I'm working on a three cell LIPO battery cell voltage monitor using three differential op-amps.

I had the circuit I designed below printed on a PCB and have assembled the board, but I am getting strange results on the output of op-amp 1 (Op1_output.) Op2_Output and Op3_Output work as expected.

When connected to the 3 cell LIPO (cell voltage for each cell = 3.8V,) Op1_output is reading 4.6V as opposed to the expected 3.8V. I have tested all the connections and also soldered 3 separate boards in case there was an issue with my connections, but the results are always the same.

I know that I can simply read the 3.8V directly because that cell is the first in the series, but I'm curious as to what my issue is, as creating the circuit in a simulator is yielding the expected results.

Any insight is greatly appreciated.

Op-amp datasheet


enter image description here

enter image description here

  • \$\begingroup\$ examine R10 in the schematic \$\endgroup\$
    – jsotola
    Jan 27, 2022 at 2:44
  • \$\begingroup\$ What's the minimum potential difference between the opamp rails? \$\endgroup\$
    – jonk
    Jan 27, 2022 at 3:15
  • \$\begingroup\$ @jsotola - R10 is connected to B3+, which the same as Vcc. Should be fine. It’s op1 that’s malfunctioning. \$\endgroup\$ Jan 27, 2022 at 4:42
  • 1
    \$\begingroup\$ Very odd opamp choice. 5532s are more at home in mixing desks and hi-fi amplifiers than battery monitoring circuits! Look at "Input common-mode voltage range" in the datasheet and note that you are operating out of spec. You may be looking for rail to rail opamps instead. \$\endgroup\$
    – user16324
    Jan 27, 2022 at 13:58
  • 1
    \$\begingroup\$ That’s a big topic. Try searching ‘selecting an op amp’ and see what comes up. In this particular case, you’d select one for single-ended and lower voltage with inputs that go to ground. So, an LM324 or a newer cousin. \$\endgroup\$ Jan 28, 2022 at 3:28

1 Answer 1


According to the 5532 datasheet you linked, the input range needs to be Vss+2V to Vdd-2V. Since you’re using ground as your negative supply, this limits the input range to no less than 2V.

With the R3/R4 voltage divider between OP1 and GND, this value isn’t being met on the (+) input of OP1: it’s only 1.9V above GND if BT1 is 3.8V, and as low as 1.5V at BT1 end of discharge (3.0V).

So OP1 is failing. As to exactly why, that is a result of this particular op-amp’s input stage design.

Try an op-amp that includes GND in its input range. The LM324 can do this. It’s popular and cheap.

So why doesn’t the sim fail? The Falstad simulator uses an ‘ideal’ op-amp which has no such input range limitations. However, if you try the Falstad 741 model it will fail; it’s similar to the 5532. There’s also an LM324 model you can try.

Or, you can get rid of the resistors R1, R2 and R4 and just wire the op-amp as a follower. Then the (+) input will be the same as the BT1 voltage, 3.0 to 3.8V or so, and (-) will follow it.

Still another option is to just measure BT1 directly. Since it’s referenced to GND there isn’t a need to use a differential amp. Save the one part, solve the problem. Win-win.

Not sure what your intention is with the 7809 driving an LED, but the overhead voltage for that regulator is 2-3V. That means it’ll be at only be outputting ~7V at the end-of-discharge voltage of 9V (3.0V per cell).

  • \$\begingroup\$ Ah of course, I figured it was something obvious that I was missing. That fixed it! Thank you. This is the first proper circuit I designed and I think I got overwhelmed by all the new processes and data. Thanks again! \$\endgroup\$ Jan 28, 2022 at 2:36

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.