Electronics > Beginners

Oscillating op-amp circuit

(1/3) > >>

Alex_Baker:
I am working on building a constant current source for a transistor curve tracer that I am building. This circuit works in concept, I tested it in LtSpice. The problem is that it wants to oscillate at about 1.6MHz.

Using what I have learned from school I know that I want to reduce the gain at high frequencies, which I can do by adding a capacitor in parallel with R5. This only gets me so far though, I can't seem to fully damp the oscillations with this method.

If there are some solutions to this problem, or if my methods here are fundamentally flawed, I would appreciate the feedback!

note: In the uploaded schematic the bypass caps are not grounded, this is only a schematic error.

temperance:
This will help.

https://www.eevblog.com/forum/blog/eevblog-1609-composite-amplifier-tutorial-practical-demo/

Don't forget that the transistor is also an amplifier and the bandwidth must be much larger than the bandwidth of the servo amp. The solution is compensate the servo amp.

Kleinstein:
The 2nd amplifier in loop with additional gain makes oscillation likely. A bit surprising that it did not oscillate in the simulation, but it is still boarderline and parasitic capacitance and details in the model can make the difference. Chances are one would want an additional capacitor as direct FB of the outer amplifier and a resistor in front of the other OP's signal to the feedback.

The transistor should be fast enough for the LM358. So no problem there.

Whales:
> constant current source for a transistor curve tracer

This circuit looks quite complex.  Is the transistor Q1 the DUT?  If not then where is the output?

If useful I can suggest simpler constant current source circuit using an LM358, a single resistor, a single pot and no need for a split-rail power supply.    But I might be completely misunderstanding what you want to build.

Alex_Baker:
Thanks for the input Wales.

In this circuit Q1 is not the DUT, it is only a pass transistor. R6 is a load resistor representing the base-emitter current path of an npn transistor, I know this is not a good approximation of a base-emitter junction, but I figured for a good constant current source it shouldn't matter.

I am not sure what circuit you are talking about but your point makes me realize that Q1 may not be necessary, I think the op-amp itself can source the small currents I am attempting to control.

Navigation

[0] Message Index

[#] Next page