Yes speakers are connected on both channels. Source is a battery operated MP3 player but did the same thing connected to preamp in my system. B+ measures as above. 550vdc with 500mv ac at about 8-10hz and then on that 120hz sine at 100mv. All the power caps in the system are brand new and good quality from digikey or mouser. I'm running Star GND. I have been doing a bunch of troubleshooting and kept notes. The first thing I did was disconnect the input leads and connected 3 inch leads with RCA right at the input board. Same results, same level of noise as back panel connection RCAs. Still going to test the shorted input and open and report back.
I'm assuming when trying that I should keep input shorted bc gain is going to go sky ~20db higher and will amplify anything on an open input??
That may work to stop the oscillation, but then the THD (Total Harmonic Distortion) will go up.
Yep, the gain of the amp will also go up if the NFB loop is broken, this may stop the oscillation, but then again, it may make the oscillation worse due to the increased overall gain, remember, in order for an oscillator to start oscillating at least two conditions need to be met, the overall gain needs to be greater than 1 and the total phase-shift needs to be 360 degrees.
You actually need Positive Feedback, not Negative Feedback, for oscillation to occur, Negative Feedback reduces THD, flattens the frequency response, and reduces the gain, Positive Feedback increases the gain of an amplifier greatly making the amp more unstable.
Negative Feedback becomes Positive Feedback when the overall phase-shift in the amp brings the feedback signal back into phase with an input signal.
The reason why some amps have a tendency to oscillate at a particular frequency is due to a phase-shift in the amp changing the negative feedback from negative to positive at that particular frequency, the output is then in-phase with the input.
I did a quick search on Google and asked "Is Positive Feedback Needed For Oscillation To Occur" this is the response I got from AI Overview:
Yes, positive feedback is essential for oscillation to occur; it is the key mechanism that allows a signal to be amplified and regenerated within a system, creating sustained oscillations.
Explanation:
Yes, positive feedback is essential for oscillation to occur; it is the key mechanism that allows a signal to be amplified and regenerated within a system, creating sustained oscillations.
Explanation:
- Positive feedback amplifies signals:
Unlike negative feedback which reduces a signal, positive feedback reinforces a signal, causing it to grow in amplitude.
- Loop gain condition:
For sustained oscillations to happen, the loop gain (amplification provided by the system including the positive feedback) needs to be greater than or equal to 1.
Don’t quote anything generated by AI at me thanks. I’m not interested in reading circular nonsense.. My point was that as he has disconnected the NFB network there isn’t any feedback at all, and therefore no oscillation either.
It's not circular nonsense, it is verifiable scientific fact and something that all electronics engineers should know already, there's been plenty of papers filed about it from when the science of Electronics first existed up until the present day.
Basically correct, but one better may find related information in reliable literature instead of making an effort with that AI crudge.
Best regards!
Best regards!
Yes thanks for all the input guys. I do not intend to run w/o NFB but if disconnected and it quits does that mean that NFB is the source or just a symptom of something else? If the source I'm assuming that I can maybe lower the level of NFB and problem is solved. I have free time today and will be looking into alot of things. I will post more after that. Thanks for the ninterest.
The NFB is the source of it, but it is also a symptom of a problem. Its not one or the other it’s both. The only time the NFB is always wrong is if you had the entire polarity backwards. If the polarity is right and it causes high or low frequency oscillation then your frequency compensation is wrong. At least wrong for the amount of feedback applied. It is always fixable if you dig deep enough, and are willing to accept the performance that results from the “fix”.
If the global NFB was NOT the source of it, it would continue to oscillate with it disconnected and input shorted. Then you have another unintended loop somewhere that needs addressing.
If the global NFB was NOT the source of it, it would continue to oscillate with it disconnected and input shorted. Then you have another unintended loop somewhere that needs addressing.