[GPS_Standard] GPS Modules with 1PPS output

[Bert, VE2ZAZ]

My understanding is that, eventhough most navigation-grade GPS are specified
to have their 1PPS accurate to 1 microsecond, their actual accuracy is more
like 100-200 nanoseconds, hence the order of magnitude I was refering to
comared to the UT+.

Cheers,

Bert, VE2ZAZ

----- Original Message -----
From: "Graham Haddock" <GrahamH at austin.rr.com>
To: "Zack Widup" <w9sz at prairienet.org>
Cc: "Bert, VE2ZAZ" <ve2zaz at sympatico.ca>; <gps_standard at mailman.qth.net>
Sent: Sunday, January 27, 2008 12:45 PM
Subject: Re: [GPS_Standard] GPS Modules with 1PPS output


As Bert says, his and most of the GPS "disciplined" oscillators average
the timing from the 1PPS pulses over a long time interval, that would
average out the error from pulse jitter, if you average for long
enough.

The jitter spec on the Jupiter 12 is one microsecond.

That is ten times the resolution of the 10 MHz standard, if you
are counting pulses. (One pulse every 100 nano seconds.)
So, if you are time averaging for long term accuracy, the
jitter dominates the timing calculations, rather than the counter accuracy,
typically assumed to be +/- one count.

The jitter spec on the Timing version of the Oncore is 20 nano seconds
and the receiver will provide you a message, should you want to use it,
to correct that jitter to be as low as 2 nanoseconds.  (See the paper
on VLBI on  http://www.gpstime.com/   on how they can do this.)

If you are going to "count pulses" for determining long term accuracy
and management of an oscillator, then you should get the same accuracy
in one tenth the time with the Oncore Timing version, versus the Jupiter 12.

As an experimenter, you may not care if you have to wait a little longer.

The pulse counting approach that Bert has taken is a simple and very
understandable system. My complements to Bert.

So, the timing version of the Oncore has "pedigreed" pulses, and is the
receiver being used for discipline in the VLBI system.

And, as Bert points out, the rest of your system may not be rigorous
enough to take advantage of the additional accuracy of the Oncore.

--- Graham / KE9H

==

Zack Widup wrote:
>
> Thanks Bert!
>
> At $50 apiece, I guess it's worth getting one and experimenting with
> it. It might be a while before I do that, but I'll post the results here.
>
> 73, Zack W9SZ
>
> On Sun, 27 Jan 2008, Bert, VE2ZAZ wrote:
>
>> Hi Zach,
>>
>>> Are these units better for our purposes than a Jupiter 12 board?
>> The instinctive answer would be yes. The Jupiter 12 board is a
>> navigation-grade GPS and the Oncore UT+ is of timing grade. Timing-grade
>> receivers "typically" deliver an accuracy one order of magnitude
>> better than
>> navigation-grade. BUT, it all depens on whether the system's overall
>> accuracy and stability will be enhanced by a better GPS. I can tell
>> you that
>> to obtain a range of accuracy typically in the low 10^-10's, an
>> automotive-grade GPS is fine on my system.
>>
>> I have not personally tried better-grade GPS receivers though. In the
>> end,
>> since we average out the GPS inaccuracies over a long period (hours),
>> the
>> benefit may not be so obvious on my system.  It might be beneficial,
>> especially for shorter sampling periods. Worst case, you would not
>> see much
>> improvement. On better (commercial-grade, more complex, better designed)
>> systems, it would definitely make a difference.
>>
>> In the end, everything counts. But your whole system has to be solid,
>> including OCXO, thermal stability, DC supply stabilit, etc.
>>
>> Cheers,
>>
>> Bert, VE2ZAZ
>>