0
\$\begingroup\$

enter image description hereenter image description hereI have a constant current source that can output current in the range of 0mA to 1mA. I am driving 100 microamps through a load resistance (1.5Kohm) and then connected another test resistor (1.0 Kohm) in series with the load resistance. I am trying to monitor the voltage drop across this resistor so that any fluctuations in the current generation can be picked up. For this, I am connecting an oscilloscope across the test resistor and monitoring the voltage waveform.

However, inspite of using high impedance probes (10Kohm) with the oscilloscope whose channel input impedance is set to be at 1Mohm, I am detecting a fair amount of noise in the voltage waveform, which makes it fairly difficult to monitor the 100mV signal that I should be seeing.

Can someone please suggest an approach where I can decrease the noise that is creeping into the measurement? I would like to monitor the current waveform continuously to be able to make sure I am indeed getting the right rating from the current source. Is there a better way to monitor the current waveform which might be far less susceptible to noise?

Thank you.

P.S: At the moment, the entire circuit is on a breadboard and I will make a PCB soon and encase it in a copper lined box (faraday cage) to decrease any EM noise pickup.

\$\endgroup\$
  • \$\begingroup\$ What's driving the 100µA? \$\endgroup\$ – Samuel Feb 26 '15 at 23:45
  • 2
    \$\begingroup\$ Please add a sketch of your circuit. In particular, exactly how have you connected your scope to the circuit? \$\endgroup\$ – Dwayne Reid Feb 27 '15 at 0:00
  • \$\begingroup\$ What is the bandwidth of the scope and what is the bandwidth of the current variations you want to measure? \$\endgroup\$ – The Photon Feb 27 '15 at 1:11
  • \$\begingroup\$ Hello all, please find the sketch of the connections added. It is a very elementary circuit. Following the connections as showed, I created a math channel on the oscilloscope to calculate (A-B) i.e (Ch1 - Ch2) to obtain the final waveform (to decrease noise). The current injected is a single positive square pulse with a time period of 100ms and a duty cycle of 50%. \$\endgroup\$ – Aditya Kumar Kasinadhuni Feb 27 '15 at 16:48
  • \$\begingroup\$ My oscilloscope bandwidth is 200MHz. \$\endgroup\$ – Aditya Kumar Kasinadhuni Feb 27 '15 at 16:55
2
\$\begingroup\$

Take a look at the MAX4239...

http://www.maximintegrated.com/en/products/analog/amplifiers/MAX4239.html

It is an "auto-zero" type amplifier that is perfectly suited to measuring tiny voltage differentials (and hence tiny currents across a shunt resistor) very accurately with amazingly low noise.

For bench-top use, you can buy the uCurrent...

https://www.adafruit.com/product/882

...which is the above part made into a product. It has a MAX4239, battery, and shunt resistors all on a well designed board with nice connectors. It lets you easily see nanoamps on your scope.

I know StackExchange frowns on product recommendations, but I consider the uCurrent to be a tool class like an "oscilloscope" or "voltmeter" rather than a specific product. If you ever need to accurately measure 1mA or less, buy or build one.

\$\endgroup\$
  • \$\begingroup\$ The page at that Adafruit link for the uCurrent says that the product is "discontinued". Seems like you have to order it from Australia. \$\endgroup\$ – Pete Becker Feb 27 '15 at 2:49
  • \$\begingroup\$ Luckily, project is open hardware so you can build one yourself based on the schematics and BOM. You can also build a much simplified version with fewer precision resistors and amps if you only need uA resolution. \$\endgroup\$ – bigjosh Feb 27 '15 at 14:26

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.