[time-nuts] Sidereal timekeeping

iovane at inwind.it iovane at inwind.it
Mon May 16 10:02:06 UTC 2011


The background of my request is an OT story. Just to mention briefly, I already 
have an ordinary (non-radio-controlled) clock machine which turns a miniature 
torsion balance in a sealed glass vessel. It runs on a single AA battery. No 
extreme accuracy needed. I wont to modify the rate to sidereal, and might have 
to replicate the setup too. I figure that the solution I should pursue is 
getting the "odd" crystals. 
Now it is clear to me that I have to explore two options, a) contacting a 
crystal manufacturer, b) modifying 32768 crystals. 
Thanks,
Antonio

hmurray at megapathdsl.net wrote:
>
>> does anybody out there have any ideas as where to find a 32859Hz crystal 
(1/
>> 2  that value would be better) to be used to replace 32768 crystals in
>> ordinary  clocks? I think that 32768 crystals cannot be dragged that much.
>> I've already  read the JimLux article somewhere on the web, but I would be
>> pleased finding a  simpler solution. Also, I already have computer programs
>> that show sidereal  time. 
>
>I think it depends upon what you mean by "ordinary clocks".
>
>Most of the recent wall clocks I've seen are battery powered (single AA) and 
>resynchronize nightly via WWVB.
>
>If you want sidereal time, you won't have anything to synchronize to.  What 
>sort of accuracy are you interested in?  If you want reasonable accuracy, 
you 
>will need an external signal.  (You can provide power over the same cable.)
>
>My straw man would be to send 32859Hz down coax or twisted pair and feed it 
>into the xtal-in pin on the clock chip.  I'm not sure how to set the time.  
>You can cut the lead to the antenna to make sure it doesn't sync to WWVB.
>
>You can make 32859Hz from a PIC (or any micro you like) running off any 
handy 
>frequency.  Given that this is time-nuts, I'd suggest 10 MHz from a GPSDO.
>
>It might be simpler to dump the 32KHz and WWVB chip and drive the motor 
>directly from a 1 PPS signal.  Just use a sidereal second rather than a 
>normal second.







More information about the Time-nuts_lists.febo.com mailing list