GPS time.
Paul Halliday
Hi.
I just connected my gps (garmin gps III plus) to my serial port
and realized that simply cat'ing cua0 displays date/time/position of the
unit. (neato). Anyway, how accurate would it be to use the time from this
output for ntp as opposed to my current setup using ntp servers.
Paul H.
"Don't underestimate the power of stupid people in large groups"
___________________
http://dp.penix.org
To Unsubscribe: send mail to majo...@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
di...@covalent.net
> I just connected my gps (garmin gps III plus) to my serial port
> and realized that simply cat'ing cua0 displays date/time/position of the
> unit. (neato). Anyway, how accurate would it be to use the time from this
> output for ntp as opposed to my current setup using ntp servers.
I believe that the CD pin has/can-be-configured to have a 1PPS on it.
Dw
Leo Bicknell
> I just connected my gps (garmin gps III plus) to my serial port
> and realized that simply cat'ing cua0 displays date/time/position of the
> unit. (neato). Anyway, how accurate would it be to use the time from this
> output for ntp as opposed to my current setup using ntp servers.
Your NTP servers are better.
I tested a III Plus, and without a 1 PPS source (which that model
doesn't provide) it's accurate to about 100ms, give or take. Since
real NTP servers are < 1ms, they really aren't that good. It's
not that the time isn't accurate, it's that they were not designed
to communicate with that accuracy to an external device.
If you NTP off the Internet, and want a local backup clock it might
be an acceptable solution. However clocks that can achieve < 1ms
accuracy can be had for < $1000, so if you really care you should
get one of those.
You might want to do some searches for NTP in google.
--
Leo Bicknell - bick...@ufp.org - CCIE 3440
PGP keys at http://www.ufp.org/~bicknell/
Read TMBG List - tmbg-lis...@tmbg.org, www.tmbg.org
Les Biffle
> > I just connected my gps (garmin gps III plus) to my serial port
> > and realized that simply cat'ing cua0 displays date/time/position of the
> > unit. (neato). Anyway, how accurate would it be to use the time from this
> > output for ntp as opposed to my current setup using ntp servers.
>
> Your NTP servers are better.
If you want to buy an appropriate GPS for this, you may wish to check
the Tucson Amateur Packet Radio site (www.tapr.org) and look at the
Motorola Oncore UT+ board. For around $200 you can have a stratum 1
quality time reference.
Regards,
-Les
--
Les Biffle
(480) 585-4099 l...@safety.net http://www.les.safety.net/
Network Safety Corp., 5831 E. Dynamite Blvd., Cave Creek, AZ 85331
Steve Ames
plus or minus 1 microsecond of UTC.
Mike Silbersack
On Sat, 30 Mar 2002, Leo Bicknell wrote:
> Your NTP servers are better.
>
> I tested a III Plus, and without a 1 PPS source (which that model
> doesn't provide) it's accurate to about 100ms, give or take. Since
> real NTP servers are < 1ms, they really aren't that good. It's
> not that the time isn't accurate, it's that they were not designed
> to communicate with that accuracy to an external device.
OTOH, 100ms is pretty close; I doubt many people need time better than
that. The one big advantage I can see with using a GPS receiver vs NTP
servers is security & reliability; I've always worried that my clock
might start to drift to a misconfigured NTP server. Taken to a paranoid
level, you could worry that someone was faking NTP replies to throw your
clocks off. :)
So, even at 100ms accuracy, it might be better to use a local GPS unit.
<shrug>
Mike "Silby" Silbersack
Paul Halliday
>
> On Sat, 30 Mar 2002, Leo Bicknell wrote:
>
> > Your NTP servers are better.
> >
> > I tested a III Plus, and without a 1 PPS source (which that model
> > doesn't provide) it's accurate to about 100ms, give or take. Since
> > real NTP servers are < 1ms, they really aren't that good. It's
> > not that the time isn't accurate, it's that they were not designed
> > to communicate with that accuracy to an external device.
>
> OTOH, 100ms is pretty close; I doubt many people need time better than
> that. The one big advantage I can see with using a GPS receiver vs NTP
> servers is security & reliability; I've always worried that my clock
> might start to drift to a misconfigured NTP server. Taken to a paranoid
> level, you could worry that someone was faking NTP replies to throw your
> clocks off. :)
This is the answer I was kinda hoping for. I think that accuracy
to ~100ms from a known source is a little more comforting than <1ms from a
server that I have no control over. I am not maintaining a space program,
just a dozen machines in my room that really serve no other purpose than
personal entertainment.
