Strange behavior of an op amp in non-inverting amplifier configuration

[Scintillo]

I setup an op amp (TL081CP) in non-inverting amplifier configuration.

My pin voltages:
VCC- = 0 V
VCC+ = 15 V
IN+ = Varied between 0 V and 15 V with a potentiometer. (X-axis)
IN- = Connected to 0 V and OUT with 1.5 kOhm resistors.
OUT = Measured (Y-axis)

Here is what my circuit roughly looks like:


Here is a picture where I vary IN+ voltage on X-axis and OUT is on Y-axis:


The circuit works correctly for the most part but when IN- is near 0 V, OUT jumps to near 15 V. Is this typical behavior for op amps or am I doing something incorrectly?

[bktemp]

TL081 has a input common mode voltage range of -12V to +15V for a +/-15V supply.
So for your single supply configuration, it will not work below 3V.
If you go outside the common mode voltage range "phase inversion" occurs. That is exactly what you are seeing.

If you need an input voltage near 0V, LM358 operates from 0V to Vcc-1.5V.

[Kalvin]

Here is a nice story and possible cure "The phase reversal story":

http://www.eetimes.com/document.asp?doc_id=1231129

[Scintillo]

Very interesting. Thanks for the explanation.

[Zero999]

Very interesting. Thanks for the explanation.

As mentioned above it's phase reversal, due to the common mode range being exceeded.

For a greater understanding of why this happens, look at the internal schematic of the TL081. Refer to page 14 of the data sheet. Notice the p-channel JFETs on the input? What would happen if the gate junctions become forward biased?
http://www.ti.com/lit/ds/symlink/tl082.pdf

[Audioguru]

I have this document that I show when people use a TL08x or TL07x opamp with the input voltage going below its allowed input common mode voltage:

[comeau]

So for a single supply opamp in a non-inverting configuration, shouldn't you tie the inverting input to the half-rail vice ground? Am I completely misunderstanding something?

[danadak]

Single supply OpAmp biasing -


http://www.analog.com/media/en/technical-documentation/application-notes/AN-581.pdf


http://masteringelectronicsdesign.com/summing-amplifier-calculator-java/


http://www.daycounter.com/Calculators/Op-Amp/Op-Amp-Voltage-Calculator.phtml


http://www.falstad.com/mathphysics.html



As an aside, not all OpAmps exhibit phase reversal when inputs go outside CM range. And
to complicate things not all datasheets state the OpAmp behavior when exceeding CM range.
Consult manufacturer or their FAE for details.


Regards, Dana.

[Zero999]

Replace the TL081 with an amplifier that has a common mode range which includes the negative rail. The old LM358 will work but its output voltage will saturate at around 1.5V to 3V under the supply voltage. I'd recommend an op-amp with rail-to-rail inputs and output, such as the TS912.

http://noel.feld.cvut.cz/hw/st/2325.pdf

[mrkev]

The question is if you need that part of the range or if it's just an educational experiment that has no real application behind it.

If it actually is for something, keep in mind that you have the simmilar problem with maximum voltage. Maximum output voltage from the op amp is 10 - 12 V and half of that is 5 - 6V. That means that your circuit works only from about 3V to 5V (!) setting on the pot.

[newbrain]

Here is a nice story and possible cure "The phase reversal story":

http://www.eetimes.com/document.asp?doc_id=1231129

Simple and educational.
Some OpAmps actually have the diodes on chip, e.g. the venerable NE5532.

This caught me when I tried to use one as a comparator.