[ntp:questions] Re: Where to obtain audioio.h

David L. Mills mills at udel.edu
Tue Sep 9 02:56:49 UTC 2003


Tim,

I assume the patch went via bugs at ntp.org and was a rename for conflict.
Changing the demodulation and decoding algorithms is very scary, as they
are theoretically optimal.

The 0.5-ms driver is nice, but the Sunses pre-8 were in the
single-sample (125-us) range. Consider how a mixer works with
multiple-rate sampling. If the input sample rate is different than the
output sample rate, the two rates will heterodyne giving nulls at the
least common multiple rate. At least I think that is happening in the
Solaris drivers.

As for the Pacific ions, I haven't been tracking lately; looks like I
should have. Usually, WWVH comes up briefly as the Sun crosses the
Pacific and the MUF is east-limited. However, the solar flux is cooling
down and the MUFs are already low. In a few years even 2.5 MHz may be
viable.

Your suggestion using L/R channels for frequency switch is neat. It
shouldn't be too hard to do. Put a zit in the qsy() routine and fiddle
the audio parameters. Consider using TWO frequency agile receivers for
diversity! The 100-Hz subcarrier demodulator is synchronous and should
work well with two receivers and two antennas. Square-law combine
(theoretically optimum). Wow, I'm getting carried away...

Dave

Tim Shoppa wrote:
.
> There is a special branch of the ALSA Linux sound drivers for low
> latency.  It promises latencies in the 0.5msec region.  I don't
> understand how the mixer can possibly influence this, but maybe
> I'm just ignorant.  Maybe the mixer works differently under Solaris?
> 
> Amazingly, the past couple of mornings I've been picking up WWVH
> on 10MHz with great clarity.  I'm on the East Coast of the US so in
> the evenings 10MHz is a deep swamp.  Like Steven I have the $40 Ramsey
> 10 MHz receiver and a couple of other cheap-o portable SW receivers
> that can pick up 5, 15, and 20, but not frequency agile.  Maybe I can
> feed 10MHz on the left and a different frequency on the right... gotta
> get more isolation transformers :-)
> 
> Tim.



More information about the questions mailing list