Hello Mike & group,
With Mike's suggestion to cut the signal for a short time during the phase
tranisitions the cure might be worse than the illness :
WOLF is (currently) used at 10BPS, so you would need to cut the signal
(worst case) 10 times a second for a short time (few milliseconds). This
will create a lot of sidebands, the signal might be worse than with 'rude
BPSK'.
Jim's signal is remarkable clean, as could be expected from a PSK-like
signal. The 2 'disadvantages' (in my opinion) are :
- 'high power' electronics involved
- giving away 3dB (by the envelope AM), versus 'rude BPSK'
An alternative way to get a 'clean' signal might be to have a smooth phase
transition. Instead of a 180 degrees instant phase jump one could have a
gradual phase change, let's say over 10 milliseconds. At 10BPS this would
leave the signal for 90% of the time in the 'correct phase'.
Advantages would be :
- additional electronics would be at the 'low power end'
- any PA (wether linear or not) could be used without any changes
- signals would be only 0.54dB down versus 'rude BPSK'
Depending on the amount of sidebands that is acceptable the transision
period could be changed. A shorter transisition period would give a better
SNR, but more sidebands while a longer transistion period would give a
cleaner signal but less SNR.
There would be different ways to implement the smooth phase shifting, going
from the classic phase modulator to digital techniques.
But before I start any project in that direction I would like the opinion
of the experts, maybe I have just overseen some facts that make my whole
suggestion worthless ...
73, Rik ON7YD
|
