I am trying to measure voltage across a 0.1R shunt resistor. When i measure voltage between input1 and input2 with multimeter i can see 10 mili volts. When i connect input1 to a voltage follower in the picture, op-amp starts to oscillate.

input1 and input2 are connected to voltage follower and after voltage follower i measure voltage across two outputs with differential amplifier.

I thought that this happens because of noise in input and added ~1hz lowpass filter to reduce noise but op-amp still oscillates.

I also added 1k resistor to output but no success. Op-amp i am using is TLC271CP which supports single supply operations.

Can you please help me how can i solve oscillation problem in voltage follower?

enter image description here


You have exceeded the input common mode range of the amplifier. Read the datasheet. With a 12V supply the inputs should both be between 0V and 10.5V.

That may not be the cause of your 'oscillation' but it most certainly will not work as drawn. If possible, put the shunt resistor on the low side of the load. You could also consider a rail-to-rail I/O amplifier but even that will not work for inputs right up to the power supply voltage- there has to be a bit of voltage for the amplifier to work.

You are also asking a lot of your differential amplifier. It also has to accept the voltage from the follower (which isn't really doing much for you) and the common mode rejection of the differential amplifier will add directly to the error. In other words, when the 12V changes the common mode voltage at the diff amp changes so that has an outsized effect on the amplification of the relatively small difference across the shunt resistor. Any small error in the resistors will cause a large output offset error which will not be very stable. If you have 1% resistors you might get 20% or 50% error, for example. This issue also goes away if you put the shunt on the low side.

  • 1
    \$\begingroup\$ Thank you so much for your reply. Putting shunt resistor to low side works for me. And even 1R load voltage across shunt resistor will not exceed input voltage range. I am trying it and i will share result with you. \$\endgroup\$ – yusuf şahin Apr 17 '18 at 12:30
  • \$\begingroup\$ Sounds good. It may help to think of (and draw) your shunt as having 4 terminals, two of which carry current and two of which are used for signal, with some small resistances between. \$\endgroup\$ – Spehro Pefhany Apr 17 '18 at 12:38
  • \$\begingroup\$ I put the shunt resistor to low side and now everything works fine. After unity gain i placed another op-amp with non-inverting 8.3 gain and i read 10mv in the input 84mv from the second op-amps output. \$\endgroup\$ – yusuf şahin Apr 18 '18 at 8:48

Even if you use a rail-to-rail input op amp, you will not get any information of the current passing trough R3 because input1 is always 12V.

You need to pick-up signal between R3 and laod, then you will have input1=12-I*R3.

  • \$\begingroup\$ You assume that voltage is 12.000V. Why? The voltage of input power is never that precise. \$\endgroup\$ – Chupacabras Apr 17 '18 at 12:25
  • \$\begingroup\$ You are right, when shunt resistor stays in high side it always sees supply voltage. Actually it was in low side when i first start to build this circuit and is completely my mistake putting resistor in high side while building circuit. @Chupacabras actually supply voltage is not important here because i want to get current usage with voltage across shunt resistor and if i get the voltage across it correct i can calculate loads current consumption correctly (resistors tolerance will also effect it) \$\endgroup\$ – yusuf şahin Apr 17 '18 at 12:38
  • \$\begingroup\$ The voltage of power rail is usually with ±2% tolerance. So it could be between 11.76V and 12.24V. Then your calculation is useless. \$\endgroup\$ – Chupacabras Apr 17 '18 at 12:40
  • \$\begingroup\$ lets say i see 0.02V across shunt resistor. Resistor is 0.1R and current will be 0.02/0.1 = 0.2A. When i put resistor in low side it stay in op-amps input range and it will not see more than 1volt with load. And tolerance in power rail doesn't effect result because op-amp's input will see voltage between 0V and 1V \$\endgroup\$ – yusuf şahin Apr 17 '18 at 12:48

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.