On 04/30/2013 06:49 PM, Adam Jacobs wrote:

Could you explain a bit more about how PSK's average power is 41% of

the peak? If you have a link that would be great.

I use Fldigi which has a "Tune" button which puts out a single tone. I

would set my output power to say 25W and then go to idles and the power

shown on my tuner would drop considerably. Why?

If you look at the PSK signal coming out of a sound card it is amplitude

modulated so that the phase transitions occur when the amplitude is

zero. This reduces bandwidth. When sending idles the phase changes

every bit period. The modulating function is a Cosine function. Fldigi

allows you to look at the waveform instead of the waterfall.

If we set the peak amplitude at 1 volt and evaluate the integral of Cos

(x) from 0 to PI/2 radians and divide by PI/2 we get the average voltage

of .6366 volts.

A non modulated signal of the same (peak) amplitude would have an

average voltage of 1 volt.

Squaring .6366 gives .4053 for the average power.

Yes, I spent quite a bit of time with the dummy load and my MFJ

SWR/Power meter verifying that when switching from a single tone to PSK

idles the average power drops to 41%. If it doesn't then the rig is

being over-driven and putting out a wide signal. Increasing the drive

to the point where ALC action occurs just means that the extra power is

going into splatter and not useful signal.

The 41% figure only holds for idles. During text communication there

are consecutive bits with the same phase so the signal does not go to

zero amplitude between bits and the average power bounces up higher.

I have been told that when reporting power one should report peak power

rather than average power but I have not seen it spelled out in any

contest rules for QRP status. Perhaps it should be.

Larry, KG4Q