0
\$\begingroup\$

In the following circuit, I'm having a hard time figuring out the reason for the output.

The op-amp is the OPA1662, powered at 3.3V. It is rail-to-rail and unity stable. When R?/Rth is not installed, the voltage on the positive input is 3.3V but the output is at 2.45V.

My understanding is that the op-amp will try to put out enough current at the output to keep the + and - at the same potential. The short-circuit output current is over 30mA and the output voltage seems to be independent of Ro (i.e., the output is the same even if Ro=100k or even open).

So why is the voltage not 3.3V (or at least near that)?

When the resistor R?/Rth is installed results are as expected, the output tracks the input.

I have run various Spice simulations and get the same results I'm getting on the physical circuit - changing the op-amp will get the output closer to 3.3V but I'm trying to understand what is going on here.

I considered the bias current but the output voltage seems to be immune to the value of R28.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ You have the correct answer from Kevin- you can fix this by using a RRIO op-amp or by adding a couple equal value resistors in the feedback path changing the gain to +2 and doubling R23, but I would suggest a different op-amp in any case. The OPA1662 is optimized for AC (audio) and you have a DC application. \$\endgroup\$ Commented Aug 11, 2021 at 0:34
  • \$\begingroup\$ @SpehroPefhany, thanks for the pointer; the only reason I was using this specific one is that I had a bunch of them in my box; I guess that's a good lesson to look more carefully when using existing parts! \$\endgroup\$
    – Yiannis
    Commented Aug 11, 2021 at 23:50

1 Answer 1

7
\$\begingroup\$

You are exceeding the opamp's input common-mode range.

Although the output of that particular opamp can get close to the positive supply rail the input has limitations and may not be able to get closer than 1 volt.

Opamps that guarantee the input will still function when at the positive supply rail are usually described as "Rail-to-rail input capability". Similarly, the output may have rail-to-rail ability (the OPA1662 does) but this does not apply to all opamps.

Not all opamps behave well when the input voltage range is exceeded, their gain can go negative or they may latch-up or even be damaged.

Most modern opamps are well-behaved under these conditions even though they do not function correctly and their operation returns to normal when the voltages are in the allowed region.

Here is the relevant section of the datasheet.

enter image description here

OPA1662 datasheet

\$\endgroup\$
1
  • \$\begingroup\$ Thank you - makes sense now. \$\endgroup\$
    – Yiannis
    Commented Aug 11, 2021 at 23:52

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.