In my application, I want to switch the polarity of an output which supplies a controlled current up to 50mA (accuracy 100uA + 1%) at up to 36V. The current sensor is on the low side of the output.

The control circuit I have only works for positive current, but I am also supposed to be able to output negative current, switching between the two at up to 50Hz. I already considered using a different output stage, but it seems to be much easier to achieve the required accuracy with a positive-only current control.

Before the idea of switching with 50Hz was introduced, I planned to use two relays for switching the polarity, like this:


simulate this circuit – Schematic created using CircuitLab

R1 is the shunt the current is measured over, while Q1 is used to control the current.

Obviously, using this to switch the output at 50Hz would create quite some noise, and would also put some wear on the relays.

Now finally here is my question: What alternatives do I have for switching the polarity of this output? I'm aiming for less than 10uA leakage under "typical" conditions, where leakage is the difference between the current through R1 and the current through the outputs. Low resistance would be nice, but isn't particularly important because this all happens inside the current control loop.

One option I am currently looking into is building an H-bridge with MOSFETs. There are fully integrated H-bridges, but they typically seem to use the same voltage supply for the logic as for the high/low side of the bridge, so it looks like I'd have to use discrete transistors.

Are there other, perhaps simpler alternatives?


You can use two form-C opto-MOS SSRs, such as the IXYS LCC110.

enter image description here

This one has only 35 ohms 'on' resistance, and switching time is better than a mechanical relay. Off-state leakage is at least ten times better than your requirement.

  • \$\begingroup\$ Thanks! I looked around a bit more in that direction, and found the Toshiba TLP222A-2F. It appears to be somewhat cheaper and has a much lower typical on-resistance. \$\endgroup\$ – Medo42 Apr 3 '14 at 14:00
  • \$\begingroup\$ It looks fine. The neat thing about the type I mentioned is that there is a "normally closed" MOSFET so the drive is a bit simpler. \$\endgroup\$ – Spehro Pefhany Apr 3 '14 at 14:08

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.