[Elecraft] K3 Repair

[Don Wilhelm]

Jack,

Yes, the K3 has to recalculate the power drive after a power cycle or a 
band change - so let's not call that part of the process "creep", it is 
just the K3 power control mechanism coming up to speed with the power 
setting that you requested using the Power Knob.  The power will start 
out a bit low and then come up to the level requested.  It will then 
stay there until a power cycle or a band change.

What power level do you have set?  That is an important piece of 
information.

Initially, ignore the power output and set the audio level correctly.  
You can (and should) use TX TEST for that adjustment.
For SSB, talk in a normal voice.
Set Compression to zero
Then adjust the mic gain until you have between 5 and 7 bars indicated 
on the K3 ALC meter display - the peaks should be at 7 bars.
Now add compression as desired.
Set your wattmeter to read the peak power (not peak-hold in the AVG 
setting).
Connect a dummy load and speak into the microphone.  After a few 
syllables (to allow the power control mechanism to stabilize), how 
closely do the wattmeter peaks come to the power you set on the power 
knob?  It should be close.

For data modes, the process is similar.
Use TX TEST to set the audio level.
Set Compression to zero (it should stay there for all data modes)
Adjust the audio into the K3 to provide 4 to 5 bars on the ALC meter - 
actual ALC does not start until after the 5th bar, so that will keep the 
level below the onset of ALC.
Set the K3 power control to the power output you want.
Transmit, and after the initial power leveling mechanism has had a bit 
of time to settle, the power should equal what you dialed in on the 
power control.

Yes, for DATA modes, the K3 is different than other transceivers out 
there - ignore the usual recommendation to set the power high and use 
the audio input level to control the power - due to the way the K3 
controls power (by sampling the actual output and adjusting drive to 
compensate), attempting to use the techniques that vary the audio level 
in order to control power will result in "strange happenings".  While it 
will work, I will not guarantee the power level will be right, and you 
will have to ride the audio input level to keep it constant (don't do it 
that way).

What I have related is nothing different than the information in the K3 
manual (except for the maximum number of ALC bars in DATA mode.
If you wnat to double check the veracity of what I am saying, look in 
the archives for Lyle Johnson's reflector post of February 8, 2009.  
Lyle designed the K3 DSP (and its control of the power out), so he 
should be considered an authoritative source.

73,
Don W3FPR

Jack Brabham wrote:
>  Don,
>
> What I'm watching to see the "creep" is the W2 meter's Windows app 
> with the meters set to AVG and the peak hold bar ON.   I'm 
> interpreting this as the "peak" of the fluctuating "average" power 
> level????  This is the number that rolls up around 50% during the 
> first 20 seconds of an initial transmission after a power cycle.
>
> If I change the meters to "peak", the numbers change as you would 
> expect, the initial "peak" power starts off around 70 watts and drifts 
> up to around 95.
>
> BTW normal PSK31 etc operation in DATA A seems to proceed as expected 
> as long as I stay above 1000 hz or so for the audio.  Below that there 
> is an increasingly noticeable reduction in PO, but otherwise functions 
> normally.
>
> I'm using my PSK31 program as an audio generator for these tests, not 
> having anything better to use.  In this role  it is not adjusted the 
> same way one would for actual PSK operation.  Do not try this at 
> home...... ;)
>
> I have tried, in the last hour or so, increasing the audio drive 
> levels and decreasing the mic and line-in gain.   This somewhat 
> reduced the power creep issue, at least in terms of the number of 
> transmission cycles before it "settles",  but the radio still seems to 
> have to "relearn" this from scratch after a power cycle.
>
> Reducing the Mic gain from 18 to 6 and increasing the output from my 
> mic pre-amp to get  4-5 bars of ALC, did not seem to help the low 
> output on SSB problem.
>
> Thanks for the tip about the ALC meter, makes more sense now.
>
> 73 Jack KZ5A
>