[ntp:questions] Indirect GPS time source options

William Unruh unruh at invalid.ca
Thu Mar 13 19:20:03 UTC 2014

On 2014-03-13, Olivier Drouin <ol.drouin at gmail.com> wrote:
> Thank you Greg, Jochen, William,
>                                Great answers.
>                                Diversity is really what I'm looking for and
> I dont really need microsecond accuracy.
>                                CDMA seemed to be the most 'convenient' way
> to add diversity without a need for a working WAN link and/or line of sight
> view to the sky.
>                                Also, I already know that I get good cell
> signal from inside the server room.
>                                The thing with CDMA is that it looks like
> it'll not be around for many years and I haven't seen any equipement for
> 4g, hspa, LTE, etc...
>                                For your information, I think I can get away
> with a 10K budget but I need to be absolutely sure It'll work beforehand.

A gps will cost you $50 assuming you have a computer already (if not
another $50 for a Raspberry Pi +power source and ethernet cable. )
It will also cost a bit of wiring up time. 

I do agree that gps will almost certainly require an unalumized  window, and in
Quebec, not a north facing window. (gps avoids the north).

 Ie, you can try out the gps option with spare change. If that does not
 work, then you can go to the more expensive CDMA option. I have no idea
 how long Bell and Telus will continue their cdma offering. Rogers never
 used it, and all the newer players do not use G2 type offerings.
>                                I'll talk with the facility owner and do
> some tests with handheld GPS so I can verify what kind of signal I can get
> (and where). I'm a bit worried because I dont have roof access whenver I
> want and I dont know how the antenna will behave with the snow we get
> around here (quebec city).
If it is on the roof, put the antenna onto
 a pole on the roof high enough that it will not get covered. But on a
 window, the chances are pretty slim that it will get covered in snow. 

>                                Ill also verify if I can synthonize the CHU
> radio frequency inside the server room if I can get my hand on a scanner.
>                                Olivier Drouin.
> On Thu, Mar 13, 2014 at 1:26 PM, Dowd, Greg <Greg.Dowd at microsemi.com> wrote:
>> You are not incorrect but I think that "indirect GPS time" may be more of
>> a marketing term.  Basically, GPS is the predominant method of time
>> dissemination.  When you think of a commercial ntp server, really you just
>> moved the GPS antenna from your PC (getting rid of the receiver and coax
>> cable) over to the ntp server, and then distributed the timing information
>> over a different cable (your ethernet cable).  Since NTP is layer 3, and
>> the bus isn't dedicated, you introduce error but it's still GPS time.
>> My term for the concept I understand you to be discussing is timing
>> diversity.  I prefer to have a number of different methods of accessing
>> time data.  GPS is the baseline but access to the L1 GPS signal is limited
>> in some locations, particularly indoors.  Keep in mind the analogy that the
>> GPS broadcast is like a light bulb shining from 22000 km away.  CDMA is
>> nice as it's a powerful signal (compared to GPS) but it is limited in
>> geographic deployment.  Most 3G/LTE cellular systems these days are tightly
>> syntonized (frequency) but not synchronized (time).  So, no, you won't get
>> microsecond level timing from GSM.  Next gen LTE has a profile for LTE-TDD
>> which will require about the same level of synchronization as CDMA.
>> If you are looking to add diversity to an installation, your first best
>> bet is to move to a GNSS receiver.  Like your smartphone, a GNSS receiver
>> typically works with both GPS and the Russian Glonass satellite systems.
>>  There's some diversity.
>> For Canada, you could also look at eLoran.  That operates on a completely
>> different frequency plan to satellite timing.  I think Canada has CHU radio
>> timing as well (3330kHz).  There are a number of signals of opportunity
>> that require custom hardware to access either time or frequency.  You
>> mentioned CDMA.  Digital television is another example.
>> As someone mentioned earlier, to get a more precise answer you have to
>> provide a more detailed analysis of your requirements and budget.  Anything
>> is possible.  There are neutron stars out there dieing to provide you with
>> a frequency reference :-)
>> -----Original Message-----
>> From: questions-bounces+gdowd=symmetricom.com at lists.ntp.org [mailto:
>> questions-bounces+gdowd=symmetricom.com at lists.ntp.org] On Behalf Of
>> Olivier Drouin
>> Sent: Thursday, March 13, 2014 7:29 AM
>> To: William Unruh
>> Cc: questions at lists.ntp.org
>> Subject: Re: [ntp:questions] Indirect GPS time source options
>> On Wed, Mar 12, 2014 at 3:59 PM, William Unruh <unruh at invalid.ca> wrote:
>> > On 2014-03-12, Olivier Drouin <ol.drouin at gmail.com> wrote:
>> > > Hello,
>> > >       I'm located in Canada and I'm trying to find indirect GPS time
>> > sources.  I like the idea of using CDMA as a time source but it seems
>> > that this technology will not be around for a lot longer.
>> >
>> > What is an "indirect gps timing source"? How does that differ from
>> > using ntp to go to a level 1 server which gets its time from gps.
>> > While cdma uses accurate time, it is not really designed to deliver
>> accurate time.
>> > GPS is.
>> >
>> Well, from what I understand the cell networks are closely synchronized
>> against the GPS constellation. So, synchronizing against the cellular
>> network is called "indirect gps", at least according to the endrun website
>> (I haven't seen this nomenclature anywhere else).
>> It is different of GPS because if your WAN links are down, which can
>> happen once every few years,  then, hopefully, you can still reach the
>> cellular network in your neighborhood or else its the end of the world or
>> something...
>> Also, from what I understand and please correct me if I'm wrong but
>> cellular networks are indeed designed to be accurately timed because it's
>> needed for the normal operations of the cell network.
>> Thank you for your reply, best regards,
>> Olivier Drouin
>> >
>> > >       For exemple, Endrun Technologies sells the Sonoma N12 which
>> > > seems
>> > to allow time synchronization against CDMA signal without the need for
>> > a carrier subscriptions.
>> > >
>> >
>> > And a gps receiver with PPS output is probably 1/100 the price.
>> >
>> > >
>> > >        Is the CDMA method still available with these other cellulare
>> > network technologies like HSPA or LTE? If not, do I have other options
>> > beside a GPS antenna with a sky view?
>> >
>> > CDMA also needs an antenna with cell tower view. What's the difference?
>> > (Note my gps on an east facing window through a large cedar tree works
>> > just fine.)
>> >
>> >
>> As Terje is mentioning, the cell antenna doesnt need a line of sight to a
>> cell antenna, so It can be installed inside the server room and you dont
>> have to deal with the maintenance problems (lots of snow around here) of a
>> GPS antenna on the roof...
>> > >
>> > >
>> > >
>> > >        Thank you and best regards,
>> > >        Olivier Drouin.
>> >
>> > _______________________________________________
>> > questions mailing list
>> > questions at lists.ntp.org
>> > http://lists.ntp.org/listinfo/questions
>> >
>> Thank you and best regards,
>> Olivier Drouin.
>> _______________________________________________
>> questions mailing list
>> questions at lists.ntp.org
>> http://lists.ntp.org/listinfo/questions

More information about the questions mailing list