[teqc] LB2 format: strange epoch on Sunday at 00:00:00

Lou Estey lou at unavco.org
Fri May 28 07:29:34 MDT 2004


This excerpt from Frank Pache at Leica Geosystems:

> we did investigate the strange Sunday epoch issue and found the following:
> We analysed the data logged over the last weekend here in Heerbrugg as well
> as logged by [..snip..] and found the quirk.
> The issue happens when the User Clock is already on the new GPS Week,
> whereas the Channel Time Base is still in the old/previous week. This
> causes a huge negative value for the expression (User Clock Time of Week -
> Channel Time) which is used to multiply the frequency, thus resulting in
> the huge negative code/phase measurement values.
> Accepting this as a tolerable receiver behaviour, what obviously needs to
> be done for the LB2 reader, e.g. RINEX convertes, is to validate the
> expression [User Clock Time of Week - Channel Time] to be none-negative,
> and if its negative to add a week worth of milliseconds (604 800 000 msec)
> to obtain the correct positive value of week seconds.

Also, this can only occur if you are using the LB2 0x0c data record; it does
not occur with the LB2 0x03 data record.

If this problem is causing any teqc user serious grief, please contact me
and send the "build" line from `teqc +id` (so I know which executable build
you are using), and I'll send you an equivalent build for the current
development version of teqc.


More information about the teqc mailing list