All, I would like to share the following experience on the AGC settings of the K3. Situation at my QTH: There is an electric fence nearby my home that gives a pulse type of noise. The K3 handled that pretty well with the AGC PLS setting to ON. Switching the AGC PLS to OFF gave me a lot of trouble because the AGC was reacting to it. So the AGC PLS thing did its job. I was happy. PROBLEM: Recently I noticed that I could hear the pulses again very well. The K3 S-meter reacted (from S7 local noise peaking to S9+20 on the spike). AGC was immediately back to normal, probably because of the AGC PLS algorithm. Switching OFF the AGC PLS immediately made the AGC work also. Defenitely wrong thing to do.
When I compared this to the reception of my FT1000MP, no pulses were heard and nothing was seen at the MP's S-meter. So the pulsed noise certainly attacked the performance of the DSP... and succeeded. The only way to fight this, was to switch on the NB. DSP NB set to 1-1 did the job immediately. But still.... what changed during the past weeks that made the pulse noise into a trouble again? ANSWER: NO, it was NOT the new FW !! Rolling back to an older version did not cure the problem. BUT it was the setting of AGC HLD. I had changed that a while ago from zero to 20 or so. (I have had good results with hanging AGC in another TRX so why not give it a try). Checking this setting out, I saw a direct relation between the setting of AGC HOLD and the performance of the RX regarding the pulse noise. Setting the AGC HLD to ZERO solved the problem. EPILOG: We have been talking about the AGC behaviour a lot. So maybe this e-mail will help some of you. I'm learning more of the K3 ways of DSP every day . 73 Arie PA3A _______________________________________________ Elecraft mailing list Post to: [email protected] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com
