[RTTY] K3 AFSK power and ALC?

[Kok Chen]

On Sep 23, 2010, at 9/23    7:09 AM, In reply to Steve ZP9EH, Richard  
Ferch wrote:

>> Wow, the big news is I have a K3...
>
> 3. Even in PSK31, which is very sensitive to IMD and requires good
> linearity, the K3's output is clean even with 5 bars of ALC showing on
> the meter. The critical adjustment on the K3 for PSK31 is not the mic
> gain, it is the transmitter power.

While the K3 is easy to adjust for best transmit IMD, the close spaced  
transmit IMD (i.e., for PSK31) on my K3 is about 3 dB worse than what  
I can get on my FT-1000MP.

You can take a look at narrow spacing transmit IMD that I'd measured  
here

http://homepage.mac.com/chen/Technical/K3/Digital/digital.html

It is not bad per se, actually quite good; but just not as clean as  
the cleanest signal that you find on on the air in PSK31.

The IMD is certainly good enough to take advantage of any narrow band  
filtering that your software might have when transmitting RTTY (see  
spectra at the bottom of that page). Notice that with filtered AFSK  
(second spectrum), the signal from a K3 is cleaner than using the K3  
in FSK mode (first spectrum).  Filtered AFSK is a linear signal (i.e.,  
mark and space can have a slight overlap) and you do need good IMD  
figures to avoid generating extra sidebands.

And Steve, if this is a used K3, be sure it has a recent (released  
final) version of the firmware.  For the measurements shown, I was  
using MCU firmware version 3.68.  An earlier version 2.23 had problems  
with AFSK.  The earlier problem was not caused by IMD -- but from  
overshoots during mark/space transitions caused by the DSP stage,  
resulting in excessively wide keyclicks.

Also, when the audio drive is too low, and the K3's RF output when set  
to low power, the output power tends to rise slowly to the final RF  
"PWR" setting each time you key the rig -- what I indicated as  
"creeping" in the tables.

If you see power creeping, increase the MIC gain control enough so  
that the "ALC Meter" starts at 5 bars each time you key the  
transmitter.  If you see RF output creeping, my measurements showed  
that the transmit IMD also is creeping.

Don't go over 5 ALC bars, since transmit IMD will start to rise from  
that point on.  3 bars work without creeping if you use more than 20  
watts of RF power.  5 bars will work at all the power levels that I'd  
measured, without increasing transmit IMD.

On the K3, the first 5 bars on the "ALC meter" does not indicate PA  
ALC.  Instead, the bars indicate the audio/microphone drive as seen by  
the DSP stage ("audio ALC" if you wish).  It is only after the 5th bar  
that the "ALC meter" indicates the amount of ALC applied to the PA.

When your audio drive is down in the 1 to 3 bars region, the DSP will  
need to bring up the audio gain so that the RF power output that you  
have dialed in is (eventually) attained.  The gain compensation  
happens relatively slowly, since it does not want to follow SSB audio  
peaks; they probably could have used a different time constant in DATA  
A modes.

Given enough audio drive, the K3 in AFSK (DATA A) mode is a "set once  
and forget" function, and you can adjust the PWR knob at will, unlike  
rigs where you control RF output power using the AF level.

Further, in DATA A mode, the VOX works well for full duty cycle modes  
(and almost full duty cycle modes like PSK31), so all you need for  
RTTY is a sound card.  You do not need to build anything else to key  
the PTT.

73
Chen, W7AY