I'm trying to build a high voltage capacitor leakage tester with current limiting feature, below circuit is the simplified and raw idea:


simulate this circuit – Schematic created using CircuitLab

Problem with that circuit is the current passing through the OA1 resistors, that's causing error, discharging the capacitor and essentially makes this circuit useless.

I can't buffer such high voltage and can't increase the resistor value too high because then I have to deal with noise...

  • How can I buffer the input voltages of OA1 differential amplifier?

Additional info:

The current is being measured in differential mode but I didn't include that in the schematic.

I want to test capacitors leakage and diodes (LED, zener), that's why I need to measure the voltage too.

Maximum current is 100 mA.

Input voltage is selectable from 10 V to 400 V in 12 steps.

  • 1
    \$\begingroup\$ Do you have to use OP07? I'm asking this because there are OpAmps and InAmps having input bias currents in pA range. OP07's input bias current rating is something around 1 to 4 nA. \$\endgroup\$ Jun 3 at 8:13
  • 1
    \$\begingroup\$ @RohatKılıç The input(s) voltage is high so the current is in nA range. but I can change the op amp if needed. thanks for the "heads up". \$\endgroup\$ Jun 3 at 8:15
  • 1
    \$\begingroup\$ Please add some explanation to your question: How will you measure current? (Through R6?) How is the OA2 circuit supposed to work? (It looks like it's just supposed to switch the test voltage on, but I think you have + and - switched.) Could you do a high-side switch and measure capacitor voltage with respect to ground? \$\endgroup\$
    – Theodore
    Jun 8 at 21:40
  • 1
    \$\begingroup\$ OA2 is reversed \$\endgroup\$ Jun 9 at 13:08
  • 1
    \$\begingroup\$ Read this differential probe design and take concepts from it: circuitcellar.com/research-design-hub/… At 400V you want series resistors to limit how bad a single failure in step down can be for safety. \$\endgroup\$
    – DKNguyen
    Jun 9 at 13:14

I suppose the point of OA1 is to measure the voltage?

You can use an instrumentation amplifier with very big resistors (100M perhaps) as a voltage divider.

But you will always have some sort of leakage if you measure something.

Note that OA1 can be at 400V in regards to the ground if M1 does not conduct, so you better be careful with your supplies. The instrument op-amp won't have this issue.

Added: you can reinject the leakage current into the cap to compensate for it, quickly drawn - probably incorrect - schematic to show the point.


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ I don't want to use a very big resistor (well unless I don't have any other choice), because of the possible noise problem. \$\endgroup\$ Jun 3 at 8:08
  • 2
    \$\begingroup\$ You can filter the noise out with circuitry and software (supposedly it goes to a MCU), you can also use high-end low noise, low bandwidth instrumentation op-amp. if you want minimum leakage, it means big impedance (resistor or other), thus you have to find the good trade-off \$\endgroup\$
    – Damien
    Jun 3 at 8:10
  • \$\begingroup\$ I've read articles about FET based voltage followers but I'm not sure how much input offset voltage they will have... \$\endgroup\$ Jun 3 at 8:13
  • \$\begingroup\$ If you really want something solid, you can use the schematic I've made with lower value resistor, measure the leakage current through a shunt and inject it back into the system with a current mirror of some sort. \$\endgroup\$
    – Damien
    Jun 3 at 8:16
  • \$\begingroup\$ @ElectronSurf have a look at the schematic \$\endgroup\$
    – Damien
    Jun 3 at 8:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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