> Doesn't that give you an inaccuracy of +/- 5ms on your 60 seconds > measurement?
Correct which means worst case you could drift +/- 7 secs/day which most would consider unreasonable. If you upped from 100hz to 1khz then you'd reduce it to +/1 .5 ms per minute which would get you +/-.7sec/day worst case drift. Possibly reasonable for a clock with only 1 minute resolution. You could also involve one of the Atmegas programmable timers and temporarily attach the 1pps calibration source to one timers control lines. You might be able to obtain even better results. You'd probably have to implement the trim routine in assembly code. But Michail made a very good point. Perfecting the software could take allot of time, and even if you are coding to learn, you have to put a price on how much time you spend on what's only one aspect of getting your entire clock completed ..c PS) I'm using a TXCO on my most recent clock :) -- You received this message because you are subscribed to the Google Groups "neonixie-l" group. To post to this group, send an email to neonixi...@googlegroups.com. To unsubscribe from this group, send email to neonixie-l+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/neonixie-l?hl=en-GB.