[time-nuts] WTB: Oscilloquartz 8607, 8600 or similar OCXO
bruce.griffiths at xtra.co.nz
Sat Jun 7 19:01:19 EDT 2008
Andreas Tschammer wrote:
> thanks for the valuable information.
> I am aware of the importance of the zero crossing slope to noise ratio in this kind of setup.
> I will only use this design as a first step, because it is easy to replicate and seems to have
> a sufficiently low noise floor for my applications. I also have to build the isolation amps with
> transformer coupling. But the most difficult part seems to find a good reference oscillator ...
> Andreas, DL2KCL
The LT1028 filter has some essential details missing like the source
impedance (ie mixer IF port termination circuit).
JPL and NIST favour an LC PI filter low pass filter to attenuate Rf
above 100kHz or so followed by an RC low pass filter.
The LM161has a maximum differential input of 5V before destruction.
At least add a pair of inverse parallel diodes across its inputs to
avoid potential destruction.
Comparators with totem pole output stages (like the LM161) tend to
oscillate when the input signal has a relatively low slew rate (this may
be why such a complex output circuit is used).
Driving the comparator from a relatively high impedance source (Thevenin
equivalent = 5K in parallel with 10K) doesnt help,
Even the early NIST designs:
are better, although far from optimum.
Note that all amplifier/filter stages are diode clamped.
JPL favour optically isolated outputs:
US patents 44011953, 4425543
Don't copy their isolation amplifier its not that good (besides the
LH0033 is out of production).
For isolation amplifier designs with high performance see:
More information about the time-nuts