Thanks for all the replies!
>
> So, even at 100ms accuracy, it might be better to use a local GPS unit.
>
> <shrug>
>
> Mike "Silby" Silbersack
>
>
Paul H.
"Don't underestimate the power of stupid people in large groups"
___________________
http://dp.penix.org
To Unsubscribe: send mail to majo...@FreeBSD.org
Erik Trulsson
> On Sun, 31 Mar 2002, Mike Silbersack wrote:
>
> >
> > On Sat, 30 Mar 2002, Leo Bicknell wrote:
> >
> > > Your NTP servers are better.
> > >
> > > I tested a III Plus, and without a 1 PPS source (which that model
> > > doesn't provide) it's accurate to about 100ms, give or take. Since
> > > real NTP servers are < 1ms, they really aren't that good. It's
> > > not that the time isn't accurate, it's that they were not designed
> > > to communicate with that accuracy to an external device.
> >
> > OTOH, 100ms is pretty close; I doubt many people need time better than
> > that. The one big advantage I can see with using a GPS receiver vs NTP
> > servers is security & reliability; I've always worried that my clock
> > might start to drift to a misconfigured NTP server. Taken to a paranoid
> > level, you could worry that someone was faking NTP replies to throw your
> > clocks off. :)
>
> This is the answer I was kinda hoping for. I think that accuracy
> to ~100ms from a known source is a little more comforting than <1ms from a
> server that I have no control over. I am not maintaining a space program,
> just a dozen machines in my room that really serve no other purpose than
> personal entertainment.
Yes, but that is why one shouldn't rely on *a* server. When using NTP
it is a good idea to get the time from several NTP servers.
The chance that all of them are misconfigured at the same time is
fairly small.
OTOH, taking the time from a local GPS receiver doesn't sound like a
bad idea either if one doesn't need extremely good timekeeping.
--
<Insert your favourite quote here.>
Erik Trulsson
ertr...@student.uu.se
M. Warner Losh
Paul Halliday <d...@penix.org> writes:
: I just connected my gps (garmin gps III plus) to my serial port
: and realized that simply cat'ing cua0 displays date/time/position of the
: unit. (neato). Anyway, how accurate would it be to use the time from this
: output for ntp as opposed to my current setup using ntp servers.
Generally the output time from these receivers via the serial port
isn't very good from a precise time keeping point of view. There's
about 10ms of variation in the reported time, which makes them very
very noisy for the sort of work that ntp does. For a local network,
it shouldn't be too bad, but expect to see a lot of strange things
happen because the error can be so large.
Warner
M. Warner Losh
Leo Bicknell <bick...@ufp.org> writes:
: In a message written on Fri, Mar 29, 2002 at 06:04:11PM -0600, Paul Halliday wrote:
: > I just connected my gps (garmin gps III plus) to my serial port
: > and realized that simply cat'ing cua0 displays date/time/position of the
: > unit. (neato). Anyway, how accurate would it be to use the time from this
: > output for ntp as opposed to my current setup using ntp servers.
:
: Your NTP servers are better.
:
: I tested a III Plus, and without a 1 PPS source (which that model
: doesn't provide) it's accurate to about 100ms, give or take. Since
: real NTP servers are < 1ms, they really aren't that good. It's
: not that the time isn't accurate, it's that they were not designed
: to communicate with that accuracy to an external device.
Also, the reference clocks that ntp servers use generally are in the
+- 1us or better rather than 1ms sync that you can get over a wan.
Of course, don't look too closely under the covers of ntp. it has
well documented startup transient conditions that some applications
may be senentive to.
Warner
M. Warner Losh
: to ~100ms from a known source is a little more comforting than <1ms from a
: server that I have no control over. I am not maintaining a space program,
: just a dozen machines in my room that really serve no other purpose than
: personal entertainment.
Also keep in mind that the US governement reserves the right to turn
off GPS at any time to selected regions of the globe.
The problem with the large variation in time is that if you are trying
to use it for a NTP reference clock, bad things happen to the
algorythm and you get substantially worse performance than you would
otherwise.
