[time-nuts] M12+T First Light
dga+ at cs.cmu.edu
Tue Apr 19 02:19:55 EDT 2005
Thanks - very informative graph! Might be interesting to see what a
GPSDO looked like with an M12+ core that used sawtooth correction.
Anyone have such a beast / such a plot?
I still need to find free time to build one of those. I love my
z3801a, but I keep thinking, "Gosh, this could be a lot better!" :)
On Apr 19, 2005, at 1:17 AM, Tom Van Baak wrote:
> Hi Brooke,
> Yes, the M12 timing receiver is very nice.
> Attached is a preliminary plot comparing 1 PPS jitter
> among four common GPS time recievers including a
> M12+ with sawtooth correction.
> In the plot the VP is dark blue and the M12+ is pink.
> In the table below, times are RMS over 10k seconds.
> 31.5 ns -- Oncore VP
> 9.3 ns -- CNS-1 (M12+)
> 3.0 ns -- CNS-2/O1 (M12+ h/w sawtooth correction)
> 1.3 ns -- standard HP 58503B (8ch VP, 10811 GPSDO)
> The M12+ with sawtooth correction is yellow and
> the GPSDO is green. You can see that although
> the M12+ with sawtooth correction has slightly
> more jitter than the GPSDO, as a time standard
> on average it looks like it might be more accurate
> more of the time.
> As you know from playing with a 58503A or Z3801A
> a GPSDO does a fair amount of wandering and you
> can see this in the plot.
> If you require a real-time 10 MHz or 1 PPS then use
> a GPSDO but if you're collecting data only for later
> processing then it looks like an M12+ with sawtooth
> correction is just as good or better.
> ----- Original Message -----
> From: "Brooke Clarke" <brooke at pacific.net>
> To: "Discussion of precise time and frequency measurement"
> <time-nuts at febo.com>
> Sent: Friday, April 15, 2005 16:23
> Subject: [time-nuts] M12+T First Light
>> To use an astronomical term "First Light" meaning my first use of the
> M12+T GPS Timing receiver. WOW!!!!
>> The manual description of the @@Hn command shows -128 to +127 ns as
> range for the sawtooth error. On the 8 channel unit I think the actual
> sawtooth error was around + and - 50 ns.
>> I looked at the @@Hn data stream in TAC32 and saw numbers in the mid
> teens, much lower than the 8 chan sawtooth error. I'm not sure if
> this is
> typical, or maybe I was just lucky. I have the El mask set to 50
> and Time RAIM turned off.
>> I've made a small number of measurements against the Cesium source
>> whose C
> field I haven't touched for a month or two and the standard deviation
> 5,000 1 PPS measurements ranges from 14 to 27 ns, whereas the 8
> channel unit
> was typically in the mid 30 ns range with excursions much higher.
>> So the way timing is being handled in the M12+T is far better than
>> the old
> 8 channel units, i.e. it's not just more channels.
>> Having Fun,
>> Brooke Clarke, N6GCE
> time-nuts mailing list
> time-nuts at febo.com
More information about the time-nuts