So ive built up this circuit, as a simple current limiter. After making sure the output is null when the differential is 0, i get different errors depending on the load impedance even when the both loads have the same current flowing through them.

When i connect a 100 Ohm resistor to the output i get a 10mV error at 100mA so the Op Amp outputs 90mV. When i connect a 10 Ohm resistor to the output at 100mA the op amps output reads 80mV.

I have tried two different Opamps, the one on the schematic and a Lm741. I have tested them with both single and split supply but the error pattern still persists. What is happening?

enter image description here

  • \$\begingroup\$ "Op Amp outputs 90mA" - op-amps output voltage not current. \$\endgroup\$
    – Andy aka
    Jul 31, 2015 at 14:39
  • \$\begingroup\$ Apologies i meant the circuit. Arghh. It should be outputs 90mV and 80mV respectively. \$\endgroup\$
    – MAM
    Jul 31, 2015 at 14:41

1 Answer 1


Your differential amplifier is seeing a change in common-mode input of 9 volts (10 ohms to 100 ohms at 0.1A).

You are seeing a change at the output of only 10mV. A single 100K 1% resistor mismatched by 1% will cause a 90mV error. Do you see your design problem? You have almost no signal and a huge common mode voltage change- in fact it's much worse than if the output voltage was fixed.

You can go with 0.1% resistors or a network (or trim the CMRR with a trimpot) and stay with this topology, but really this is not the best way to measure current with a high-side shunt. It's too sensitive to component tolerances so it requires excessively precise and expensive parts for mediocre results. We should try to design circuits that use cheap and widely available parts to provide superior results.

In this case, it's better to use a transistor and amplifier to level-shift some current balancing the voltage across the shunt, which is really another question so I'll sin() off.

  • \$\begingroup\$ Thankyou! I do understand what is going on now which is a relief. Could you please point me to the better way of measuring current? \$\endgroup\$
    – MAM
    Jul 31, 2015 at 15:31
  • \$\begingroup\$ Not in this thread but - if you open another question related to current measurement with a wide common mode range (be specific with numbers- both signal and common mode range) then I or someone else can help you. There may already be answered questions that will help- ISTR at least one or two. \$\endgroup\$ Jul 31, 2015 at 15:56

Your Answer

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