0
\$\begingroup\$

I am trying to design a DC circuit that allows a high input voltage of about 1 to 20kV and which can measure a low current output of about 1nA to 10mA.

This is my approach but I need feedback. I want to use the Thevenin's theorem. I want to set up a chain of resistors (Thevenin equivalent). The input voltage is the Thevenin voltage. Then I introduce a load resistor in series with the Thevenin resistance. The voltage across the resistor would then be measured. I am thinking of using Phidget's bridge to measure it. The output current is then gotten using the Ohm's law formula.

This is just my basic knowledge. Are there more efficient ways to create a circuit the allows high voltage supply and measure low output current? What are the real-life challenges to consider and how can I solve these challenges e.g clamping, isolation etc.

\$\endgroup\$
3
  • 4
    \$\begingroup\$ Are you doing highside or lowside sensing? Draw a diagram showing how you intend to connect your source, load and sensing parts. Phidget appears to be a tradename for modules. The only hit I get for phidget's bridge is an interface to Wheatstone Bridge type sensors such as load cells. \$\endgroup\$
    – Neil_UK
    Commented Feb 25, 2020 at 7:01
  • \$\begingroup\$ Thanks for your answer Niel. I'm trying to design a Circuit to measure the current in a high-voltage (1-20kV) ,low current (1nA-10mA) DC system. I think I need to use a couple of high voltage resistors and op-amp??? I am not sure about this. Quite confused. Any idea? \$\endgroup\$
    – John Bart
    Commented Feb 25, 2020 at 8:40
  • \$\begingroup\$ Are you happy to put the current sensing element between load and ground, or must it be at the high 20kV voltage? What does 'about' mean, in the phrase measure a low current output of about 1nA to 10mA. Given the large dynamic range, will +/- 10% accuracy do? \$\endgroup\$
    – Neil_UK
    Commented Feb 25, 2020 at 14:04

1 Answer 1

1
\$\begingroup\$

The 1nA to 10mA dynamic range, \$10^7\$, that you're asking for is extreme, this would not be possible with a single resistor, so switched resistors would be necessary. An alternative would be to use a silicon diode as the sense element, which would give you log(current), but you would only expect about 10% accuracy, even when using a similar diode run at (say) 100uA as temperature compensation. Maybe that accuracy is all you need?

Having a high side sense and using high voltage high value resistors to bring the differential reading down to ground is a total non-starter. The degree of matching needed to reject the common mode voltage would not be possible.

Two alternatives exist. One is to use low side sensing, interrupting the ground to the load with a sensing element, whether it's a number of switched resistors, or a diode. This may or may not be possible, especially if the load leaks current as corona into the air, not possible to intercept that going to ground. The other is to use high side sensing, in series with the power supply output.

The trick is to power the sensor, and to get the reading to you. For short runs, using batteries is simple and practical. For indefinite runs, a specially wound transformer should be able to get you 20kV isolation. You could use WiFi (arduino compatible ESP8266 for instance), or fibre-optics, or a free-space IR LED to a cheap remote control receiver, or even a current to frequency converter driving a loudspeaker with audio!

\$\endgroup\$

Your Answer

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

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