[kai_yip]

After reading threads about Jan's amp, it is said that bad topology may drive LM6171 crazy.

What is the meaning of bad topology? Is it the components are too closed? Is it the soldering is too bad? Is it that some rules should be followed when deciding the topology? or anything else?

Thanks for your attention

 

[tangent]

I'm no expert, but as I understand it there are several aspects to good topology:

1. You have to be very careful with the wiring: use "star" grounding, don't run signal wires to close together, cross them at right angles instead of paralleling them, etc.

2. The circuit itself has to be correct: things that need to be in the opamp's feedback loop need to be in there, things that should be outside need to be. Deciding which is which is a matter of experience.

3. Use lots of decoupling caps between gain stages, and between the power supply and opamps/buffers. Also, the caps should be as close as possible to the parts they decouple.

By way of counterexample, a random layout on cheap protoboard with haphazard wiring and a minimal circuit (i.e. very few decoupling caps, etc.) will be very likely to cause high-speed opamps to oscillate.

Again, this is all hearsay. I don't speak from experience on this matter.