Warner
Sean Mathias
the receiver is configured properly and acquires reference satellites fairly
regularly, this should provide almost the best possible time. For LBS, the
norm is to acquire three primary satellites and an additional satellite if
possible for reference. As each GPS satellite has 2 or 3 onboard atomic
clocks, this would seem like the best possible reference available and given
the availability of receivers in a PCI form factor, inexpensive and broadly
available to all.
SM
-----Original Message-----
From: Erik Trulsson [mailto:ertr...@student.uu.se]
Sent: Sunday, March 31, 2002 9:32 AM
To: Paul Halliday
Cc: Mike Silbersack; Leo Bicknell; hac...@FreeBSD.ORG
Subject: Re: GPS time.
On Sun, Mar 31, 2002 at 10:44:55AM -0600, Paul Halliday wrote:
> On Sun, 31 Mar 2002, Mike Silbersack wrote:
>
> >
> > On Sat, 30 Mar 2002, Leo Bicknell wrote:
> >
> > > Your NTP servers are better.
> > >
> > > I tested a III Plus, and without a 1 PPS source (which that model
> > > doesn't provide) it's accurate to about 100ms, give or take. Since
> > > real NTP servers are < 1ms, they really aren't that good. It's
> > > not that the time isn't accurate, it's that they were not designed
> > > to communicate with that accuracy to an external device.
> >
> > OTOH, 100ms is pretty close; I doubt many people need time better than
> > that. The one big advantage I can see with using a GPS receiver vs NTP
> > servers is security & reliability; I've always worried that my clock
> > might start to drift to a misconfigured NTP server. Taken to a paranoid
> > level, you could worry that someone was faking NTP replies to throw your
> > clocks off. :)
>
> This is the answer I was kinda hoping for. I think that accuracy
> to ~100ms from a known source is a little more comforting than <1ms from a
> server that I have no control over. I am not maintaining a space program,
> just a dozen machines in my room that really serve no other purpose than
> personal entertainment.
Yes, but that is why one shouldn't rely on *a* server. When using NTP
it is a good idea to get the time from several NTP servers.
The chance that all of them are misconfigured at the same time is
fairly small.
OTOH, taking the time from a local GPS receiver doesn't sound like a
bad idea either if one doesn't need extremely good timekeeping.
--
<Insert your favourite quote here.>
Erik Trulsson
ertr...@student.uu.se
To Unsubscribe: send mail to majo...@FreeBSD.org
Vladimir Egorin
> In message: <Pine.LNX.4.43L0.0203...@saruman.xwin.net>
> Paul Halliday <d...@penix.org> writes:
> : This is the answer I was kinda hoping for. I think that accuracy
> : to ~100ms from a known source is a little more comforting than <1ms from a
> : server that I have no control over. I am not maintaining a space program,
> : just a dozen machines in my room that really serve no other purpose than
> : personal entertainment.
>
> Also keep in mind that the US governement reserves the right to turn
> off GPS at any time to selected regions of the globe.
Hopefully European GPS project (Galileo) will provide an alternative.
It still has a long way to go though.
--
Vladimir
Terry Lambert
> > Also keep in mind that the US governement reserves the right to turn
> > off GPS at any time to selected regions of the globe.
>
> Hopefully European GPS project (Galileo) will provide an alternative.
> It still has a long way to go though.
Galileo strikes me as unnecessary, unless the receivers will be
cheaper to get the same resolution. The 1 meter resolution seems
a little poor, compared to differential.
I guess if the U.S. government reserves the right to turn it off,
then there's a reason to route around that...
-- Terry
sth...@nethelp.no
> > It still has a long way to go though.
>
> Galileo strikes me as unnecessary, unless the receivers will be
> cheaper to get the same resolution. The 1 meter resolution seems
> a little poor, compared to differential.
Galileo may be unnecessary *if* you trust the US. As a European,
my view of US and European politics is that they're sufficiently
different (both in methods and goals) that I don't trust the US
that much. Thus I think Galileo is a good thing, even if it'll be
very expensive.
Steinar Haug, Nethelp consulting, sth...@nethelp.no
Terry Lambert
> > > Hopefully European GPS project (Galileo) will provide an alternative.
> > > It still has a long way to go though.
> >
> > Galileo strikes me as unnecessary, unless the receivers will be
> > cheaper to get the same resolution. The 1 meter resolution seems
> > a little poor, compared to differential.
>
> Galileo may be unnecessary *if* you trust the US. As a European,
> my view of US and European politics is that they're sufficiently
> different (both in methods and goals) that I don't trust the US
> that much. Thus I think Galileo is a good thing, even if it'll be
> very expensive.
