[Elecraft] K2 & PSK-31

[Julian (G4ILO)]

Rod,

> I do not have the KIO2, I do have the updated FW. I want to use
> narrow filters. I like the wide waterfall of digipan.

I'm assuming you meant to say you DON'T want to use narrow filters. That's
your prerogative, of course. However I'm personally of the view that
Digipan is a flawed approach to PSK that may be appropriate when using
dedicated radios such as the PSK20 which have (I assume) a broad filter
covering the whole of the band where PSK signals are to be found, and also
(I assume) no AGC.

Trying to work PSK using the standard SSB filter on the K2 is fine if the
band is quiet, but when there is a lot of activity (and there is ALWAYS a
lot of PSK activity these days) strong signals will activate the AGC
reducing the percentage copy from the wanted station unless it is a strong
one. I don't think it is possible to disable the AGC in the K2, only change
it from FAST to SLOW. Because of this, it is really valuable to be able to
use a narrow filter when working PSK. Many a time I have lost a signal
completely when a strong station came on nearby, until I switched in the
narrow filter.

I'm also convinced Digipan is responsible for some of the weird signals I
see, where one side of the PSK signal is much stronger than the other. I
guess that this occurs because people are working on an audio frequency
that's on the steep part of the SSB filter passband. If you use narrow
filters, you get accustomed to tuning the wanted station to a frequency in
the center of the passband, so that's never a problem.

> Why? It seems to work just fine. Like I said I'm getting good on
> air results. KSB2 manual says 2:1 compression is OK to use.

Compression is never going to do any good to the IMD of a PSK signal.
Perhaps the K2 manuals have changed since 1999, but I don't think there is
any reference to recommended settings for PSK operation in *my* manual at
all. I don't think you can rely on on-air reports as a measure of signal
quality. Only a couple of days I worked a guy who was so overmodulated that
even his CW ID had multiple IMD products. I pointed this out, but he took
no notice. A few minutes later I noticed him asking his next contact what
his signal was like and the reply was "excellent signal, no problem." It's
so frustrating...

> I'm using the cable as described in the Digipan Help files. It
> works fine. The cable plus the sound card control enable me to
> set the audio just fine. The issue is setting it with a tracking
> ALC, In my 746, using the exact same cable, I can vary power
> output evenly from 0 to Max_Pout_setting. From the KSB2 manual:
> 
> "Tracking ALC: SSB transceivers use Automatic Level Control (ALC)
> to reduce transmit drive if output exceeds a desired level. Many
> transceivers require that you adjust the mic gain manually until
> the ALC indication is in a certain range. The K2's ALC is
> different: the drive level is automatically adjusted, over time,
> to keep ALC close to the optimal level. This tracking ALC (TALC)
> eliminates the need to adjust mic gain in most cases. You can
> set the POWER control, mic gain, or speech compression to any
> level, and the ALC level will average between 1 and 3 bars (in
> ALC meter mode) as you talk."
> 
> So it stand to reason that if 1-3 bars is good for voice, it
> should work just fine for PSK-31.

I don't think that is true, because any extra distortion added to a PSK
signal by ALC action is much more evident than it would be on an SSB
transmission. I think you need to read KI6WX's explanation of his
modification to the KSB which changes the ALC time constant. With that
modification in place, some ALC action won't noticeably worsen the IMD on a
PSK31 signal. Without it, your signal is wider than it need be. PSK31 is a
narrowband mode whose potential is being wasted because of the number of
overmodulated signals that interfere with reception on several hundred Hz
either side of the frequency. It's in everyone's interest to take the steps
to ensure our signals are as narrow as they can be, even if it means paying
more attention to the audio drive levels instead of just letting the
transceiver take care of it.

By the way, with my cable I CAN adjust the power output smoothly using just
the mixer level control, if I so wish, so if your cable isn't delivering
too much audio to the mic input I can't think of any other reason why you
should find the power increase so sudden. Unless (DING! Light bulb goes on)
the reason is that TX only happens once there's enough audio for the VOX to
kick in, of course...

> [VOX] Works fine at the 0.4 setting,

That seems to be on the assumption that up to three bars of ALC is
acceptable. I didn't consider that to be acceptable, and on my K2 I could
not get the VOX to trigger with the amount of audio necessary to just reach
5W output *without* any ALC action (this was before KI6WX came up with his
mod, so I didn't consider *any* ALC to be conducive to a clean signal.)

> Many new laptops don't have a serial interface, mine is one of them. I
> use VOX on my Icom 746pro, I also used a homebrew VOX circuit on a
> PSK-20. Now with the K2, never a problem, works great.

The other reason for not using VOX is the ease with which it is possible to
broadcast Windows sound effects and other audio when people forget they
have their radios connected to the computer. Many people are using USB -
serial converters. However, I take your point.

I guess it *may* be acceptable on the newer K2s or those with the KI6WX mod
to use VOX and drive the rig to one or two bars of ALC without unacceptable
widening, but personally I'd like to get the OK from Wayne and Eric, and
see some figures on the effect of the K2's ALC on the IMD of a PSK signal,
before trying it. On most transceivers, driving them into ALC is
responsible for the many lousy signals we see, which is why I'm so strongly
against it.

73,
--
Julian, G4ILO