When the KPA100 (the K2's 100w amplifier stage) came out, some of us noted
similar issues.
The gain available from a transistor varies in the opposite direction to
frequency, and
the circuits include feedback mechanisms to try to balance it. The
characteristics of real
components have some effect on the behavior of the feedback network; in the
case of the
KPA100, some resistors had more inductance than expected, which resulted in a
severe drop
in gain on some bands.
I guess what I'm saying is that gain variations are expected, and unless the
efficiency is
bad enough to endanger the devices, I wouldn't worry about it.
On 2/7/2012 5:35 AM, Brian Alsop wrote:
> Guys,
>
> I'm also curious about this-- especially what's going on on 15M.
> My values at 500 W out into a dummy load are:
> (amp), [volts], {efficiency}
> 1.8MHz DC (11.5) [63.1] {70}
> 3.5MHz (11.2) [63.5] {71}
> 7MHz (12.2) [62.5] {67}
> 10.1MHz (not measured)
> 14MHz (13.4)[61.8] {60}
> 18.1MHz (12.9)[62.8] (63}
> 21MHz (15.8)[61.2] {52}
> 24.9MHz (14.6) [60.1] {56}
> 29.7MHz (14.7)[60.4] {56}
>
> These are in reasonable agreement with David's except for perhaps 10M.
>
> The problem with these measurements are the wattmeter accuracy (10%?)
> and the current measurement inaccuracy. Default wattmeter calibration
> factors were used above. Elecraft is using a Hall effect device to
> measure PA current. It would be nice to know just what accuracy to
> expect from it.
>
> The efficiencies appear to be in the right ballpark. They tend to fall
> on a straight line log plot vs frequency, decreasing at higher freq's.
>
> If nothing else, the above numbers serve as a benchmark for future
> comparisons down the road for my amp.
>
> Note: The current measurement can be significantly affected by RF on the
> outside shield of the coax. Until a choke was employed on the feedline,
> the 24.9 MHz current with a real antenna measured 3 amps too low. With
> the choke it agreed with the above dummy load value. I suspect lots of
> KPA-500 users do have this condition-- given the types of
> antennas/tuners et al described by uses on this reflector.
>
> 73 de Brian/K3KO
>
> On 2/4/2012 20:27, david park wrote:
>> I have recently assembled KPA500 serial number 641. When testing into a
>> dummy load I find that I get a large variation in the DC power input when
>> driving for 500W CW when I go from band to band. My worst band is 21MHz,
>> where I get the highest current (lowest efficiency). The drive was adjusted
>> to give 500W as indicated by the LCD display. Here are my results:-
>>
>> 1.8MHz DC current 10.0A fan hardly ever comes on
>> 3.5MHz DC current 10.0A fan hardly ever comes on
>> 7MHz 12.0A
>> 10.1MHz 12.4A
>> 14MHz 14.3A
>> 18.1MHz 13.8A
>> 21MHz 16.4A temperature rises rapidly and fan
>> soon gets to full speed
>> 24.9MHz 13.5A
>> 29.7MHz 12.3A
>>
>> I have no results for 50MHz band as I have no transceiver covering that band.
>>
>> The readings drift as the heatsink (and the dummy load) warm up, and so I
>> took readings quickly and allowed time for the heatsink to cool down between
>> each band. I find that the KPA500 current sensor tends to give low readings.
>> I would be interested in knowing how other amplifiers perform.
>>
>> David Park G3PSV
>>
>
>
> -----
> No virus found in this message.
> Checked by AVG - www.avg.com
> Version: 10.0.1424 / Virus Database: 2112/4794 - Release Date: 02/07/12
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft@mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
--
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco/
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:Elecraft@mailman.qth.net
This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
