2
\$\begingroup\$

I want to monitor the waveform of a high voltage (8kV) 3Hz to 1000Hz DC pulse form. It has been suggested that I use 1pf capacitor in series with a 1000pf capacitor, then put the scope across the 1000pf capacitor for a 1000 to 1 voltage divider.

I have a few questions:

  • Will a 3 Hz DC pulse with a 50% duty cycle, look true, (i.e. flat tops without droop) with a capacitive divider?
  • How will the 10X scope probe load such a circuit and alter the waveform and amplitude?I will alter values as needed.
  • Would I be better off using a resistive divider?

I want to keep the current low, under 0.5 uA.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Do you actually need to see the true shape of the waveform? Just detecting the edges will likely be far easier. \$\endgroup\$
    – pipe
    Jun 3 at 21:39
  • \$\begingroup\$ I wanted to see the duty cycle, and maybe down the road an arbitrary wave for generator would be used, maybe want to see that HV waveform.. \$\endgroup\$ Jun 3 at 22:20

1 Answer 1

3
\$\begingroup\$

Don't mess with Mickey Mouse solutions because you'll never know whether the divided waveform is a true representation of the actual high-voltage waveform. If the 1 pF capacitor breaks down, it will destroy your scope. Also, it could be dangerous.

Instead buy a high-voltage scope probe: reliable, accurate, well built.

EDIT:

OP added "build it into the unit".

In that case, the correct way to do this is with 21 resistor and 21 capacitors:

  1. Build a parallel set of 1 MOhm 1/2 W resistor and 1 nF 1 kV capacitor.
  2. Make 19 more of them
  3. Connect them in series in a string
  4. Place the whole string across the voltage to be sampled
  5. Place the monitor port across the last set, the one connected to ground.

That will give you a 20:1 divider that is faithful from DC to high frequencies.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
7
  • \$\begingroup\$ Actually planning on five- 5 pf in series to help with the breakdown problem. I would like this monitor point as a bnc connector on the unit to connect the scope to. I actually have a 100MΩ HV probe on it's way. but I had no plans to build it into the unit this was just to verify things. \$\endgroup\$ Jun 3 at 22:21
  • \$\begingroup\$ "build it into the unit". That's an important requirement. Edit your question and that crucial point. \$\endgroup\$ Jun 3 at 22:45
  • \$\begingroup\$ Sorry I had a mistype, I want the current under 0.5 ma. \$\endgroup\$ Jun 3 at 22:56
  • \$\begingroup\$ "I want the current under 0.5 ma". The circuit I cave you draws 400 uA at 8 kV. \$\endgroup\$ Jun 3 at 22:59
  • 2
    \$\begingroup\$ It should be noted that your monitor point will be loaded significantly by a standard 10x oscilloscope probe (10 MΩ//5~10 pF in parallel with your 1 MΩ//1 nF), and even worse by a high-impedance oscilloscope input (1 MΩ//20~30 pF in parallel with the 1 MΩ//1 nF). It may be a good idea to compensate for that with a different resistance for R36 (the capacitance won't be significantly affected though). Or use a 100x probe, with an input impedance more like 100 MΩ//5~10 pF. \$\endgroup\$
    – Hearth
    Jun 3 at 23:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.