[time-nuts] Fix Trimble Oscillator??

[John Miles]

> > jmiles at pop.net said:
> > > I can confirm that the whole oscillator including oven runs
> on 12 volts.
> > > However, you do have to connect two different pins to +12 and
> > two different
> > > pins to ground to make it work (see attached).  ...
> >
> > Nice pictures/graphs.  Thanks.
> >
> > What's the spike at 60KHz?
>
> No telling, probably leakage from a switching regulator somewhere.
>
> Spurs under -150 dBc/Hz are elusive enough to track down, and once you get
> down below -160 it's often impossible to observe them from one measurement
> to the next.  Digital analyzers like the TSC5120A attempt to remove their
> own spurs from the measurement, but this process isn't always
> perfect.  The
> only sane approach is to ignore low-level spurs unless/until they are
> associated with specific problems.  (An actual defect in the
> oscillator will
> usually show up at well over -140 dBc/Hz.)

Another point: it was lazy of me to use "dBc/Hz" to describe spur
amplitudes -- they should be labelled with either dBc or dBm units, as they
don't undergo normalization to 1 Hz BW the way the noise does.  In this
case, this older plot was taken before my software knew how to parse the
spur table generated by the TSC 5120A, so it didn't draw the spurs as dashed
lines with the correct amplitude, or display a list of the spurs in a table
next to the graph.  That means the actual amplitude of these spurs is
several dB higher than shown, but they're still small potatoes.

-- john, KE5FX