[time-nuts] Phase noise of HP8663A 640 MHz reference?

[ed breya]

Thanks all, for replies and possibly running some tests. The reason I'm 
asking about this is that now that I finally have the HP8663A, I'd like 
to rig it up with the HP11179C, and its documentation mentions all this 
option 3 stuff on the 8663A. So I started wondering, just how good is 
it, and if there's a "typical" or representative spec. Again, I believe 
option 3 just means that the particular unit was tested at the factory 
before delivery, and the results documented and included with the unit. 
Over time, this info could be lost.

My unit has option 2 only, which is fancier phase modulation, that could 
be handy for some things. Before it arrived, I studied up the design and 
evolution in the manuals. It appears that the 8662A and earlier 8663A 
models may have had the 10544 OCXO installed, while mine is one of the 
later versions, officially sporting the 10811-60111. I found that this 
is indeed what's in there, and the cal seal is marked ca 2014 - likely 
the last time it was messed with. It seems to work so far, and it is 
quite an impressive beast.

On the 11179C, I had a stroll down memory lane for a while. I kind of 
forgot how much I had worked on it about a dozen years ago. Since I had 
no prospect of getting a 8662A or 8663A back then, I had been collecting 
and mounting parts to make a fairly high grade built-in 640 MHz source. 
The scheme was based on a 80 MHz VTOCXO, to be doubled up and filtered 
thrice to 640 MHz, and phase locked to 10 MHz if needed. I already had 
all the major pieces fitted, and a bunch of notes on the design details.

I started yanking all this stuff out to restore it to stock, but then 
started thinking "not so fast - maybe this could be pretty good after 
all." Since I really don't know how good the 640 MHz from the 8663A is, 
I'll keep all the pieces and info from the other scheme together, just 
in case.

Ed