[Elecraft] K3: Odd ALC behavior

[Wes Stewart]

Wayne,

There are no beat notes. These frequencies, at least from the FT8 program in 
Tune mode are single frequency.  Based upon output RF spectrum measurements*, 
with for example 1010 Hz input and 4-5 bars of ALC, I see spurious or 
intermodulation sidebands, down -65 dBc.  If I change nothing else but adjust 
the tone frequency to 1000 Hz  the spectrum is the same but the ALC bars are 
cyclic, going from none to 4-5, at about the 2 HZ rate Brian reported earlier.  
(Slow ALC in action?)  That's the good news.

If I set the tone to 1500 Hz, the FT8 mid-band target, I'm presented with the 
cyclic ALC and some ~193 Hz, -50 dBc sidebands.

I have tried this with both the K3S internal sound card and via Line In/Out into 
my laptop sound card.  The results are more-or-less the same.  But IMHO, you 
need to put a 30 dB attenuator on the input of the internal sound card; it is 
way too sensitive.  I'm having to run the Windows gain at less than 5% and Line 
In at 10 or less.

I would suggest that you do some measurements with a low-distortion sine wave 
audio source run into the radio to eliminate the possibility that this junk is a 
result of all of the digitally generated tones.

If this (what I will call a) problem continues to exist then I think a fix is in 
order.  Expecting users to select particular frequencies to use to set gain 
levels isn't an acceptable workaround.

Regards,

Wes  N7WS


* My "spectrum analyzer" is my SDR-IQ which I have used to measure TX IMD on 
both my K3 and K3S.  Your IMD
  measurements on my K3S concurred with mine within 1 dB or less.

On 8/4/2017 4:45 PM, Wayne Burdick wrote:
> Is it possible that a beat note between tones (and their harmonics) is what’s modulating the ALC reading? If so, moving the tone slightly off the standard pitch may be an acceptable workaround.
>
> Wayne
> N6KR
>
>
>> On Aug 4, 2017, at 1:42 PM, Wes Stewart<wes_n7ws at triconet.org>  wrote:
>>
>> Good find Brian!!! I see it too in FT8.
>>
>> This probably explains the great variability in results reported. Some of us have a problem, others say, Huh, what the heck are you talking about?
>>
>> Wes, N7WS
>>
>>
>> On 8/4/2017 11:33 AM, Brian Hunt wrote:
>>> Here's some followup:
>>> FLDIGI shows the "pumping" affect with carrier tones of 500, 1000, 1500, and 2000 Hz.
>>> WSJT-X 1.8 shows the "pumping" affect with carrier tones of 1000, 1500 and 2000 Hz (couldn't select 500 Hz).
>>>
>>> The affect is very narrow with frequency.  10 Hz either way and the output and ALC display and MON output is rock solid.  Changing the K3 filter center frequency has no effect on the above.
>>>
>>> 73,
>>> Brian, K0DTJ
>>>
>>> On 8/4/2017 10:58, Brian Hunt wrote:
>>>> I spent some quality time with my K3, FLDIGI and an oscilloscope today.  I had the same problem as described.  Looking at my USB soundcard output to the K3, the amplitude of the audio from FLDIGI in TUNE mode was rock solid. Looking at the MON output from the K3, the level oscillated in time with the ALC "pumping" at about 2.4 Hz (timed with a stop watch).  At the same time the K3 power output into a dummy load was rock solid at a selected 10 watts.
>>>>
>>>> Then I changed the carrier frequency in FLDIGI slightly off 1500 Hz and the "pumping" went away!  At 1515 Hz I could easily set the ALC to the desired 4 bars solid, 5th flickering.  Then I remembered (and verified) that the default filter center in DATA A mode is 1500 Hz. I changed the filter center frequency to 1450 Hz and moved the FLDIGI frequency around but couldn't reproduce the "pumping".
>>>>
>>>> I plan to repeat the same tests with WSJT-X 1.8 but wanted to get this out to the group.
>>>>
>>>> 73,
>>>> Brian, K0DTJ
>>>>