RF amplifier positive feedback problem
If you have discoverd "strong positive feedback" you also should be able to find a way to change it to negative feedback.
Without any further information (crcuit diagram) no further advice is possible.
if you are designing an amplifier put a shunt resistor at the output, simulate and optimize resistor for K>1. read some material on amplifier stability.
if you are designing a oscillator put a tank circuit at the output to choose a frequency at which the oscillator works.
Is it necessary to comment on such "recommendation"?
Actually this "recommendation" is even stronger.
K is not good stability criteria but mu. See agilent ADS examples on stability.
K is neither a good criteria nor is it a bad criteria. It is a simple letter. What is the purpose of such a posting?
I mean if K>1 then it is not sure that amplifier meets stability. It also should meet determinant of S matrix less than 1 to have un conditional stability.
If mu>1 then no need to check any other parameter.
Hence I mentioned K is not good stability criteria and you may refer to
Application Note - AN 5441; LNA Design with the MGA-16516 Matched Pair, Low Noise Amplifier (708 KB)
Hi kspalla,
I could imagine (that means: I could guess) what you mean. But I have serious doubts that everybody knows what you mean by simply mentioning a letter like K.
To make it clear: I repaeat that K is not a criteria. Without any additional explanation it is simply a letter - nothing else.
Remember: The outer feedback factor of an amplifier with feedback is also called "k". And in many cases, the voltage divider factor also is called "k".
Thus, you can see that one should spend some explaining words rather than only mentioning simply a letter - and call it "criterion". OK?
sorry, What I describe is wrong. Actucally, the circuit has a negative feedback. When I simulate it with OCSTEST in ADS . The close loop gain is greater than 1, but it's not up to the condition of Nyquist criterion for oscillation. So if the close loop gain is much more than 1, the circuit performance will be affected or not. Thanks
david, are you really familiar with the stability criterion?
The Nyquist criterion is based on the loop gain - not on the closed-loop gain!
Normally, a closed-loop gain larger than 1 is the main purpose of each amplifier.
Another hint: Consider that negative feedback (for low frequencies) can turn into positive feedback for higher frequencies (due to phase shifts)
and, thus, cause oscillations.