[time-nuts] Re: What GNSS module to buy for a good time reference?

[John Ackermann N8UR]

On 6/21/23 13:11, Erik Kaashoek via time-nuts wrote:
> John, thanks
> 
> Could you elaborate a bit on these points:
> 
> -- L1/L5 bands only (no L2 option)
> I read the amount of satellites transmitting L5 is a bit lower compared to
> L2, are there other reasons why L2 is important? I guess one would have to
> choose between L2 or L5 anyway in configuring a F9T, anything else?

In the future, L5 will be a solid choice for the second frequency, and 
it's fine now if you want only real-time results.  But the NRCan 
post-processing service currently does not use L5 in its corrections so 
you won't be able to upload RINEX files to get back precise location 
information or clock offsets.  At some point, I'm sure they will process 
L5 but that's not the case yet.  (Last time I checked they ignored L5 
data included in an L1/L2/L5 RINEX, and I haven't been able to find 
anything on the web indicating that they've started.)

> -- only one TIMEPULSE output (could be a very big deal)
> I failed to understand why this may be a big deal, can you explain?

One of the neat things about the prior u-blox timing receivers (and some 
of the non-timing ones as well) is that they have two independent 
timepulse outputs that can be programmed to RF frequencies (as high as 
25 MHz for some).  You can use one output for PPS, and the other to 
generate an RF signal that can be used, with appropriate cleanup as the 
phase noise/jitter is horrible, as a frequency reference.  There are a 
couple of modern GPSDO designs that take advantage of this, using the 
second TP output to drive a Silicon Labs jitter attenuator/synthesizer chip.

Without the second TP signal, you can do PPS or RF but not both.

John