[time-nuts] Lost GPS lock or 1PPS recently?

[Bob kb8tq]

Hi

> On Sep 8, 2018, at 1:53 PM, Scott McGrath <scmcgrath at gmail.com> wrote:
> 
> This is precisely the scenario even a short GPS blackout of 1-2 weeks would cause.   Its not that GPS is not the finest time transfer system ever devised.   Its that with the loss of legacy systems we’ve lost the ability to degrade gracefully.

How will the presence of a “backup” system that in no way at all interconnects with a cell tower have any 
effect at all on it’s performance with loss of GPS ??? The legacy system (which does not even exist) 
has *zero* impact in this case. 

> 
> With a eLORAN system cell networks during a prolonged outage would probably degrade to 3G,

Except that they have no way at all to do that. They simply are not designed that way and can not function 
that way. The systems that *could* function at lower timing tolerance all got scrapped out long ago. That 
of course *assumes* that eLoran can deliver < 100 ns timing. Is there any data to support that ? 

Do any members have data on the 1 second timing out of eLoran? We’re about as deep into it as any
group not working for the eLoran people. If we haven’t seen any performance data …. I kinda doubt 
anybody else has. 


>  but they would still be up.   No you cant stream HD video or play GTA Online X,  but talk,text email and Facebook would still work Time transfer for most applications would still work.  The HFT boyos on Wall St would be SOL.  Not sure how to evaluate that eventuality.
> 
> 
> People like US need to educate political and business leadership on the need for BACKUPS to GPS

The reason the systems are designed just with GPS is *not* because “nobody mentioned the need
for a backup”. There are indeed people out there who spent a lot of time talking about this with the
guys who designed and spec’d these systems. The very simple answer *always* came back:

1) There is no alternative out there

2) There is not FCC requirement to do so = cost is not justified

3) There is absolutely zero demonstrated need

That’s very much three strikes you’re out.

Bob


> mainly because things like the Carrington Event have happened before and WILL happen again. 
> 
>   And having terrestrial systems mean that you can get techs onsite to repair by horse if necessary unlike a space based system where some idiot retired the fleet of repair trucks. So the only remaining option is to launch new ones.
> 
> 
> On Sep 8, 2018, at 10:25 AM, jimlux <jimlux at earthlink.net> wrote:
> 
>> On 9/7/18 10:05 PM, John Reid wrote:
>> Hi all,
>> discussion of how to keep accurate time without access to GPS seems very
>> on topic to me.
>> These people involved in major catastrophe ('end of the world' as you
>> put it) scenarios have a wealth of experience in other ways of keeping
>> accurate time.
> 
> 
> Actually, they don't necessarily have a wealth of experience, because they may have marched themselves down a path where they have a *requirement* for much better timing than they realize, because it is so easy and cheap to get good time today.
> 
> Imagine this scenario - you're a bank, and you batch process checks and deposits in one physical location, so you don't much care about when the check was written or the deposit made.  Then you move to a distributed system across the US, where the reconciliation is done on the basis of the date of the transaction - still probably ok, because there are no transactions during non-business hours, so as long as you reconcile at 1AM, if transaction time stamps are off by 5 minutes, it doesn't matter.
> 
> 
> Now say "we're going to charge you, the customer a fee, if your balance goes negative" and go to 24/7 operations, where transactions are journaled immediately, rather than batch processed at night  If a deposit that was made at 12:00 (but timestamped 12:05)  is followed by a withdrawal made at 12:03 (but timestamped 12:00), you get unfairly charged the overdraft fee.
> 
> For small problems, banks have ways to "unwind" errors.  But if it becomes a systemic thing that's a problem.
> 
> So the bank sets up GPSDOs at each transaction point - problem solved.
> 
> Until GPS fails.
> 
> 
> 
> _______________________________________________
> 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.
> 
> _______________________________________________
> 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.