[Elecraft] K3 RF Output Level Variation

[Don Wilhelm]

Dave,

You would be better off not touching the power knob once it is set for 
your desired power.

Unlike most other transceivers, the K2 and K3 control power output using 
a closed control loop.
The setting on the power knob is compared to the power output (as 
measured by the wattmeter) and changes the drive to make them equal.
The power should start out slightly less than that requested by the 
power knob and build to the set value.
The control loop operating point is reset after changing the power knob, 
a band change or at power on.
Every time you move the power knob, it is reset and the control point 
must be re-established.
It takes a couple dit times to come up to the full requested power at 
normal keying speeds.

Yes, you may benefit from a wattmeter calibration if your requested 
power is different than your measured output power.  You might also want 
to investigate the accuracy of your external wattmeter.  Many are not 
accurate at the 5 watt level - that is particularly true of analog 
meters which typically specify power as a percentage of full scale.  on 
a 20 watt scale, even 5% accuracy will be an error of 1 watt, and that 
error applies to any point on the scale.

73,
Don W3FPR

On 2/17/2014 12:17 AM, Dave Barr wrote:
> K3-100 sn 7901 is working great;  I am still in the learning cycle, 
> but I have noticed an interesting and slightly disturbing  thing about 
> the rf output levels.
>
> During measurement with an external (HP) power meter, the power output 
> increases rapidly for about 10 seconds and then slows and eventually 
> settles, with small variations from every few seconds afterwards.  For 
> example, to achieve a 5 watt reading on the power meter, the initial 
> setting on the K3 is 4.3, and within 10 seconds must be reduced to 3.9 
> to maintain 5 watts output.  (Watt meter calibration has not yet been 
> performed.) This occurs to similar extents at all power levels, i.e., 
> to maintain an even power level, the K3 setting must be reduced over 
> the first 10 seconds or so.   My K2 does much the same thing.  Is 
> there an explanation for this behavior?