Posted by
wc1m on
Dec 06, 2011; 7:00pm
URL: http://elecraft.85.s1.nabble.com/Fw-CQ-WW-DX-Contest-CW-2011-Unofficial-claimed-scores-tp7060148p7067968.html
Awesome, Guy. I think your test will answer the questions I have about this
problem. Please post the results of your test with a different title than
above so we don't have to comb through a lot of posts to find it.
73, Dick WC1M
-----Original Message-----
From: Guy Olinger K2AV [mailto:
[hidden email]]
Sent: Tuesday, December 06, 2011 3:17 AM
To: David Gilbert
Cc:
[hidden email]
Subject: Re: [Elecraft] K3 AGC Mush ... or "blur"
Not to argue but to move toward clarity...
Low level signals are by definition closer to the noise. That would also
mean that there would be more "stuff" between CW bauds. "Crisp"
to some means that the artifacts which define the beginning and the ends of
a baud are there. The extreme artifact is a key click. A normal artifact
is additional bandwidth occurring at the edges of the baud, but reduced in
level. These are down 6, 12, 18, 24 dB from the pure sine wave. Without
these, the baud sounds soft, and the transitions on and off only excite the
directly on-tone nerves in the ear, instead of some number of tones.
Once the signal is close enough to the noise the baud start and stop
artifacts become obscured by the noise, but the signal is still copyable by
itself. If there are multiple signals in this state and the tones are
close, our ears lose all clue of baud on/off other than slow (not "sharp")
level transitions. We lose harmonic artifacts, and now must rely only on
very soft level changes. Ancient Bell Labs studies on these issues for
telephone would say that for most people, a change from one CW signal to
another at close to the same frequency, less than three dB in amplitude,
would not be a perceptible change.
On these issues in that QRM situation, one would EXPECT that kind of
"muddyness" in the S unit above the noise level. This would happen in a
perfect receiver. This is why someone's signal sending fast when they go
into the noise "blurs" the CW and the only way to overcome that is to slow
down signficantly.
**IF** the AGC was using settings that removed even those level differences
from hearing, the muddying would begin at higher levels above noise. This
would happen in a perfect receiver.
**IF** the hardware AGC was exposed to significant CW signals outside the
DSP skirts due to use of a much wider roofer than the DSP width setting
(such as one listing of 500-800 Hz DSP under a 1000 roofer) the variable IF
gain stage would be tracking the HAGC imposition on the gain, and raising
and lowering the signal of the collective IN-DSP signals with foreign on/off
information not related to any of the in-band signals. Muddle would be a
good word to describe that effect.
This is in a volume region where one is already losing clarity keying
artifacts to the noise floor.
***This effect would occur the worst in the range between where the DSP AGC
started to take up down to the noise***, and would otherwise be flattened
back by the DSP AGC responding only to the in-band signals and treating the
HAGC variations as "propagation changes".
Slowing down the hardware AGC would prevent in-roofer-out-of-DSP baud-speed
external signal gain modulation to weak, under AGC in-band signals.
I should note for the record that in contests I am either running a
400 8 pole with DSP at 450 or a 250 8 pole with DSP at 350 or 150.
When the offsets of the filters are carefully adjusted, these produce the
sharpest composite skirts down 30 and 40 dB. There is never any signal
outside the DSP limits making it to the hardware AGC. I always use slow AGC
with the fastest possible decay. I always have my PRE/ATT and RF gain
pushed back to match the band.
There may be something else going on, but with my settings I have never
heard any of the reported muddiness, other than the expected perfect RX
conditions listed above.
I am going to try the cap modification to my KRX3 only and listen in
diversity to a single antenna. I will also set the main RX to keep the 400
8 pole as roofer all the way down. Then at 350 bandwidth, I will be
listening to a wide roofer on one and an equal roofer on the other.
Should be easy to put an XG3 + step attenuator tone IN band with a loud CW
signal between the skirts and see what happens. Anything audio happening to
one RX and not the other should be really obvious.
73, Guy.
On Mon, Dec 5, 2011 at 12:39 PM, David Gilbert <
[hidden email]>
wrote:
>
> "The only time I heard mush on the K3 was when we had worked down the
> pileup and only had very low strength guys all calling on the same
> frequency (the packet/reverse beacon spotting phenomenon)."
>
>
> Multiple low level signals close in frequency is EXACTLY the situation
> that I and at least several others are concerned about. I don't
> notice the problem on stronger signals either, and my low level
> hearing is still pretty good. Operating from ZF-land where you
> almost always have somebody strong calling you is not the typical
> situation many of the rest of us experience from our home QTH's, and I
> would therefore expect that you are less likely to experience the
> problem. Great sensitivity and dynamic range are less useful if what
you hear cannot be deciphered.
>
> By the way, it has been pointed out to me that "mush" may not be the
> best descriptor for what I hear, and I agree with that. I think "blur"
> might be a better term. When the problem shows up I can distinguish
> the individual tones (the signals are not dead zero beat) but the
> crispness of the keying disappears. It's as if something is filling
> in the spaces between the keying elements, and I tend to believe that
> the culprits are the sum/difference products of other nearby signals
> that happen to overlap the signal I'm trying to copy.
>
> 73,
> Dave AB7E
>
>
>
> On 12/5/2011 8:34 AM, K5WA wrote:
>> I really can't agree with the mush comment. I've just returned from
>> ZF1A where we put almost 8000 Qs in the log over the CQWW CW weekend.
>> I took my
>> K3 and was able to evaluate it against the station's existing TS-850
>> since we were M/S and rotated positions. In my opinion, the 850 WAS
>> mushy and the
>> K3 was very clear however, K6AM, our host (who is a VERY serious
>> contester), prefers the 850 since he has been using it for 10 years.
>> He also owns 4 K3s but didn't bring them on the trip since the 850s
>> "live" in Cayman and he brought many pounds of other equipment.
>> Personal preference plays a huge part of this discussion. In this
>> case, K6AM feels like his hearing is limited to a frequency range of
>> 6-7K while mine is still near 20K. Maybe that is part of the puzzle.
>>
>> We regularly saw the 10 minute QSO rate meter in the 300 range and
>> topped
>> 400 occasionally. I loved the way the K3 was able to plow through
>> the pileups which were calling us and work through them as fast as
possible.
>> The only time I heard mush on the K3 was when we had worked down the
>> pileup and only had very low strength guys all calling on the same
>> frequency (the packet/reverse beacon spotting phenomenon). These
>> must have been micro-watt QRP'ers or stations with minimal/indoor
>> antennas. I had my AGC on and didn't have any time to try multiple
>> settings but I am extremely happy with the K3's receiver and think it is
in the high end competition grade for me.
______________________________________________________________
Elecraft mailing list
Home:
http://mailman.qth.net/mailman/listinfo/elecraftHelp:
http://mailman.qth.net/mmfaq.htmPost: mailto:
[hidden email]
This list hosted by:
http://www.qsl.netPlease help support this email list:
http://www.qsl.net/donate.html