DC including any ripple component.
I say that having spent 1000's of hours scoping PSU's, line output stages, flyback pulses etc with standard probes. Remember that is real hf power stuff and I have never had one go up in smoke

To "resolve the ripple" as you put it is best done by AC coupling the scope. You can see millivolts of ripple then even if sat on 400 volts.