"Joseph S. Myers" wrote on 1999-12-04 12:14 UTC:
Do you happen to know what GPS receivers are available that provide user access to TAI (or some other form of the leap second offset information transmitted in GPS; ideally all the parameters from Table 2-11 (UTC Parameters) of the GPS Standard Positioning Service Signal Specification 2nd Edition)? This doesn't seem to be the sort of feature manufacturers of GPS receivers advertise, but it's of more interest to me in a GPS receiver than the route storage features that get promoted.
I don't have much experience with current low-cost GPS receivers and their serial port protocols. When we played in Erlangen around with NTP, we had a Mainberg GPS167 time receiver, i.e. a GPS receiver that was especially designed for reference clock applications. And even that one did *not* provide TAI, or the current UTC-TAI or UTC-GPS difference. It's output format looked suspiciously compatible to that of other Meinberg DCF77 receivers, which naturally have no clue about TAI. More info on Meinberg receivers is available on http://www.meinberg.de/english/index.htm including the full manuals. Also the IRIG-A and IRIG-B US military standard time telegram formats supported by some GPS reference clock receivers only provide UTC and do not allow you to reconstruct TAI. I have seen output from some Trimble GPS receivers that contained the GPS time as a (week, second) pair, which can be converted easily into TAI, but I am also not sure, how much of the leap second announcement you can access here. Trimble clocks such as the Thunderbolt on http://www.trimble.com/oem/om_timng.htm support the Trimble Standard Interface Protocol (TSIP) via RS232 as documented on http://www.trimble.com/support/files/thunder.htm Even there, they don't mention whether TAI is available. A bit more information is available on ftp://ftp.trimble.com/pub/tracking/pla400-TAIP-Reference.pdf which describes the Trimble ASCI Interface Protocol (TSIP). It *does* provide in the TM message a 2-digit "GPS/UTC Time Offset" field that indicates the number of seconds, but there is no leap second announcement field. Conclusions: If you look long enough, you might get TAI extracted out of a good GPS receiver, but not all GPS receivers support this and even for those that do, the leap second announcement is not necessarily available or well supported, or the documentation is very obscure. No wonder, hardly anybody is using TAI in practice and everything is UTC based. Markus -- Markus G. Kuhn, Computer Laboratory, University of Cambridge, UK Email: mkuhn at acm.org, WWW: <http://www.cl.cam.ac.uk/~mgk25/>
Markus Kuhn schrieb:
"Joseph S. Myers" wrote on 1999-12-04 12:14 UTC:
Do you happen to know what GPS receivers are available that provide user access to TAI (or some other form of the leap second offset information transmitted in GPS; ideally all the parameters from Table 2-11 (UTC Parameters) of the GPS Standard Positioning Service Signal Specification 2nd Edition)? This doesn't seem to be the sort of feature manufacturers of GPS receivers advertise, but it's of more interest to me in a GPS receiver than the route storage features that get promoted.
It is possible to extract TAI from GPS time. Those scales differ just by an integral amount of seconds. As for GPS to UTC conversion, the GPS has correction information that give the current and the future GPS/UTC offset and the time when the switch occurs. Thus leap second announcement can be calculated when approching a switch event. New ntpd code (4.0.x) support Trimble TSIP decoding. Newer Meinberg GPS receivers feature a binary protocol that will give out some more information, though no direct GPS time. GPS time can be reconstructed by converting the Meinberg time stamps to GPS via the GPS/UTC correction value. Regards, Frank Kardel
Dear Markus and others, Markus Kuhn wrote:
"Joseph S. Myers" wrote on 1999-12-04 12:14 UTC:
Do you happen to know what GPS receivers are available that provide user access to TAI (or some other form of the leap second offset information transmitted in GPS; ideally all the parameters from Table 2-11 (UTC Parameters) of the GPS Standard Positioning Service Signal Specification 2nd Edition)? This doesn't seem to be the sort of feature manufacturers of GPS receivers advertise, but it's of more interest to me in a GPS receiver than the route storage features that get promoted.
I don't have much experience with current low-cost GPS receivers and their serial port protocols. When we played in Erlangen around with NTP, we had a Mainberg GPS167 time receiver, i.e. a GPS receiver that was especially designed for reference clock applications. And even that one did *not* provide TAI, or the current UTC-TAI or UTC-GPS difference. It's output format looked suspiciously compatible to that of other Meinberg DCF77 receivers, which naturally have no clue about TAI.
I'm working for Meinberg and have been involved in the development of GPS receivers. Initially, our GPS receivers were designed to be replacements for DCF77 receivers in cases were DCF77 could not be received, so the default serial protocol was made compatible with our DCF77 receivers. Of course, GPS receivers provide increased accuracy and information which is not available from DCF77. So we have implemented a simple binary protocol which can be used to read mostly all of the information the receiver decodes from the satellites' navigation data. This includes raw almanach, ephemeris, and UTC correction data, and also the current time in raw GPS format (week number and second-of-week). An example C source code can be made available which shows how to read this information.
More info on Meinberg receivers is available on
http://www.meinberg.de/english/index.htm
including the full manuals.
Right ;-) Martin -- M. Burnicki Meinberg Funkuhren Bad Pyrmont Germany
participants (3)
-
Frank Kardel -
Markus Kuhn -
Martin Burnicki