The killer for that wide roofer approach is that loud enough stations can
pump the hardware AGC before the signal gets to the DSP.
The "full body" (whatever that means) CW technique will NOT work in a DX
test I've been in (Will knows where) with that 45 over S9 Italian station
running 15 kW and a 4 element beam on 40m pointed at the US when the band
is wide open, and who has parked 400 Hz above or below me, AND I am trying
to copy an S0 (if even that strong) basement noodle antenna QRP station.
I want the skirts of the roofing filter and the DSP to COINCIDE and when I
hit the edge of the 450 or 350 Hz, I want the skirts do a VERY sharp dive
to oblivion. This gives me a 450 kHz running bandwidth and a 350 running
bandwidth, which I can shift in 10 Hz steps, dropping a signal on the edge
about 10 dB per 10 Hz step.
This way I get the same excellent selectivity with a K3 that I had with my
MP with cascaded INRAD filters in the 8 and 455 IF's. K3 is not superior to
the MP in selectivity or audio quality, but the K3 front end and low noise
opened up a new layer on 40 meters, going down past the DL and OK basement
noodle QRP stations and exposing the RUSSIAN basement noodle QRP stations.
We thought that was band crap listening to the MP, but turned out to be RX
IMD, RX noise crap generated by the MP. Finally exposed by not being there
in the K3.
If my skirts do NOT dive maximally, then the 15 kW Italian is going to pump
the K3's hardware AGC and make copy of my basement noodle QRP stations
I do know those (so far all in the western US) who insist that the DSP
selectivity is all they need. They might be right, out where they are.
Idaho is not exactly the land of monster signals. If one really wants to do
"full body CW" out there, a 700 Hz roofer would be a better choice to be in
front of the DSP. Do you really want your AGC pumped by the next two or
three stations above and below your frequency?
I've spent my time with radios with too-wide high IF selectivity. NEVER
Besides, having the steep diving combined selectivity converts key clicks
to spikes, which the K3 NB and AGC circuits handle splendidly, allowing key
click elimination, another K3 advantage trying to keep a run frequency 400
Hz from a 45 over S9 15 kW Italian station.
You think I'm kidding about 15 kW and 4 elements on 40m? Think again. Not
kidding about 45 over 9 on a calibrated K3 S meter either. Though we *were*
listening on a five element wide spaced (190 feet end to end) 40m quad.
73, Guy K2AV
On Mon, Sep 19, 2016 at 9:21 PM, WILLIE BABER <wlba...@bellsouth.net> wrote:
> Hello All,
> I did M/M with a big station in the NW running Icom 7700. Needless to say
> comparison between K3 (one was on site) versus Icom radios came into
> discussion during rest periods. Yes, the Icom 7700 played quite well,
> giving what some called "full-body" cw and easy to pick-out-stations while
> running them.
> I like all radios...well almost all of them. Having said that, you can
> easily get "full-body" cw and better running of station if you simulate the
> front-end of the Icom radio using K3. To do this, set the first I-F filter
> to ssb bandwidth (2.7khz) and then use the DSP to deliver 500 hz passband.
> You will hear stations that are within the 2.7khz window but not quite on
> your frequency (similar to 7700) while running stations, and without having
> to fool so often with RIT.
> Once you go to 400 hz in the first I-F of K3 you are in serious high
> dynamic range territory...better have a k-pod on RIT and forget about
> "full-body" cw, which requires a wider first I-F. Of course when you need
> a narrow first I-F you can have several of them in K3.
> My two cents.
> 73, Will, wj9b
> CWops #1085
> CWA Advisor levels II and III
> 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
> Message delivered to k2av....@gmail.com
Elecraft mailing list
This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to arch...@mail-archive.com