[Elecraft] k3/psk on laptop

[Jim Brown]

On Wed, 11 Jun 2008 06:29:58 -0700, Ed Muns wrote:

>What do you think the "limiting factor" is?

I suspect that much of it is poor level matching. 

One thing I can measure with a great deal of accuracy is the audio 
distortion at the output of a sound card. This is quite important 
for narrowband modes like PSK, because intermod is another signal 
that interferes with others on the band!  Several years ago, I 
measured the distortion at the output of my laptop sound card. It 
was 30 dB down from the carrier at maximum output, but 40 dB down 
from the carrier if I reduced output the level by 6 dB. That is, 
my signal was 10 dB cleaner simply by running the sound card 6 dB 
down from max output!   

More distortion will be created if the audio input of the radio is 
overdriven, AND if the RF stage is overdriven. In both cases, 
reasonable care with level matching solves the problem.  

As Simon and Joe have noted, noise and distortion in the sound 
card can degrade decoding, but again, proper level matching can 
improve that situation a LOT. Good operating practice for audio 
systems is to maximize the output of the driving stage and reduce 
the input gain of the receiving stage. That is, the output of the 
radio should be set so that it is operating just below the level 
where its distortion begins to rise, and the input gain of the 
computer should be turned down so that it is not overloaded. 
Sometimes, especially if the computer has only a mic input, a 
simple resistive voltage divider (pad) is needed between the radio 
and the computer. That's two resistors, one in series, another in 
parallel with the input. Guidelines for choosing values are in the 
tutorials I cited in a previous post. 

In summary, the noise and distortion problems Simon has cited can 
usually be solved, and can nearly always be greatly improved, 
simply by proper level matching. Of course, that won't be true if 
the sound card is junk. :)

73,

Jim Brown K9YC