I am fairly new to electronics and I am encountering some issues in using the INA122 instrumental amplifier. I am using it as in Figure 3 in the spec sheet for the device https://www.ti.com/product/INA122

enter image description here

Now, I am seeking to measure the difference between Vin+ and Vin-, which will have slightly differing values around 13V. Originally, I was powering the inamp with a single supply at ~13V, but found that the output voltage was ~3V even when the difference in the outputs was 0V. After some digging, I gathered that my input voltages were higher than the allowed "Common Mode Voltage Range", and so I was saturating the outputs. In the spec sheet, it only lists this range (0-3.4V) for when the inamp is powered with 5V.

I tried using 19V to power the inamp, but the output is still 0.6V when there is no difference. My question is, am I still outside the common mode range? Any ideas how much voltage is needed to power the inamp that will allow my setup to work? Or am I missing something entirely?

  • \$\begingroup\$ Never used INA122. If I see the lower left picture page 4 of datasheet, common mode is -15V to + ~10 V when Vs=+/-15V. But in first page "COMMON-MODE RANGE TO (V–)–0.1V" ? Seems to be an internally shifting by 0.5 V to allow unipolar supply. Must simulate if I found its model. \$\endgroup\$
    – Antonio51
    Commented Jul 12, 2021 at 14:06

1 Answer 1


"am I still outside the common mode range?"

Simulation show that if INA122 is supplied with +/- 15V, common mode (in the circuit tested) goes from -12V to +11V with 0.1 V differential input. So it is ok. Two others configurations added lower.

Don't forget that for this amp, a return path is needed for the inputs, see datasheet, page 8 :

... "Without a bias current path, the inputs will float to a potential which exceeds the common-mode range of the INA122 and the input amplifiers will saturate. ! ... "

Provide at least two resistors 1 or 10 MegOhm ... to ground. This input bias current is approximately –10nA (current flows out of the input terminals).

Do you see the function of the two 10 MegOhm resistors ?

Simulation made with microcap12 from http://www.spectrum-soft.com/index.shtm

enter image description here enter image description here enter image description here

  • \$\begingroup\$ I'm not sure that I follow... I only have a single power supply at ~13V (so V- = 0V), which I can step up to ~19V. My differential input "must" be able to measure ~12-13V. When you mention return path, what are these resistors connected to and where? I don't understand why they are necessary. \$\endgroup\$ Commented Jul 12, 2021 at 15:00
  • \$\begingroup\$ Not a problem, I think. Unipolar power supply can be higher. Add only the two resistors to ground, and I think all will be ok. I try simulating this configuration. \$\endgroup\$
    – Antonio51
    Commented Jul 12, 2021 at 15:03
  • \$\begingroup\$ As in, put a high resistance resistor between each input (the two of which I am measuring the differential voltage) and ground? Why will this help? \$\endgroup\$ Commented Jul 12, 2021 at 15:09
  • \$\begingroup\$ Inputs must be biaised. See the remark and new simulations in my updated answer. I will add another simulation without resistors. If model is ok, it will saturate. \$\endgroup\$
    – Antonio51
    Commented Jul 12, 2021 at 15:16
  • \$\begingroup\$ Thanks for your help. I tried to implement it but even with no differential input, I am still saturating at 600mV. I downloaded your software and tried my own simulations, and my setup "should" work - 19V supply should result in 13V being within the appropriate range. But maybe the range is still slightly too low in practice? \$\endgroup\$ Commented Jul 13, 2021 at 1:21

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.