Voltage Follower Saturation

I'm trying to use a 0.5V signal as a source.

I've wired a TL072CN as follow:

1. VCC+ to 11V
2. VCC- to GND
3. OUT2 to IN2-
4. IN2+ to 0.5V

The thing is when I measure the voltage on OUT2 I get 10V where I am expecting 0.5V.

I made some test and for different values of IN2+ I get the values under:

• IN2+ = 0.5V : OUT2 = 10V
• IN2+ = 1.0V : OUT2 = 1.3V
• IN2+ = 3.0V : OUT2 = 3V
• IN2+ = 5.0V : OUT2 = 5V

Do you have any idea where this effect could come from?

Edit: the schematic (I didn't know you could do this)

simulate this circuit – Schematic created using CircuitLab

• Please take the effort of adding a circuit diagram. If you click edit to your question, there is a schematic editor button to easily include one. Mar 13, 2013 at 17:02
• Way too much hand waving. No, I'm not going to go thru your verbal descriptions and try to build the schematic in my mind or on paper. It's your job to present a circuit in a understandable way, which means to provide a neat and clear schematic. Mar 13, 2013 at 18:13
– Leo
Mar 14, 2013 at 12:43

From the datasheet you linked to:

ELECTRICAL CHARACTERISTICS VCC = ±15V, Tamb = +25°C (unless otherwise specified)

Vicm : Input Common Mode Voltage Range min +/-11V, typ +15V -12V

Which means that at +/-15V supply, it is guaranteed to work with the inputs are more than 4V away from either supply, and will usually work from V-(+3V) to V+.

Yours is already doing at least this well...

Now the good news : you can find a drop-in replacement if you search for "rail to rail opamps" which are designed to avoid this limitation. Depending on your application, you can either use one of these, or re-design to use the TL072 within its specifications.

• Do you see any other solution to get a 0.5V source than to redesign the whole circuit?
– Leo
Mar 14, 2013 at 13:40
• As I said in my answer, you can use the existing design if you substitute a rail to rail opamp for the current part. Mar 14, 2013 at 14:20

Your input is out of range of what the input differential pair is operational over. i.e. the front end is shutting down.

Unfortunately the datasheet only hints at this and doesn't explicitly state "Keep the inputs away form the rails by XXX & YYYY amount" The closest the datasheet comes is the input common mode range of +/- 11 volts on +/- 15V rails. Implying 4V off of each rail, but things are rarely symmetrical at the limits.

This is a common enough part, perhaps find a datasheet from another manufacturer that has that data.