[dsp] PWM/phase relationship

[Scott Miller]

> My gut also tells me that the higher the sampling
> frequency, the lesser this effect will be as the
> total amount of shifting (in a time sense) will
> be lessened, but how much will that effect the
> intelligability of the voice signal 'payload'.

Not a DSP expert by any means, but I've had good luck with using PWM for
Bell 202 and PSK31/BPSK modulation.  Both used waveforms from lookup tables
that could just have easily been voice samples.  I think Bell 202 especially
is likely to be more sensitive to any phase error than human voice.

In the Bell 202 example, I'm using a PWM clock rate of 7.3728 MHz.  With 4
bits of resolution (16 step PWM) that comes out to a frequency of 460 kHz.
A simple RC filter removes the high frequency component.  I'm really not
sure about any phase shift.  The sample rate is 57.6 kHz, so in this example
the PWM register stays at the same setting for 8 periods in a row.

The BPSK code uses 8 bit resolution, so the PWM carrier is 28.8 kHz.  That's
still easy to filter for a 4 kHz audio passband.

> All of this is for a 'next gen' laser communicator--
> inspired by my first laser contacts in this last
> VHF contest. :)

Sounds cool.  Are you sending PWM with the laser, or just using that for
audio output?

I'd like to experiment with pulse position modulation.  On a high-altitude
ballooning list we were talking about the possibility of using a xenon
strobe to send position and telemetry data.  The same thing should work with
a pulsed laser.  what makes it interesting is the very high ratio of peak
power to average power.  Data rate's not very high, though.  Say the strobe
has a half-second recharge time, and you have millisecond timing accuracy.
You'd start with a reference pulse, and the window starts .5 seconds later.
The next .5 seconds are divided into 256 slots, so the slot during which the
pulse arrives determines the value of an 8-bit byte.  That gives you a data
rate of 1 byte per second.  Not good for voice, but you can get position and
altitude in 10 seconds.

Scott
N1VG