[Ground-station] Experimental Channels

[Phil Karn]

On 2/12/20 11:57, KENT BRITAIN wrote:
> NOPE!!!!!     Do some research on Class C amplifiers vs Class A linear
> amps.
>
> Kent WA5VJB
>
> Well ...... if the transmitter is only running a fraction of a watt,
> then the bent pipe would
>    be more energy efficient.
>
I'm actually not so sure. Even though I'm always talking about the
greater transmitter efficiency of single-carrier digital downlinks vs
linear transponders, overall power efficiency is not necessarily less
for a linear amplifier vs a digital downlink carrying a raw digitized
uplink. I think it depends heavily on the bandwidth. If you don't know
anything about the signals you're repeating, you'll have to sample at
least twice the bandwidth, with some number of bits per sample, and
transmit those PCM bits uncompressed even when there's no perceptible
signal on the uplink. (There might be a spread spectrum signal buried in
the noise.) In the idle situation a linear transponder could idle at a
low power level. I'd have to do the calculations.

It might be possible to vary the PCM bits per sample according to
measured uplink power. If you know the thermal noise level in the uplink
receiver, you wouldn't have to waste bits precisely sampling it.
Increase bits/sample as the uplink power and SNR increase. You could do
this automatically by setting the A/D gain so that the thermal noise is
~1 bit. Then send the PCM bits with variable length encoding (big
numbers get more bits).

But saving downlink HPA power is not the reason for all this. The
satellite's primary job would still be the retransmission of digital
data that is received, demodulated, error corrected and multiplexed onto
the downlink. The "pseudo-bent-pipe" (PBP?) mode would be the exception,
provided only for occasional experimentation with modes the satellite
doesn't support natively.

Phil