[time-nuts] Can M12+T be improved?

[billdobson at uk2.net]

   For some time now I have been planning a GPSDO based on a M12+T and an
   Oscilloquartz 8663 double OCXO. I have been following the discussions
   on Sawtooth Correction, but wonder if there is another approach which
   will remove the problem.
   It is based on the assumption that the sawtooth correction message is
   in effect a measure of the phase error between the
   16.368MHz TCXO in the M12 and the incoming code rate from the
   satellite. The idea is to remove the onboard TCXO and replace it by a
   16.368MHz clock synthesysed from the OCXO, using a clock generator such
   as the TI CDCE913. The Sawtooth Corection Message would then be used to
   control the 10MHz OCXO. If this were to work it may lead to an improved
   10Mhz stability as well as removing the need for sawtooth correction if
   the 1pps output is used. What do you think?