[time-nuts] Datum Starloc II GPSDO issues
holrum at hotmail.com
Tue Jun 28 19:08:35 EDT 2016
Ok, these things are a little less horrid than they first presented themselves as. Whenever Lady Heather sends a command to change a setting, it follows it up with a command to read that setting back from the receiver. It turns out that the Datum units don't have any (or possible a very small) received message buffer. Basically they were ignoring any commands sent to them. Increasing the delay after sending something to the unit got things working better.
You can't stop a survey using the normal Trimble "stop survey" command, but you can send a fixed position to the unit which will stop the survey. They definitely don't have an EEPROM for saving settings and location between power cycles.
Sending the proper Trimble "set broadcast packet mask" message (that tells it what messages to send automatically) to the unit causes it to stop sending the primary and/or secondary timing messages. I need to figure out a packet mask setting that won't hose up both the Datum or any of the Trimble units...
Also it appears that it occasionally repeats or skips a primary timing message and that causes a "duplicate time stamp" or "missing time stamp" warning if you send the receiver a "request almanac health (0x29) request. The primary timing message should be sent exactly once per second and each message should have sequential and unique time values.
More information about the time-nuts