Measuring high voltages using a scope probe. Safely

I have built an “oscilloscope emulator” front end for my QA401 that:

  1. Provides the AC line 3rd prong ground connection that all scopes provide
  2. Properly terminates oscilloscope probes at both AC & DC (one Meg + the usual few pF)
  3. Provides a DC block
  4. Provides a unity gain buffer after the DC block

With this box I can use any oscilloscope probe with my 401. However I have a set of 4KV scope probes this interface was built for.

I built it out of available parts and it surprised me by measuring pretty good on a loopback test. And it can be seen measuring THD of a signal that sits on 1650VDC. Works great ":^)


Very nice! Can you share more details of your design with us? Also some of your use cases?

1 Like

Use case? Every vacuum tube product ever! ":^)

…but in my instance I built this to measure one specific thing, a weird amp from an electrostatic speaker that is “all kilovolts all the time” under the hood. It’s the silver unit being measured in the photo.

Got it. Thx. Would you share your design with us?

My priority was safety before precision. The most important design aspect is the ground wire.

I used a FET opamp as it seemed like a good idea, the only one I had in stock was OPA2134. And the DC block is a 1uF film chosen simply because I had some. It’s not a calculated value. My input cap is adjustable - I wanted to calibrate this new box to my probes, not the other way around. And I used a nice beefy metal film for the 1Meg termination.

There’s one more thing I forgot to mention in there:

  1. There are back to back 3.3V Zeners at the output “just in case” to make sure my 401 never gets hurt. They see a 1K resistor from the OPA2134 output to limit fault current.

Thx @Chicago . Do you have a schematic?

Thx much Carl! I’ll be putting one of these together.

You can do better than I did. My termination is spot on at DC but slightly off at AC, I need to reverse that by nudging up the 1 Meg resistor.

Also there are 22uF dipped tantalums as rail bypasses in mine. I really did an available parts thing and plenty of improvement is possible.

… and the REAL most important part is the scope probe you use! I have the GE3425 from Cal Test and can recommend it. Not expensive.

I’ll post my loopback tests, like I said it surprised me by performing pretty well ":^)

Thx Carl. What are the specific changes you’d make? I was thinking of going with an opa1656 and a couple 12v batteries.

The OPA2134 worked out better than expected so I’ve no real complaints.

But you could do a discrete FET buffer instead, maybe with those new super quiet TI FETs? I don’t know the OPA16xx series well, sorry, can’t comment re: 2134.

The little box I found did not allow me to place a higher voltage DC block. Mine is 63V which isn’t awful, but this is a box were safety counts, I 'd like to have higher. And there’s no need for 1uF, that’s bigger than required.

My probes are 100:1 which means my kit should have an overall gain of exactly -40dB, you can see in my photos I’m more like -41.

Fine tune the 1M input R to calibrate that specific buffer channel & that specific scope probe to 100:1 overall ratio. Keep them together as a calibrated pair.

Final thought - be damn sure you know the voltage rating of your scope probes! Many 10:1 probes are only rated to 300 Volts. For vacuum tube audio stuff that is not enough.

Ok. I’ll do some experimenting. Thx for your help!