[Elecraft] K2 Frequency Stability Data and AGC Action

Jack Smith jack.smith at cliftonlaboratories.com
Wed Jan 17 16:06:01 EST 2007


That's a good point. In the AM broadcast receiver world, the ideal AGC 
action, I believe, is to maintain the IF at a constant level into the 
detector, so (1) the detector does not overload if you are close to the 
transmitter; and (2) so that the demodulated audio is always the same 
level, regardless of the signal strength, for a given modulation waveform.

In that regard,  the classic AGC curves found in the Racal and Drake 
receivers, as well as others that I've looked at match the ideal action. 
No AGC action up to a threshold and then very aggressive AGC action 
above that threshold, so that 100 dB of signal level change produces 
essentially zero change in audio output.

Whether the behavior desired in the AM broadcast world should be carried 
over to amateur radio is an interesting question. The fact that most 
receivers are designed one way does not mean that is necessarily the 
best way. Of course, the reason most receivers are designed that way may 
be because it is the easiest from an engineering prospective, or it may 
be a case where it also provides the optimum user experience.



Jack K8ZOA

Sverre Holm wrote:
>> I've finished additional K2 receiver measurements and have posted them to
>>     
> my web site, 
>   
>> www.cliftonlaboratories.com.
>>     
>
> Very interesting results, in particular the AGC performance curve. 
>
> I wonder about the statement "In an ideal receiver, once the threshold point
> is passed, there will be little or no change in audio output with changes in
> signal level." The way people speak so highly of the AGC action of their K2s
> seems to contradict this. But then maybe this is called the ideal curve just
> because it is easy to comprehend and draw for engineers.
>
> I have worked enough with acoustics and psychoacoustics to have great
> respect for the ear and our hearing. It is a very subtle instrument that is
> not so easy to quantify in a simple curve. Maybe design of AGC
> characteristics could learn something from the psychoacoustics that is built
> into all modern sound coders, like mp3 and its descendants?
>
>
> 73
>
> Sverre
> LA3ZA
>
> Unofficial Guide to K2 Modifications
> http://www.qslnet.de/member/la3za/K2/mod.html
>
>   


More information about the Elecraft mailing list