After some detailed measuremetns on eight GPS receivers from different
manufacturers, I've found that the modern ones mostly exhibit around 100ns
jitter from second to second. Older units like the Garmin GPS25 and
Motorola Oncore (old favourites among the amateur community) give a few
hundred ns. But by averaging over longer periods, the jitter from these is
just as good as the later ones when averaged over the same period.
This pulse to pulse jitter can often be reduced to a few tens of ns on some
modules by operating in position fix mode, but you need to have a module
specifically optimised for timing purposes in order to do this; these
usually cost somewhat more than the standard navigation version.
So far my favourite is the Jupiter T with its 10kHz output, making locking
of an oscillator straightforward, as described earlier on this reflector. I
tried a Jupiter module directly locking a 10GHZ local oscillator, but the
raw signal is not good enough when effectively multiplied by one million -
teh tone of a CW signal sounded much too rough. For high microwaves a long
PLL time constant with a high spec OCXO is required. But for use up to UHF,
the simple GPS disciplined oscillator described earlier is satisfactory.
Andy G4JNT
-----Original Message-----
From: José Manuel [mailto:[email protected]]
Sent: 2004 March 10 10:11
To: [email protected]
Subject: LF: Re: Ionospheric doppler ?
Hi Alberto ad all:
When you install the GPS I think that you´ll probably find short-term
variations in the same order, + - 100 nsec.
73 de José, EA1PX
----- Original Message -----
From: "Alberto di Bene" <[email protected]>
To: "LF Mailing List" <[email protected]>
Sent: Tuesday, March 09, 2004 12:57 PM
Subject: LF: Ionospheric doppler ?
Hello Group,
I made yesterday an interesting experiment and would like to know your
opinions about it.
Waiting for the weather to become such to allow me to go on my roof
to install there a GPS antenna, in the meantime I started to play with an
inexpensive radio-controlled clock, made by Conrad, bought a few years
ago at the Friedrichshafen Messe in Germany, which receives the DCF-77
signal.
This clock has an output meant to drive an external electro-mechanical
hand clock, and on this output there is, of course, an 1pps pulse.
I have an HP-5328B Counter, with a 10811 OCXO which is always (24/7) on.
My shack is in the basement, with a constant temperature of 21 Celsius,
no drafts,
so any variations in the measured frequency or time is real, and not an
artifact
of the counter.
The 5328 has a sort of reciprocal counting feature, where you can use an
external signal as a gate for an internal 100 MHz oscillator, phase
locked to
the OCXO. In addition you can prescale the external signal.
So what I did was to prescale by ten the 1pps signal from the clock,
then used
this 10 second interval to count the internal 100 MHz oscillator, giving
a resolution
of 1 ns. If everything were perfect, I should have obtained a count of
exactly 10^9.
What I measured was a value that differed from the ideal by an amount
slowly
changing with time, ranging from -80 ns to + 120 ns. The count was very
consistent from period to period, showing no short term random jitter.
In one case I measured a variation of about 100 ns in a time lapse of
roughly
one hour.
I am by no means an expert in propagations and ionospheric effects, so
my question
is : are the values I measured compatible with what is known about
ionospheric doppler ?
If not, what else could be an explanation of that slow change ? I would
tend to exclude,
for the reasons reported above, an artifact of the HP counter.
Thanks for any explanations
73 Alberto I2PHD
"This e-mail is intended for the recipient only. If you are not the
intended recipient you must not use, disclose, distribute, copy, print,
or rely upon this e-mail. If an addressing or transmission error has
misdirected this e-mail, please notify the author by replying to this e-mail."
"Recipients should note that all e-mail traffic on MOD systems is
subject to monitoring and auditing."
|
