[time-nuts] time sync by moonbounce
EB4APL
eb4apl at gmail.com
Mon May 25 14:46:07 UTC 2020
Hi Jim,
This explains a lot of things. As an old time user of the Moon Bounce
time synchronization in the Madrid Space Complex, I can say that the
system was never popular among the users, it was cumbersome to use and
there were already other systems with better accuracy and availability,
such as VLBI and Loran-C. We used Loran-C as our daily reference
because we had a station at a distance of 650 Km and the signal was very
good, even with an old receiver who needed manual estimation of the
delay the results were very good. Maybe the aliens who designed the
system were not very aware of the humans idiosyncrasies
I think that the idea was good, but the implementation was not so good.
I can summarize the "peculiarities" as I remember them:
- The transmissions had to be scheduled for a period of common moon
view. The receiving station was manned 7/24 but the transmitting station
(DSS13, if I remember it correctly) had to be manned specially for the
event by a crew probably from Barstow, some 40 miles away.
- Normally there were not provisions for voice communications between
the two end points, so if we did not got correlations we didn't know if
the transmitting station was working ok or even if it was not manned due
to some problem.
- There were not monitoring aids in the receiver, so if we did not have
correlations, we has to climb to the roof, verify the pointing of the
antenna (there were a rifle scope for that), and if it was, tried to
contact DSS13 by phone and pray.
- As you can figure it out, the antenna was not remotely controlled. One
had to climb to the roof, set the moon declination for that day using a
handwheel, slew the hour angle with a switch and select the hour angle
rate with another switch. The moon position and rate was obtained from a
nautical almanac and when the sky was clear we used the above mentioned
rifle scope.
The receiver was quite dumb, all intelligence was on the transmitter
site. The TX equipment generated a PN code that lasted about 1s, and a
full observation cycle lasted about a minute, I don't remember the exact
figures. The code was sent advanced to take care of the round trip light
time but an additional time bias of 30 us was also introduced which was
the basis of the measurement. The bias was decreased 1us/s, so
theoretically it was received just on time in the second 30.
The receiver generated the same PN using the station timing reference
(from a HP 5065A Rb) and it was correlated with the received code. The
output of the correlator was integrated and sent to a strip chart
recorder. The graph consisted in one trace with a quite noisy ramp and
the other trace with 1PPS from the station reference. Now the weird
thing: after finishing an observation, we put the graph in a desk and
using a drafting rule we draw a straight line that tried to be the "best
fit" to the noisy ramp. Were the line crossed the zero we read the PPS
mark there and counting back to the start of the minute we got the PPS
offset with respect to the transmitting station. Of course we averaged
the values obtained from several minutes, after discarding the noisiest
ones.
Since we didn't had any faith in the system we didn't tried to suggest
improvements or improve it ourselves. A good one could be to use a
computer program to perform the best fit analytically, but this would
mean to type the hundreds of points manually from the graph and we never
tried this. There were not a digital version of the output, we also
could use a digital voltmeter for acquiring it. We suggested or
implemented a lot of improvements to other operational things, but this
contraption was felt as a dead horse from the beginning and its
operational life was short. Later it was replaced with a GPS based one:
2 full height racks filled with equipment and an antenna made from a 10
or 20 gallon hermetic paint drum for housing the front end electronics,
topped with a fiberglass radome about 1 1/2 ' in diameter. It was
painted white, but the cylinder origin was discovered during a
maintenance. It was a beautiful prototype that worked very well during
its shot life.
I don't keep pictures of this equipment, but I have one with the same
antenna used for other purposes.
Sorry for the bandwidth, but the thread brought me old memories.
Best regards,
Ignacio
El 23/05/2020 a las 19:17, jimlux escribió:
> On 5/23/20 9:18 AM, Mike Millen wrote:
>> Probably a good idea... there are two page 19s and no page 20 in the
>> pdf. :-(
>>
>
>
> That's the page where the aliens came and told us how to build the
> DSN, then the story resumes with 26m antenna design and operation.
>
> (If anyone's interested, I can probably ask the librarians to find it
> at JPL - correcting the pdf/microfilm is probably beyond scope)
>
>
--
El software de antivirus Avast ha analizado este correo electrónico en busca de virus.
https://www.avast.com/antivirus
More information about the Time-nuts_lists.febo.com
mailing list