[time-nuts] Hanging bridge question

Tom Van Baak tvb at LeapSecond.com
Wed Mar 26 01:08:10 EDT 2014


> The lowest cost solution is to do the correct entirely in software.
> After the measure the phase, simply add the correction.
>
> All you need to know is the phase.  There is not point in correcting
> the pulse, you don't need a corrected pulse.  What you want is a
> measurement of the phase.

This depends on the goal. There are two types of GPS timing products. Those which output time & frequency and those which output time only.

For time and frequency you design a GPSDO, in which case you have a choice of h/w or s/w sawtooth correction. Most people choose s/w since, as you correctly assume, a GPSDO already includes some sort of phase or time interval measurement circuit along with a microprocessor to do the math.

But for a timing only GPS product (e.g., the base models from www.cnssys.com) the goal is just a precise 1PPS output. This class of product tends to use h/w sawtooth correction, since by design there is no TIC or OCXO in the box.

There's also a third way to do it -- sawtooth correction provided by PC software tools like Tac32Plus or DSPmon (similar to TBoltmon). In this case the PC reads correction messages from the receiver and measurements from an external TIC and applies the sawtooth correction before writing the composite result to a log file. DSPmon has support for the hp 5334 and 53132.

I hope this helps more than it confuses.

/tvb




More information about the time-nuts mailing list