3
\$\begingroup\$

I am trying to build a non-inverting amplifier using a LM7171 op amp, and I am implementing it with potentiometers as Rf and R2 so that I can smoothly control the gain of the amplifier (and easily check if its working correctly). I am applying +/- 15 V to the op amp. When I adjust the potentiometer's resistance the gain does not change as expected, or noticeably at all. Setup is shown below with traces when the resistances are Rf=Ri=540 Ohm and Rf=1.8 kOhm, Ri=540 Ohm, and as can be seen the gain/amplitude is the same for the different resistances used. The yellow output on the oscilloscope is the output of the amplifier, and the green is directly from a t-connector to the function generator's output.

I have used two different op amps and got the same behavior, so it is not broken, and all connections seem good. I measure the values of the potentiometer so am confident they are as described. Any ideas what might cause this? Thank you in advance for your help.

enter image description here

enter image description here

enter image description here

enter image description here

enter image description here

100 kHz sinusoids. enter image description here

enter image description here

Setup as voltage follower

enter image description here

enter image description here

enter image description here

\$\endgroup\$
4
  • 2
    \$\begingroup\$ You also have a large phase shift. Something odd going on. You have no DC path on the non-inverting input. That's going to cause the input capacitor to charge up or down until it hits one of the supply rails. \$\endgroup\$
    – Transistor
    May 8, 2019 at 22:06
  • \$\begingroup\$ The LM7171 has a very, very fast slew rate and 200 MHz at unity gain. I think your output is showing a gain of about 1.5 in the setup. This looks more to me about the breadboard's pF, which is getting well into the 1k zone with that frequency. Use dead-bug construction and see what happens. \$\endgroup\$
    – jonk
    May 8, 2019 at 22:11
  • \$\begingroup\$ Jonk, when you say pF does that stand for parasitic feedback or are you simply referring to the capacitance (i.e. picoFarads)? You're saying that with such a high capacitance from the breadboard it gives a higher equivalent impedance in place of the pots' values changing the gain from what would be expected? \$\endgroup\$
    – wondersm
    May 9, 2019 at 16:01
  • 2
    \$\begingroup\$ @wondersm I mean the fact that there is about 5 pF between adjacent rows on those boards!! Those breadboards aren't mean to be used with those frequencies. (Well, I don't think so.) \$\endgroup\$
    – jonk
    May 9, 2019 at 20:43

2 Answers 2

Reset to default
5
\$\begingroup\$

The phase shift is the give-away. You are bandwidth limited. You need a faster op-amp, or a lower frequency.

\$\endgroup\$
8
  • \$\begingroup\$ Yup, 18.88 MHz. I hadn't spotted that. \$\endgroup\$
    – Transistor
    May 8, 2019 at 22:08
  • \$\begingroup\$ The op-amp has a 200 MHz unity-gain bandwidth so I was thinking it should be fine with only a gain of 2-3x around 20 MHz. I tested again at 100 kHz and the same behavior happens as well. I added two pictures to the original post to show this, and they are for the same resistance values as at 19 MHz; the waveform with the more pronounced artifacts has the higher Rf value, which actually yields a lower amplitude. I also added a DC path to ground as Transistor suggested, though that doesn't seem to make a difference. Does this suggest anything else? \$\endgroup\$
    – wondersm
    May 9, 2019 at 15:54
  • 1
    \$\begingroup\$ OK, based on your 100kHz test you probably aren't bandwidth limited. I don't see any decoupling caps. Put a small ceramic as close as you can to the op-amp. If that doesn't work, keep simplifying until it works as expected. Remove the pots and connect the output to the minus input for a gain of 1. Remove the input cap. Lower the frequency even more. \$\endgroup\$
    – Mattman944
    May 9, 2019 at 16:29
  • \$\begingroup\$ Mattman, I used bypass capacitors earlier but removed them as they seemed to have no impact. I've reimplemented them and converted it to a voltage follower as you suggested, picture above. Note I need the input capacitor, otherwise huge and strange oscillations are induced. In this configuration low frequencies are passed normally, but higher ones are not (pictures above), even though 1 MHz should be well within the bandwidth of the opamp. Would this imply that the problem lies with the frequency response of the breadboard itself? \$\endgroup\$
    – wondersm
    May 9, 2019 at 21:08
  • 1
    \$\begingroup\$ OK, I found another clue. "The LM7171 is stable for gains as low as +2 or −1". Most op-amps are stable at unity gain, but some high performance ones are not. Sorry, I lead you astray on that one. Try this test: Use 100kHz. Set your gain to 3 with fixed resistors. A resistor from the + terminal to ground as Transistor suggested is very important, try 100k. Always use decoupling caps, issues caused by the lack of them will get you when you least expect it. \$\endgroup\$
    – Mattman944
    May 9, 2019 at 22:16
4
\$\begingroup\$

You need to connect a high-value resistor between the op amp’s + terminal and ground. All op amps have small dc currents that flow in or out of their input terminals, and they need a dc path for these currents. If you measure the dc level at the output of the op amp, you will see that the output is not at the ground, even when there is no input. So, the op-amp “bias point” is incorrect. Try 100k or so.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.