[time-nuts] Thunderbolt GPS rollover

k4cle at aol.com k4cle at aol.com
Fri Jun 8 01:45:31 UTC 2012


Mike,
The problem with your solution is that the receiver can get an error in  
calculating the year value and if it writes a number greater than it  
actually is into eeprom, the next time it powers on it will be forced to go  
forward 19 years in time.  And since it is not allowed to go backwards in  
time, it will always have the wrong date.  I have worked with many software  
engineers doing GPS receiver design, believe me, there is no fool proof  
solution to this.
Doug, K4CLE

Connected by DROID on Verizon Wireless

-----Original message-----
From: Mike S <mikes at flatsurface.com>
To: time-nuts at febo.com
Sent: Fri, Jun 8, 2012 00:51:21 GMT+00:00
Subject: Re: [time-nuts] Thunderbolt GPS rollover

On 6/7/2012 8:02 PM, Bob Camp wrote:
> One could put in a routine that looks at the date the software was
> written and "fix" any date that shows up as being in the past.

After all the issues seen after the last rollover, I'd think receivers 
would have been made robust against this.

One obvious method would be to keep a checkpoint date (or 1024 week 
cycle # + week offset) in EEPROM. To avoid issues with using up EEPROM 
r/w cycles, only update it anytime the currently received date is more 
than a year beyond the currently stored date.

_______________________________________________
time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to  
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.



More information about the time-nuts mailing list