4

When the input voltage is smaller than 300 mV, the output voltage drops to almost zero.

When it is above 300mV, the output follows the input.

Why? If there is any discrepancy between input and output, I would expect it to be that the output is not able to reach the extremes (0 V or 5 V) of the supply, as it is not a rail to rail op amp. However this behaviour is totally unknown to me.

Circuit

Circuit

This is the measurement.

Measurement CH1 Vin, CH2 Vout

Raul M.
  • 310
  • 1
  • 9
  • 1
    All - *Please be nice* and make sure to comply with the site's [Code of Conduct](/help/conduct) in everything you write. Nothing snarky / sarcastic / condescending - or worse. Some comments were already flagged by other site members & deleted. [Polite constructive criticism](https://meta.stackexchange.com/questions/19756) is allowed - with the emphasis on *polite* & *constructive*. Don't answer the question in comments. If your comment includes the word "you" or "your", that's probably a sign that you should check your comment still follows the Code of Conduct. Thanks for your understanding. – SamGibson Aug 31 '21 at 05:39

2 Answers2

5

That op-amp has a minimum recommended supply voltage of 6V and an input range that typically gets to the negative rail +3V.

You are outside both of those constraints, so whatever it does is not clearly defined.

Probably you're preventing the bias circuitry from functioning and causing the output to shut down. Give it supplies such as +/-10V (or even +8/-4) and it should operate properly.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
  • Where the 6V is comming from? Apparently +5V and -5V in [page 13](https://www.ti.com/lit/ds/symlink/tl071h.pdf?ts=1630307198789) – Raul M. Aug 31 '21 at 19:30
  • @RaulM. [table 2](https://www.st.com/resource/en/datasheet/tl071.pdf) but the TI ‘H’ version says 4.5V is the minimum recommended, in 6.5, so OP may be okay in that regard. – Spehro Pefhany Sep 01 '21 at 00:20
5

You are violating the the "common-mode input voltage range", which defines the maximum and minimum acceptable voltages that may appear at the opamp's inputs, given its power supply potentials.

From the TL071 datasheet, on page 19, we find \$ V_{ICR} \$. It tells us that this range could be as constrained as ±11V, but is more typically +15V and -11V, when powered from ±15V.

You can infer that if you're really unlucky you can't take the inputs to within 4V of either supply potential, which is a real bummer if your supplies are 0V and 5V.

More likely, though are the typical values of +15V and -12V. These tell us that the manufacturer is confident that you can take your inputs all the way to the positive supply voltage, and to within 3V of the negative supply. Still rather restrictive, and this explains why the TL071 and its cousins are very ill suited for such low power supply voltages.

For inputs that lie outside of that range, the behaviour of the opamp is undefined. Most opamps' outputs just get stuck against one or the other of the supply rails.

Your particular device seems to be very forgiving. It allows you to get right down to within 300mV of the negative rail potential. If you want better performance than this, you must choose a device with correspondingly better input common mode voltage range, such as the LTC6258

Simon Fitch
  • 27,759
  • 2
  • 16
  • 87
  • On page 17, there is a more relevant spec of (Vcc-)+1.5V to (Vcc+), so for +5V supply it is 1.5V to 5V. – jpa Aug 31 '21 at 12:01
  • @jpa That's for the H version, which seems to be an extended range variant. On page 13 there's a section entitled "Recommended Operating Conditions: All Devices Except TL07xH" which supports what I said. There's nothing in the question which suggests that the asker is using an TL071 in his circuit - only that he's using that model in his schematic. If he is, it's unlikely to be the H variant. Also, I used that datasheet as an illustration only. – Simon Fitch Aug 31 '21 at 12:59
  • Oh, indeed. Looks like the plain TL071 isn't even rated for single-sided +5V operation. – jpa Aug 31 '21 at 13:21
  • 1
    Thanks. I must point out that the measurement is taken from a TL071H so limits are not as extreme as the TL071. I will update the title. – Raul M. Aug 31 '21 at 19:19
  • 1
    @jpa I was wrong, OP is using the H version! That made me laugh. – Simon Fitch Sep 01 '21 at 00:50
  • 1
    These letter suffixes can be tricky, sometimes the entire part behavior is different and the main type number is kept just because it is "well known". – jpa Sep 01 '21 at 04:06