Electronics > Projects, Designs, and Technical Stuff

EEVBlog #85 High Voltage Oscilloscope Probe Discussion

(1/1)

dannybeckett:
Hi guys, I have a few questions on this video:

1) What would the freq response of the probe look like if no caacitors were used in parallel with the resistors at all?
2) If that copper tube is grounded, surely precaution has to be taken to ensure the signal being probed does't arc to the tube? I want to build a probe capable of 20 to 25kV, any ideas on how to insulate the shield from the signal without making it huge?
3) In EEVblog #279 - How NOT To Blow Up Your Oscilloscope!, Dave talks about the importance of the ground clip, and how a differential probe allows you to overcome the issues described. I'd like to create an active, 25kV differential probe with a decent freq response. Im assuming that all I need to do is build a floating differential amp, then use two of the decribed passive HV probes to feed the inputs of the amp. Sound like a good idea?

I'm quite an exerienced electronics enigneer so please dont fret about playing with HV etc ;)

Dan

amspire:

--- Quote from: dannybeckett on October 18, 2019, 09:55:29 pm ---Hi guys, I have a few questions on this video:

1) What would the freq response of the probe look like if no caacitors were used in parallel with the resistors at all?

--- End quote ---
You would be luck to get a 10KHz bandwidth. I didn't look at that video, but if you are talking about 100M resistor in the probe, then don't expect to get to 1kHz.

--- Quote ---2) If that copper tube is grounded, surely precaution has to be taken to ensure the signal being probed does't arc to the tube? I want to build a probe capable of 20 to 25kV, any ideas on how to insulate the shield from the signal without making it huge?

--- End quote ---
25kV is pretty scary  - it will find out any weakness or oversight in your design.

Kapton tape it a pretty good insulator. I remember testing it years ago and we got 8kV per layer. You would never expect to get that reliably, but if you used a conservative design of about 500V per layer, then 50 layers of Kapton tape should be reasonably safe.

--- Quote ---3) In EEVblog #279 - How NOT To Blow Up Your Oscilloscope!, Dave talks about the importance of the ground clip, and how a differential probe allows you to overcome the issues described. I'd like to create an active, 25kV differential probe with a decent freq response. Im assuming that all I need to do is build a floating differential amp, then use two of the decribed passive HV probes to feed the inputs of the amp. Sound like a good idea?

--- End quote ---
Successfully making a 600V - 1kV differential probe is not that easy. If it was, you would see $20 probes everywhere on eBay. There are a handful of companies that do it very well.

25kV would be taking it to another level. Chances are very low to a successful result with decent bandwidth.

Richard.

jmw:
1) the capacitors are there to compensate for the scope's capacitance. If you took out all the capacitors in the design, the 50k probe resistor would be in parallel with the scope's internal 1 M resistance and 16 pF capacitance and this would distort the waveform. If you do the math, the capacitance is canceled at all frequencies when \$R_p C_p = R_s C_s\$. The shunt part of the probe that is in parallel with the scope (50k || 2nF) is way lower than the scope impedance (1M || 16pF), so the scope impedance can be safely ignored in favor of the probe's shunt impedance. The speaker picked the smallest cap available (10 pF x5 in series = 2pF), and the formula above requires 2 nF shunt capacitance to cancel it out.

2) in the video the speaker is talking about potting, so that means it the tube was filled with epoxy, which they go on to talk about introduced its own problems of adding stray capacitance and throwing off the frequency response.

3) sounds like a good start but there's also coupling between the differential amp's inputs and the amp's bandwidth that will affect the frequency response...

dannybeckett:
This was a perfect explanation, thank you.

Navigation

[0] Message Index

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod