1
\$\begingroup\$

I'm trying to design an inverting amplifier that has a gain of about ~30. It is required to be a two stage design with the first stage being none inverting and have an input impedance of >=500k. The input frequency would be ~160kHz to 200kHz. Below I've built a preliminary circuit that only has a gain of 20 for testing and ran into some oscillation issues. I've attached the scope capture of the oscillations and a schematic. Channel one is attached to Node1 of the attached schematic. The oscillations seem to occur with no input.

enter image description here

schematic

simulate this circuit – Schematic created using CircuitLab

Reducing the gain of the first stage (reducing R10) or increasing R6 stops the oscillations but then I can no longer meet my specification. I'm wondering what is causing this issue. I've read this post: Why is my op-amp oscillating? and am wondering if I need to go into the math of the control system and feedback stability. Any help is greatly appreciated.

UPDATED: As suggested by Spehro Pefhany I tried grounding the input. This actually stops the oscillations. Is it possible this is due to poor layout? Layout of the circuit. Highlighted trace is the input to the none inverting in the first stage:

enter image description here

I've also updated the schematic to show bypass capacitors.

\$\endgroup\$
4
  • \$\begingroup\$ What does your layout look like? And do you have any bypass capacitors on the power supply? And does the oscillation stop when you ground the input as opposed to leaving it open? You have a very high impedance node at the input (500K). \$\endgroup\$
    – Spehro Pefhany
    Jan 23, 2020 at 3:28
  • 1
    \$\begingroup\$ @SpehroPefhany I've updated the question! \$\endgroup\$
    – Mackenzie Goodwin
    Jan 23, 2020 at 3:42
  • \$\begingroup\$ To avoid the obvious positive feedback, show detailed photos of cables and define signal source impedance. It think it is a mistake to use this configuration instead of a balanced differential amplifier. R5 needs a filter cap. Signals needs to be balanced and low impedance to avoid reactive coupling \$\endgroup\$
    – Tony Stewart EE75
    Jan 23, 2020 at 4:28
  • \$\begingroup\$ @TonyStewartSunnyskyguyEE75 I'm going to first try and reduce the capacitance feedback. Could you point me to an example of a differential balanced amplifier for a single ended application like this? \$\endgroup\$
    – Mackenzie Goodwin
    Jan 23, 2020 at 5:27

1 Answer 1

Reset to default
1
\$\begingroup\$

Your input (when open) looks like 500K||6pF maybe (Edit: maybe a bit more with the ground (??) plane), and there's a gain of almost 50, so even 0.1pF of parasitic capacitance from the output to the non-inverting input could be enough to cause oscillation at the ~1MHz you are observing. It would take even less from the 2nd op-amp output.

So it's probably layout-related. It's difficult to get something like this to work on a solderless breadboard. A dead-bug construction with a ground plane is far more likely to succeed.

Edit: From your posted layout, the output pin 1 crosses over the non-inverting input under the chip. That may be the source of your problem. Of course if the amplifier is normally connected to something that's relatively low impedance it might not be an issue. Your input impedance is maybe 100-200K at 200kHz I think (ignoring the inadvertent bootstrapping anyway). To get 500K at 200kHz is not easy without trickery that makes it happy to oscillate.

If the blue plane under the input is not AC grounded that's really bad. I don't see where it goes.

The fact that oscillation stops when the input is grounded tells us that it's getting in through the non-inverting input.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ The blue plane is a GND plane. I'll accept this as the answer until I am able to adjust the layout. Thank you! \$\endgroup\$
    – Mackenzie Goodwin
    Jan 23, 2020 at 5:19
  • 1
    \$\begingroup\$ You can try lifting the leg on pin 3 and cutting the trace from C4, then loop some thin wire 1/2" or so above the chip between C4 and the isolated pin 3. \$\endgroup\$
    – Spehro Pefhany
    Jan 23, 2020 at 5:21
  • 1
    \$\begingroup\$ I have tried a new design with an updated layout and it was actually a combination of two things. First being the issue Spehro suggested with capacitive feedback. The second issue was actually power supply noise on the 30V rail. This was causing noise on the input of the non inverting amp OA2. Thank you everyone for the help! \$\endgroup\$
    – Mackenzie Goodwin
    Jan 29, 2020 at 1:13

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.