[comp.protocols.time.ntp] CHU observations - some confirmation?

mrapple@quack.sac.ca.us (Nick Sayer) (05/05/91)

Well, since I posted the STREAMS CHU driver, I've been using it, and
I've got some observations. First, the setup.

The CPU is a Sun 3/160 running SunOS 4.1. The radio is a Sony 2010
with a 10' wire antenna (I'd make a dipole, but I live in a condo).
The modem is (for now) an old accoustic 300 baud modem. I want to make
my own, but no one has mailed me the data on the chip (hint hint).
It's not your fault if you've mailed me. My mail/news connection has
gone sour the last week. The whole thing is in Stockton, CA. propdelay
says I'm about .03 seconds away. The CHU is at stratum 0, and there's
a stratum-ceiling driver at stratum 9. The CHU clock has the propdelay
time set and has flag1 set. I've done all the tickadj nonsense.

First, CHU itself. The signal is good on 14.670 from about noon or so
until about 9:30 PM. Then some clown starts running RTTY right on top
of it, so it's time to move to 7.335. The noise there is deafening,
but at night the signal comes up enough to get something. I have yet
to hear anything at 3.330.

I'm really, really impressed at this setup's ability to pull the AFSK
out of the mud. Signals that are so bad the voice ID is inaudible
decode perfectly. However, the offsets as reported by xntpdc seem
to be all over the place. I can't figure out many of xntpd's numbers,
but xntpdc's loop command is useful at least. It seems to think my
clock's frequency number is about -0.025 or so. The 3/160 at school
peers to 3 internet stratum 1 clocks and it's frequency is about
0.37. Using a recently posted naval observatory modem program
tells me that the clock is about -0.05 from UTC. As far as I know,
Suns have a basic clock granularity of 0.02 seconds.

So all of this tells me that I'm doing pretty well. Does this corespond
to anyone else's observations? clock.txt listed only one CHU chimer
on the internet. Is no one else doing this sort of thing?

-- 
Nick Sayer              | Think of me as a recombinant    | RIP: Mel Blanc
mrapple@quack.sac.ca.us | Simpson: Homer's looks, Lisa's  |   1908-1989
N6QQQ                   | brains, Bart's manners, and     |  May he never
209-952-5347 (Telebit)  | Maggie's appetite for TV.  --Me |  be silenced.