Hi As explained to me, the corrections result in an improvement in location and that’s is their objective. They don’t care about any time error they might generate part of the correction process. You could call this sloppy math. Without a time reference in their system, they don’t have anything to “correct to” in order to reduce that error. Simply put, if all the corrections get you to the right location by being 200 ms off (for all of them), that’s fine as far as the augmentation system is concerned. ( = add 0.2 seconds to all the sat clocks is not a problem for them ). Bob > On Apr 17, 2019, at 8:45 AM, Greg Troxel <gdt@lexort.com> wrote: > > Bob kb8tq <kb8tq@n1k.org> writes: > >> I have never seen any manufacturer recommend SBAS for improved timing. The >> simple answer seems to be that the reference system does not have time sources >> in it and that the software that generates the corrections does not consider time to >> be important. >> >> SBAS (or WAAS or whatever) will improve your location estimate. In the survey in >> part of the process things might go a little better. My guess is that pretty much everybody >> *assumes* you already have a well known location for your timing device. > > I don't follow this. Differential corrections are basically deltas to > pseudoranges. Sure, they are often encoded as a pseudorange error and > an origin time and rate for pseudorange error (NDGPS), and as I > understand it, this is gridded for interpolation in WAAS. So they > should help time as well as position. > > But, it may be that the short-term dynamics of these are unhelpful in a > GPSDO control loop, and that's a far harder question.