[time-nuts] eLORAN will be on the air GRI 99600

[Bob kb8tq]

Hi


> On Aug 6, 2020, at 9:38 PM, Forrest Christian (List Account) <lists at packetflux.com> wrote:
> 
> I probably need to clarify where I'm coming from in relation to my previous
> message.
> 
> I have a fair bit of background in dealing with using GPS clock sources for
> synchronization at communication sites.   Many of these sites are lucky to
> have a rubidium oscillator in place for holdover, although some do (usually
> the larger ones).  And we're talking about thousands and thousands and
> thousands of these sites.
> 
> Where I was coming from is that at these sites, GPS can be a challenge -
> it's in a narrow band, very low signal, and, at least from my worldview, it
> seems like GPS interference is becoming more prevalent instead of less.
> As a result, whether terrestrial or via satellite, it would be nice to
> have a second alternative in a different band where one could obtain a
> reasonably aligned 1PPS (+-1uSish).
> 
> I'm hopeful that eLORAN will result in this result.   I'm also curious
> about the GPS L5 signal once it becomes operational since that is far
> enough away frequency wise that one would hope that it wouldn't be affected
> by the same interference source.

The problem is that you have a moving target. As system speeds have gone up,
so have timing requirements. 10 us used to be the “drop dead” number for a 
cell system. Now it’s more likely to be 1 us (or less …). At 1 us, you need a 
timing reference that is sub-100 ns, pretty much all the time. Getting that from a
single terrestrial broadcast source is unlikely. 

The “simple" answer is the one already being used. Send the timing over fiber. 
Yes, there are problems with various implementations. They can be worked
out a *lot* easier than putting up a broadcast (and reception) system from 
scratch.

Bob



> 
> On Thu, Aug 6, 2020 at 6:19 PM jimlux <jimlux at earthlink.net> wrote:
> 
>> On 8/6/20 4:28 PM, Forrest Christian (List Account) wrote:
>>> If you look at generally-available GNSS PNT solutions, you'll find a few
>>> failure modes:
>>> 
>>> 1) Loss of a satellite (or two).   This is why the constellations have
>> more
>>> satellites than is strictly necessary, so not a big deal.
>>> 
>>> 2) Loss of control/failure in the control system/constellation wide
>>> software failure, aka the recent Galileo failure.   This is why you have
>>> multiple GNSS constellations.
>>> 
>>> 3) Ground based interference (jamming, spoofing), etc.    This is why you
>>> need a terrestrial backup, which doesn't really exist.
>>> 
>>> For timing, I wouldn't be opposed to someone flying (or adding a payload
>>> to) a couple of geostationary satellites which live in a separate band
>> from
>>> GNSS.  It would be interesting to be able to put up a small satellite
>> dish
>>> and get a highly reliable and hard to interfere with timing alternative
>> to
>>> GNSS.    I know there are two way time transfer options out there, I'm
>> more
>>> thinking basically a fixed-location cesium clock in the sky.
>>> 
>> 
>> Well, the GPS folks found that a Rb works better than a Cs, and both
>> need ground monitoring and updating.
>> 
>> One could rent a Single Channel Per Carrier transponder slot on a GEO
>> satellite, feed a carrier derived from your ensemble of clocks to the
>> uplink, and there you go.
>> 
>> My google-fu is failing and I can't find even a rough cost for such a
>> service.  Maybe something like $50k/month? $600k/year.
>> 
>> That 600k would probably buy you *one* space qualified Rb oscillator of
>> "good enough" performance, maybe.   USOs like used on GRAIL were
>> $1M/each sort of items in qty 4. And then you need some TWTAs to amplify
>> the signals for transmission - those are also pretty pricey.
>> 
>> And then the launch cost.. I happen to know (because I just bought it at
>> work) that you can push about 100-120kg to a GEO+1000km orbit for right
>> around $10M.  Rocketlabs might do 20-30kg to a similar orbit for half that.
>> 
>> 
>> All in all, I suspect that there are better uses of the $50M it would
>> cost - You could buy a LOT of Cesium clocks for terrestrial use.
>> 
>> If you're willing to do CSAC level performance, and you're willing to
>> have it be in LEO, so you see it a couple times a day for 20 minutes,
>> and you're willing to do some design and fab in your garage - under $5M.
>> 
>> Look up CHOMPTT  - CSAC and optical links.
>> 
>> 
>> https://digitalcommons.usu.edu/cgi/viewcontent.cgi?article=4407&context=smallsat
>> 
>> 
>> 
>> 
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>> 
> 
> 
> -- 
> - Forrest
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.