1
$\begingroup$

In digital receivers, at the RF level (not audio), I often read about "slow AGC" and "fast AGC", and sometimes even "medium AGC" (but I think the latter is for audio). From my understanding, the slow AGC reacts to changes in noise level and variations in signal level over a relatively long period of time (fading for example). And from what I understand, the fast AGC is used more on symbol level, trying to keep them at a relatively stable power level, but not too fast so that amplitude-dependent symbols would not be affected (like M-QAM or M-PSK with M > 4).

Is there more to it? Also, would the "medium AGC" have any use in the RF chain?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
3
$\begingroup$

I don't believe there is much more to it than that, especially noting that you are aware not to track so fast that you track out amplitude information for modulated signals where the amplitude is used as part of the modulation (so more so for M-QAM than M-PSK). You can also design AGC's with different rise/fall times so can have "fast-attack" and "slow-decay" for example.

A reason you may want to have a dynamic AGC is distinguishing operation between acquisition and tracking. During acquisition it would be likely to use a much faster AGC at the expense of noise performance for the benefit of acquiring the signal faster. Then once acquired the loops (AGC, carrier tracking, symbol tracking) can run slower at the benefit of better noise performance but fast enough to track all expected dynamics.

Improve this answer
$\endgroup$
2
  • $\begingroup$ I'm trying to understand this dynamic AGC you mention. During acquisition, how is having a fast AGC beneficial? I guess it depends on the characteristics of the signal. The signal I'm working with starts with a CW preamble, then depending on the modulation used there may be a timing sequence, which I use to synchronize, then data. I can expect 10 dB difference between different transmitters, bursts last roughly 300 symbols. I use a power detector to detect the start and end of the burst. How would the slow and fast modes work in this situation considering acquisition and tracking? $\endgroup$
    – Signal
    Commented Jun 13, 2020 at 15:56
  • $\begingroup$ @TehWan The idea is to acquire as fast as possible at the expense of a lower bit error rate. Setting loop BW is a trade between SNR in the loop and ability to track fast changes (smaller bandwidth means more filtering but a slower result). $\endgroup$
    – Dan Boschen
    Commented Jun 13, 2020 at 18:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.