[time-nuts] 5 MHz from 10 MHz

[Marco IK1ODO -2]

At 14.01 23/07/2007, you wrote:
>); SAEximRunCond expanded to false
>Errors-To: time-nuts-bounces+ik1odo=spin-it.com+ik1odo=spin-it.com at febo.com
>
>No point using an expensive divider with an 8640B.  The close-in noise is
>dominated by the cavity, especially at the bandwidths used by their lock
>mechanism, and the broadband floor is ECL at ~-150 dBc/Hz.  Just stick a
>74xxx74 in there (what *is* considered to be the lowest-noise modern TTL
>family, anyway?)
>
>-- john, KE5FX

I used 74ABT logic (Philips, see Digikey) with good success up to 
over 200MHz clock, driven by SN65LVDS34 as squarer, with good success 
(it is an SDR QSD mixer, and had much less noise than using 
74-shottky series). No jitter data measured, however. I think that 
the same circuit - or even a much simpler one - would be perfect to 
drive an 8640B. A 74AC74 would be just fine at 10 MHz. 
Metrology-grade dividers are another thing, of course.

The schematic of my front-end is at 
http://www.spin-it.com/sdr/ik1odo_sdr1.html - use only what needed, 
of course :-)

73 - Marco IK1ODO - AI4YF