I am trying to create a 100uA constant current source to power a homemade sensor. I bought a few REF200 ICs http://www.ti.com/lit/ds/sbvs020/sbvs020.pdf since each has two 100uA current sources. Problem is I'm getting only 7uA for each of these sources, so my wiring must be wrong. How do I wire this correctly? Are there better options to create a ~100uA current source?

Here's what I did: pin(8) to +5V (power supply) pin(1) to the sensor and from the sensor to my power supply GND Added a multimeter in series to measure the current.

  • 1
    \$\begingroup\$ Did you ground "substrate"? Is the sensor too high impedance for that current at 5V? Does it work properly when driving e.g. a potentiometer? \$\endgroup\$ – pjc50 Jul 10 '13 at 12:36
  • \$\begingroup\$ I did not ground the substrate. Do I need to do this? The resistance of the sensor is ~5k ohm, that's why I believe +5V should be plenty! I also tried using +15V, with a different (smaller) resistance and I still get the same output current. \$\endgroup\$ – Michael Santiago Jul 10 '13 at 14:10
  • \$\begingroup\$ I wonder if the problem is that I need to use negative voltage at the low output pin. I read somewhere that some ICs need both negative and positive voltages. I'm just using a plain (+) power supply(?) \$\endgroup\$ – Michael Santiago Jul 10 '13 at 14:14
  • 1
    \$\begingroup\$ @MichaelSantiago Per the datasheet: "A substrate connection is provided (pin 6), which is isolated from all circuitry. This pin should be connected to a defined circuit potential to assure rated DC performance. The preferred connection is to the most negative constant potential in your system. ... For best AC performance, leave pin 6 open and leave unused sections unconnected." - Whatever is your ground or negative line from the supply will do fine. No negative voltage is needed. \$\endgroup\$ – Anindo Ghosh Jul 10 '13 at 16:30

The issue is likely to be insufficient voltage for the desired operation.

Based on the description, this is the schematic for the circuit being used:


simulate this circuit – Schematic created using CircuitLab

For the current source to function as designed with a 5 Volt supply, the maximum resistance across the terminals of the sensor would have to be under 25 kOhms.


  • If the sensor has a resistance of, say, 100 kOhms, then it would require V = I x R = 10 Volts across the sensor alone, to allow 100 uA to pass. Add a minimum of 2.5 Volts for operation of the REF200, and the minimum required supply voltage is 12.5 Volts. For stable operation, take 15 Volts as a safe margin.
  • If the sensor has a resistance of, say, 22 kOhms, then it will develop just 2.2 Volts across its terminals. This leaves 5 - 2.2 = 2.8 Volts across the REF200 current source, which is within its recommended operating range of 2.5 - 40 Volts.
  • In the schematic shown above, add a supply voltage via a battery, replace the sensor with a potentiometer, and vary the potentiometer to simulate this behavior for yourself.

Thus, depending on what the sensor's effective resistance is, at the desired operating point, the supply voltage for the current source to provide a stable 100 uA may need to be higher than the 5 Volts provided.

  • 1
    \$\begingroup\$ You beat me to it! \$\endgroup\$ – Andy aka Jul 10 '13 at 13:02
  • \$\begingroup\$ @Andyaka Oops :-) \$\endgroup\$ – Anindo Ghosh Jul 10 '13 at 13:03

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.