[time-nuts] Fury Realhamradio listing

[SAIDJACK at aol.com]

In a message dated 4/28/2007 22:21:55 Pacific Daylight Time,  
bruce.griffiths at xtra.co.nz writes:

>Obtaining 100ps resolution, particularly when one is only making 1  
>measurement per second, is relatively easy and inexpensive to do with  a 
>few discrete components.

>Obtaining much better than 150ps  resolution is somewhat more difficult 
>when using the technique  employed by the HP53131/2 in its CMOS gate  array.

>Bruce


Hi Bruce,
 
that may be, but I don't believe the Z3801A has 100ps hardware resolution  
until someone proves it to me. Maybe we should ask Jack Kusters, he worked on  
that unit.
 
I just don't think that HP would have given the Z3801A 100ps resolution,  and 
the 53132A "only" 150ps resolution. Doesn't make any sense, since the  
counter absolutely needs the resolution, but a GPSDO does not.
 
Quite frankly there is no need for that kind of resolution since the noise  
on the VP Oncore is 104ns peak to peak. Why would you have 100ps per second  
resolution if that is three orders of magnitude below your inherent 1 second  
noise??
 
Another fact is that the GPSDO will average that resolution over say 1000  
seconds anyway in it's control loop, so 100ps average over 1000s is orders of  
magnitude better than any OCXO local oscillator could achieve anyway.
 
Also, what is the granularity of the VP Oncore 1PPS output? I am sure that  
it's orders of magnitude above 0.1ns, so even if the Z3801A TI hardware had 
that  kind of resolution, it would be way more than needed.
 
They probably have 1ns - 5ns hardware resolution, and simply average down  to 
0.1ns apparent resolution.
 
I hope someone familiar with the Z3801A hardware here can shed some light  on 
this :)
 
bye,
Said



************************************** See what's free at http://www.aol.com.