I am trying to test my inverting opamp circuit, as seen from the spice model, on a breadboard. Before I run any AC power through the circuit, meaning I only have the +-15V DC going through the opamp, I tested the signal coming from the output of the opamp after the feedback resistor on an oscilliscope, but I see a 4.6V pk-pk 27kHz signal, which is causing lots of additional "ringing" and noise when I do add in the AC power to the circuit. I do not know where this 27kHz signal is coming from or how to get rid of it.

 

Do you have decoupling caps on the supplies?

 

It's either poor decoupling of the power (100nF ceramic caps directly at both power pins to ground at the opamp power pins, or a bad layout.

 

Just as a test, connect the oscilloscope to the output through a 10k resistor. The capacitive loading of the scope, though small, will affect the circuit. The series resistor will largely isolate the capacitance. Again, this is just a test and in no way a solution even if it makes the instability go away. There are certain circumstances where low value resistors are legitimately used when capacitive loads must be driven, but your circuit is not such a circumstance. As others have said, lack of power supply decoupling is the likely culprit. Capacitance at the summing node (inverting input) can cause grief with high value feedback and source resistors with some amps.

A couple of photos of your breadboard would be helpful - one of the parts on the board and another that shows the wiring from the board to the power supply.

 

What value is V1?

 

Is there some reason you've chosen resistor values so high? They're WAYYYYY higher (like ≈100X higher) than what would seem appropriate, like a few tens of kilohms or so. With the values you've chosen, stray capacitances can strongly affect circuit operation and can even cause oscillation.

I would never design a circuit that way, and expect it to actually work correctly.

 

OBW, it is a JFET input amp so presumably the high resistances are to limit loading of a high impedance source, though they are asking for trouble if care isn't taken to minimize summing node capacitance. Feedback resistors of hundreds of megs aren't uncommon in FET op amp circuits.

I note that this is a rare op amp that has a pin for the can and substrate. And say what you like about amps in tin cans (usually rude things about Kovar-copper thermocouples), they do make it easy to put the summing node on the top of a teflon insulated terminal. Not what I'd call an inexpensive amp!

 

OBW, it is a JFET input amp so presumably the high resistances are to limit loading of a high impedance source, ...

That wasn't the issue I was thinking of; given the extremely low Ib of this part, resistors that large are certainly no problem with regard to DC errors.

...though they are asking for trouble if care isn't taken to minimize summing node capacitance.

Yes, that was the concern. Also, the capacitance between the R3-R4 junction and ground, and the R4-R5 junction and ground. Together with the op amp's input capacitance, I suspect beaucoup phase shift, which can invite oscillation.

On top of installing decoupling caps as suggested above, it may help to add a small compensation cap (10 pF or perhaps less) across the feedback resistor string.

Feedback resistors of hundreds of megs aren't uncommon in FET op amp circuits.

True. Resistors of even a gigohm or more are often used in charge amplifiers.

 

I do know that the resistor values are high, but this does not seem to be a big problem, and I have looked into taking the stray capacitances into account already. The oscillations exist even when I make a simpler feedback circuit with smaller resistors and different opamps, so the problem probably lies with the decoupling capacitors, which I have not had the chance to put in yet. I will do that the next time I am in the lab. I will reply again if there still seems to be issues. Thanks!

 

Also, the value of V1 is irrelevant because the oscillations appear before I even connect the AC power supply

 

Not all caps are equal in bypass applications -

Regards, Dana.

 

Also, the value of V1 is irrelevant because the oscillations appear before I even connect the AC power supply

The value of V1 is necessary, because otherwise, we have no idea what the voltage being applied at the non-inverting input is. You have a voltage divider between V1 and your OpAmp output. In order to tell you what the OpAmp is doing, you need to understand what the voltage is on the inputs so you can determine the relationship between the inputs.

You have a problem with oscillation. Contrary to what you may have been told, an OpAmp when coupled with a negative feedback loop generates and oscillating output.