Re: Wanted: A Very Accurate Timer

In message <t2q5c1tvoncjud7q700smltovls8l719s6@xxxxxxx>
John Fields <jfields@xxxxxxxxxxxxxxxxxxxxx> wrote:

> > ... I suggested you use an 120Hz clock, so you will actually obtain
> > the sampling resolution you want...

> For me, your suggestion that he use a 120Hz clock implies that you
> don't know what you're talking about.

It seems to me a perfectly valid point, illustrating the universal
truth, hopefully familiar to anyone with an address such as yours, that
the accuracy/resolution of any measuring-instrument must be higher than
that required of the measurements to be made.

> Considering that the OP has specified that: "It must to be accurate
> to within 1/60th of a second over the course of 6 hours." ... means
> that he needs an accuracy of one part in 1,296,000 ...

> Now, what was it you were saying about that 120Hz clock?

Nothing about how its accuracy could be achieved; only, by implication,
that this should be better than 1 part in 2.6 million rather than 1.3

Non-trivial in isolation, but 'standard frequency/time' transmissions
make available to everyone within range the accuracy in which national
standards-bureaux have heavily invested: propagation-variability still
leaves the result 'better' than anything feasible to achieve otherwise.

'Radio-controlled' clocks locked to these are cheap domestic commodity
items, though for any serious application modification/additions would
be needed to produce/use an 'electronic' output instead of the normal
'visual' one. (They don't usually even have a 'stopwatch' mode, for
starters, though £8/$15 watches available here do)
This, however, would need more knowledge/skills than the OP claims.

Though still unconvinced of the relevance of such accuracy-requirement
to what little I've grasped of the clock-watching/button-pushing(!)
application, the discussion IMO remains of some academic interest, if OT
in a 'repair' NG ...

--
Peter Duck <pduck@xxxxxxxxxxxx>
.