Are you guys sure this is a frequency domain problem? Has anyone measured those IMD products under the mentioned conditions and concluded from the results that they are the culprit for the mushy (somewhat vague description) audio that makes impossible to copy the different callsigns?
Some W6s measured in a lab test setup a similar effect on the FT1000MP years ago and found that the problem was the AGC recovering too fast (in between symbols) making all signals sound as loud as others. The solution was to increase the time constant. The loudest signal would then set the radio gain level. Then characters in-between from weaker signals sounded weaker making them distinct (or maybe buried). I cant think why you would ever need a recovery time constant any shorter than say 150 ms, while shorter time constants are normally used -reading Clifton I think the FAST setting in the K3 is 73 ms-. At 60 WPM, a dit is 20ms long if I am not mistaken, and an intersymbols spacing (7 time units) to full gain recovery seems the shortest time that would be necessary for the receiver to recover full gain. 73 de Juan EA5RS -----Mensaje original----- De: elecraft-boun...@mailman.qth.net [mailto:elecraft-boun...@mailman.qth.net] En nombre de Geoffrey Mackenzie-Kennedy Enviado el: sábado, 06 de febrero de 2010 11:56 Para: Guy Olinger K2AV CC: Elecraft Discussion List Asunto: Re: [Elecraft] K3 in a cw pileup - needs work Hi Guy, The problem reported has the classic characteristics of in-passband IMD caused by a small dynamic range (IMDDR3) IF system, rather than being something caused by the AGC sub-system. Yes changing the AGC's loop characteristics will alter the effect, but the root cause is still non-linearity in the signal path. I suspect the second mixer. Trouble is that large IMDDR3 IF systems can consume a lot of power, and low power drain by the K3's receiver was a design goal I believe. 73, Geoff GM4ESD On Saturday, February 06, 2010 at 5:17 AM, Guy Olinger K2AV wrote: > Hi Dave, > > Using slope 000 is likely most of your problem. With that setting all > signals regardless of strength are set to the same level regardless of > how loud they are coming in. If they are zero beat or close, it will > mush the dickens out of the copy by making them all the same level in > your ear. I set AGC SLP to 15 for contests. I also use slow AGC the > entire time. If there are key clicks, I use NB with IF off and dsp to > t3-7 or t2-7. It rounds off CW bauds some but does not make them > uncopyable. > > Make sure your ATT/PRE/RFGAIN use conforms to something like: > > 160m ATT plus RF gain at 2 oclock > 80m ATT plus RF gain at 3 oclock > 40m ATT plus RF gain fully clockwise. > 20m " " > 15m off plus RF gain fully clockwise. > 10m PRE plus RF gain fully clockwise. > 6m " " > > ...if you are listening on your transmit antenna. At least to start. > Make sure the ambient noise on the band is moderately low audio > listening on a clear frequency. > > Using the NB with those settings, including the ATT/PRE/RFGAIN > settings. Just today I was listening to an S4 Cuban underneath what > were S9/5 key clicks with the NB off. Offending station up 500 Hz at > 30 over 9 (really), and was using 250 filter running at WIDTH 350. > > Good luck in Sprint this weekend! > > 73, Guy. ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html __________ Informacisn de NOD32, revisisn 4576 (20091105) __________ Este mensaje ha sido analizado con NOD32 antivirus system http://www.nod32.com ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
