[Elecraft] K2/100 Power Fluctuates Cyclically on 40, 30 Meters; K2 s/n 06660

[Don Wilhelm]

Tom,

The short answer it that the addition of the gain provided by the KPA100 
is inside the power control loop, and because of that additional gain, 
the control loop is no longer being critically damped so it oscillates 
about the correct power level for  a while.

There is a fix - see http://www.elecraft.com/Apps/Power_Control_Mod.html.

Each K2 is a little different, so you will have to experiment with the 
correct value of R98 for your K2.  Use a potentiometer as Wayne suggests 
and adjust for the highest value resistance that will still allow an 
acceptable amount of power output from the base K2 on 10 meters (you get 
to select how much power reduction on 10 meters is OK with you.  Power 
only the base K2 when doing that test, so that the KPA100 is off line, 
but the wattmeter in the KPA100 is operational - you can also turn the 
KPA100 off in the menu, but I find it easier just to pull the APP 
connector and put a coaxial plug into the base K2.

Since you asked about the K2 power control, I will give you the long 
answer too.
The K2 microprocessor knows the power you have requested via the power 
knob, and it also knows the power output that is detected by the KPA100 
wattmeter, or in the base K2, by the RF Detector diode (this is signaled 
to the uP on the VRFDET line).  If the power is lower than requested, 
the uP increases the voltage on the VPWR DAC output, and after 
waveshaping, applies the resulting voltage to RF Board Q34 as the VALC 
signal to control the BFO injection level to the transmit mixer.  Since 
the gain of this power control loop contains the overall RF gain of the 
transmit stages, if the RF gain is increased, the control loop can 
overshoot the correct level, then it corrects again and overshoots on 
the low side, etc.  When the control loop is critically damped, it will 
not overshoot and oscillate, but if it is overdamped, the power control 
response will be too slow, and may never get up to the full requested 
power output (that is why it is tested on 10 meters where the TX RF Gain 
is lowest).  The element that controls the amount of loop damping is R98.

73,
Don W3FPR

TRBOAG at aol.com wrote:
> *Just finished K2/100 assembly and verified power output on all bands.*
> *I notice at about 15 watts output a cyclic (1-2/second) fluctuation 
> in power from 13 to 16 watts as shown on Elecraft display (in Tune 
> Mode). CW Mode, XFIL1. Driving MFJ-949E as Dummy Load Mode and see 
> same fluctuation in forward power meter there. SWR is consistent 
> 1.0:1. Power from Alinco DM-340MV. K2 contains SSB and NB modules. Fan 
> runs slow some of time during tests but, after shutting off K2 and 
> power supply, felt for warm components and found nothing - Heatsink 
> only mildly warm.*
> *Wondering what could account for that cyclic fluctuation and also how 
> power output regulation works... software? sampling 1-2/second???*
> *Regards, Tom*