[Elecraft] K3 SSB average power output

[Larry Phipps]

I haven't checked the K3 speech processor accuracy yet, but will get to 
it shortly. I did run a power check with carrier though. I calibrated my 
K3 internal meter against my LP-100, and verified it against my HP 438A 
power meter. The calibration was done according to the manual at 5W and 
50W on 20m. Here are my results for requested powers of 5W, 10W, 50W and 
100W for three bands, plus 120W on 20m...

160m  5.1  9.7  51.0  95.3
20m   5.0  9.5  50.0  94.1   110.8 (120W)
10m   4.8  9.1  48.5  91.0

It looks like the coupler in the K3 has a little falloff in response 
with rising frequency, and a little non-linearity.  This would be 
expected. When the dust settles, they could probably add a little code 
to compensate for this in the firmware. That's the beauty of firmware ;-)

Mine may not be fully representative, since I had a slight amount of 
damage to a T/R switch component on my KPA3 (self inflicted during 
installation. I substituted a part, but am awaiting the correct one). 
But since the ratios below 12W and above 12W seem to track, I think 
these are probably typical (12W is the switching point between low power 
final and high power final for those who don't know). BTW, my output on 
20m tracks yours pretty well.

73,
Larry N8LP
>
> Message: 4
> Date: Mon, 04 Feb 2008 19:45:10 -0500
> From: Bill Tippett <btippett at alum.mit.edu>
> Subject: [Elecraft] K3 SSB average power output
> To: elecraft at mailman.qth.net
> Message-ID: <200802050045.m150jCHW025267 at smtp4.mail.ctc.net>
> Content-Type: text/plain; charset="us-ascii"; format=flowed
>
>          N8LP's LP-100 has a function for evaluating
> Peak to Average performance of speech processors.  SSB
> isn/t of much interest to me until 10 meters comes back
> so I haven't played with this mode in the LP-100.  Larry
> has a K3 on order and maybe he will compare the K3
> speech processor to others when he gets his K3.
>
>                                  73,  Bill  W4ZV
>
> See LP-100 at http://telepostinc.com/
>
> * Peak to Average ratio screen. Allows accurate checking
> and setting of compression in your speech processor.
>
> Peak-to-Average Mode
> This mode lets you determine the average power in a signal by taking 
> 40,000 samples/second, and compares this to the peak power in
> the signal. The result is displayed as a ratio in dB. I provide a 
> couple test tones, which are available on my website at
> http://www.telepostinc.com/Files/two-level-tone-loop3.zip and 
> http://www.telepostinc.com/Files/loud_tone.zip. The loud tone is used to
> set the maximum power and proper ALC range (with processing OFF). The 
> two-level tone is used to determine the peak-to-average
> ratio of the output signal. It can be played back on your PC, or 
> converted to mp3 and played on a portable player. It can be played over
> a speaker into the microphone, or directly into the mic input. The 
> two-level tone provides alternating loud and soft tones with 20 dB
> difference in level. This tone should provide the following 
> Peak-to-Average ratios vs. effective compression ratio.
> Compression Ratio Peak-to-Average Ratio
> 0dB 5.2dB
> 5dB 5.6dB
> 10dB 3.6dB
> 15dB 2.2dB
> 20dB 0dB
> I plan more test tones with different characteristics in the future, 
> which is why I decided to keep the display as Peak-to-Average as
> opposed to Compression, which would only be accurate with one test 
> signal. I will provide additional tables such as the one above with
> the additional test signals