The only thing that matters is consumer pricing of the receivers,
and whether or not the resolution is high enough to keep computer
operated vehicles safely within their lane markings.
The U.S. ability to turn off the GPS isn't really the threat that
you think. Most U.S. airports had equipment installed to broadcast
a "correction" signal for GPS, well before they turned the "wiggle"
off. They turned it off because it was a losing proposition, and
a few meters weren't going to make that much difference to a 200
megaton device anyway. Leaving it on only gave cruise missles a
nice target lock for transmitters at all U.S. commercial airports
that were installed to de-wiggle the signal. 8-).
If it comes down to it, the U.S. has the capability of "turning
off" Galileo, if it ever felt that it needed to do so. 8-) 8-).
So if it's being built out of distrust, well, it's already nor
getting around the problem.
-- Terry
Wilko Bulte
> sth...@nethelp.no wrote:
> > > > Hopefully European GPS project (Galileo) will provide an alternative.
> > > > It still has a long way to go though.
> > >
> > > Galileo strikes me as unnecessary, unless the receivers will be
> > > cheaper to get the same resolution. The 1 meter resolution seems
> > > a little poor, compared to differential.
> >
> > Galileo may be unnecessary *if* you trust the US. As a European,
> > my view of US and European politics is that they're sufficiently
> > different (both in methods and goals) that I don't trust the US
> > that much. Thus I think Galileo is a good thing, even if it'll be
> > very expensive.
..
> nice target lock for transmitters at all U.S. commercial airports
> that were installed to de-wiggle the signal. 8-).
>
> If it comes down to it, the U.S. has the capability of "turning
> off" Galileo, if it ever felt that it needed to do so. 8-) 8-).
While not stepping up to solve the world politics: the US government
claiming the right to define the law for everything is unnerving to
lots of non-US (and US I suppose) people alike.
GPS is just one of these things..
--
| / o / /_ _ wi...@FreeBSD.org
|/|/ / / /( (_) Bulte Arnhem, the Netherlands
We are FreeBSD. Resistance is futile. Prepare to be committed.
Terry Lambert
> While not stepping up to solve the world politics: the US government
> claiming the right to define the law for everything is unnerving to
> lots of non-US (and US I suppose) people alike.
>
> GPS is just one of these things..
I disagree. The problem is the engineers.
When the U.S. bombed Iran during the Gulf War, the communications
came back up very quickly -- or never went down, in many cases --
because the TCP/IP protocols on which those command and control
systems were based, developed on U.S. Government funding through
DARPA, and were resiliant to damage, even to large portions of
the infrastructure.
This was a pain in the butt for the U.S., but it validated that
the same system, in the face of an attack on the U.S., would
remain operational here, too.
The GPS system, on the other hand, was designed so that "the
right people" could turn it off. This probably means that
"the wrong people" can also turn it off, and it means that if
"the right people" ever become "the wrong people" (John Travolta
has played enough of these characters in recent movies ;^)),
they can also "turn it off".
This is actually a design flaw, from one perspective: it means
that you can't rely on the system, 100%, in a crisis situation,
since who "the right people" are may not fall on your side of
the argument, even if you were/are "the right people".
As long as engineers are willing to design things with holes
in them, then we will have things with holes in them.
If it's possible for someone to "turn off" Galileo, then the
only thing this argument devloves to is a disagreement about
who "the right people" are ("the right people" are defined as
"the people who are initially entrusted with the power switch").
To take this back around: for either system to be useful for
a timebase, you have to be relative sure that it can't lie;
I think because of this, any time server based on it can't
truly be considered "tier 1", because the time base is not
under the control of the system operator.
To note Galileo: cool name, but if it's not less expensive to
buy a receiver, and all other things are equal, then who
really cares? All we are winning by using it is a lower
commercial value system, with a different set of hands that
can pull the plug on us.
Actually, the name is a bit inappropriate, considering the
whole project revolves around the Earth... literally. ;^).
PS: Personally, I'd never trust a robotically driven vehicle
based on a system where someone could drive me off the cliff
by transmitting bogus data, or that would scream for me to
get back up in front of the R.V. in the middle of me making
a sandwitch because some idiot thought economic sanctions
included the inability to safely use the highway system in
an area not under their political control.
-- Terry