KPA power fluctuations

[David Walker-8]

I think I have to some calibration because I only get 85w out on the kpa-100.  No worries. I read this is easily fixed.

However I notice my power output fluctuations between 17 and 22 watts with the tune button.  It clearly goes up and down in a tick-tock fashion.  

Can anyone exlain why this would be happeing?


Dave
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Don Wilhelm-3]

David,

This question comes up from time to time, so let me attempt a long
explaination for general information of those interested.

The power control for the K2 is speced at 10%, and you are seeing a +2 watt
and -2 watt variation at the 20 watt level, so it does fall within spec
limits.

The power variation you see during TUNE is a result of the power leveling
controls attempting to bring the power into the range of power requested
(forced to 20 watts during TUNE).

The K2 controls the power by means of a loop function (I will refer to it as
the 'ALC loop') to provide constant power equal to the level requested by
the power knob - the loop includes the entire trnsmit chain since the actual
output power is sampled right at the antenna jack.  This method of power
control while allowing constant power output without dealing directly with
changes in the gain of the transmit chain from band to band, does have one
'drawback' - such a control loop must have hysterisis in order to be stable.
Without the hysterisis, the power control could overshoot drastically and
vary wildly from minimum to max while searching for the proper central
level.  The hysterisis range limits are embedded in the firmware function,
and only Wayne B. could change them (if the limits could indeed be reduced
at all).

The overall gain of the ALC loop in the K2 can be reduced by increasing the
value of RF Board R98 in the K2 which may help smooth the variation cycles
(but it will not eliminate it).  When the value if R98 is increased
dramatically, the power output available on the higher bands will decrease
(because the transmit chain gain is lower).  The best value can be
determined by temporarily substituting a 1K or 2K pot for R98 and adjusting
it to the highest value that will provide adequate 10 meter output power (it
is your decision for how much is 'adequate') - then measure the pot value
and replace it with a fixed resistor equal to or less than the measured
value of the pot.

Some hysterisis must be included in any loop function like this for the loop
to be stable (it is a mathematical condition for loop stability).  The power
variation during tune is the result.  It is not a problem even though it may
be disconcerting when 'watched'.  If you do a long TUNE, it will eventually
settle down to the 20 watt level.

Bottom line is that there is no 'fix' for the variation.  It might be
considered similar to the control provided by a heater thermostat - there
must be a range of control (say it is 1 degree F) to work effectively - set
it at 72 degrees and if the temperature in the room drops to 71 degrees the
heater turns on and stays on until the temperature goes up to 73 degrees.
This is the hysterisis range at work - if the range were zero, the heater
would cycle on and off rapidly (which would not be a good nor effective form
of operation).

73,
Don W3FPR


> -----Original Message-----

> I think I have to some calibration because I only get 85w out on
> the kpa-100.  No worries. I read this is easily fixed.
>
> However I notice my power output fluctuations between 17 and 22
> watts with the tune button.  It clearly goes up and down in a
> tick-tock fashion.
>
> Can anyone exlain why this would be happeing?
>
>
> Dave
>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Torsten Clay]

I still think there are some oddities (firmware bug?) in the K2 ALC algorithm. I still haven't figured out why I sometimes get 170W out on 160m.

Torsten
N4OGW


_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Don Wilhelm-3]

Thorsten,

If I recall your original description, that high power glitch is a result of
the 'attack time' delay in the power control circuits.

Yes, I do believe there is a firmware oversight - the power leveling
controls must change (to compensate for the gain variations) when switching
from band to band, and when on 10 meters, the transmit chain gain is low, so
the power controls are set for a higher level - when going directly from 10
meters to 160 meters (where the transmit chain does not need as much drive),
the control circuits are still primed for the higher drive level and will
drive hard until the power output is actually measured by the microprocessor
the drive level decreased accordingly.

The alternative is to always reset the power drive to a very low setting
when changing bands and build up to the requested setting in a few dot times
of power output.  I believe some of the past firmware releases actually did
it that way.

Perhaps if you ask Wayne about it, he may consider doing something different
for the next firmware release - your operating situation (overdriving a
linear, etc.) may be important pieces of information in helping Wayne decide
which method of power initialization is the better one. Right now it seems
like the power drive level is not changed when switching from band to band.

73,
Don W3FPR


> -----Original Message-----
>
> I still think there are some oddities (firmware bug?) in the K2
> ALC algorithm. I still haven't figured out why I sometimes get
> 170W out on 160m.
>
> Torsten
> N4OGW
>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Vic K2VCO]

Don Wilhelm wrote:

> The alternative is to always reset the power drive to a very low setting
> when changing bands and build up to the requested setting in a few dot times
> of power output.  I believe some of the past firmware releases actually did
> it that way.

Mine (2.04p, IOC 1.09) certainly does.
--
73,
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com