[time-nuts] another Ebay mixup, 5370
jmiles at pop.net
Sun Jun 10 16:04:48 EDT 2007
If you send it to a cal lab, they will hook up a cable between the START and
REF jacks, note that the intrinsic jitter is less than 100 ps, slap a
sticker on it, and return it to you with the bill. As long as it passes
this and other tests in the manual, they won't tweak it any further, nor
(arguably) should they.
If it proves impossible to meet the 100-ps spec with the level control in
the PRESET detent, they will either fail the cal, or charge you a lot to fix
So I think it makes sense to measure the input sensitivity first, to see if
the front end and its switchgear is OK. Measuring a weak sine wave, or
measuring one anywhere except at the zero crossings, will make your jitter
performance look artificially poor. I can't exaggerate how important that
One of these days I'll get bored and go through mine with the
factory-specified cal equipment, after which point I will be in a better
position to calibrate other peoples' 5370s. Right now, though, there's no
guarantee that I would be able to improve things. Had some good results
with Dennis Tillman's counter (one of the 'broken' eBay units that was sold
as inoperative with the EXT REF setting) but I never could get it to perform
quite as well as my own, jitter-wise.
-- john, KE5FX
> To everyone: would my unit perform better with very(!) careful
> Is there any cal-lab that can do it inexpensively, and not mess
> up the unit (I
> have had instruments returned to me after cal that were
> mistuned, and worked
> less well than before)?
> There are so many trimmer caps inside that for sure I am not
> going to move a
> single one of 'em :)
> Then again I get <50ps jitter with 8.2 readings per second using the
> external Ref input, so maybe I should be content with that.
> That's a lot better than
> my 53132A.
> Thanks again everyone for your useful hints!
More information about the time-nuts