[time-nuts] New NTBW50AA

[Charles Steinmetz]

Dave wrote:

>I set the temp scale to 250m°C/div last night 
>and it remained steady for quit a while, but 
>then it started the jumping again.  It seems to 
>jump on 1° increments, like from 37.750 to 
>38.749 in less than 1 minute, but there are step 
>variations during the 1°  movements.

I do not know whether LH just displays every 
"n-th" sample to produce long-time-scale plots, 
or whether it averages "n" samples to get the 
displayed value.  If it averages, each pixel 
would include data from more than one sample, 
which would produce intermediate values that the 
GPSDO did not actually report.  You may want to 
set the time scale so you can see each one-second 
sample (I think one minute per division will do 
this).  At that scale, you may find that all of 
the samples are in one degree increments.  In 
other words, the appearance of intermediate 
values may be an artifact of display averaging -- 
the actual data from the GPSDO may be strictly in 
1v increments.  Mark could say for sure.

>So what you're saying is it won't affect the freq stab or adev?

We (the time-nuts community) do not have hard 
data about how the temperature data is used in 
the Trimble GPSDOs, but it appears that it is not 
used during locked operation except to "teach" 
the unit about the OCXO's temperature 
coefficient.  What it has "learned" is applied 
during holdover in an attempt to improve holdover 
stability.  I doubt many time nuts would make 
critical measurements while their GPSDO was 
unlocked, so there is probably no practical 
difference as far as we are concerned.

Best regards,

Charles