[time-nuts] New precision watch

Chuck Harris cfharris at erols.com
Mon Dec 19 19:12:43 EST 2011

Looking at the datasheet for the DS3232, it doesn't appear
that they mean for it to run off of a small coin cell in a
watch.  Its battery operating capability is purely to keep
it running when the main power is turned off.

I would suspect that the DS3232's power consumption is due
to its being able to drive sizable loads on its various pins.

There is nothing inherent in temperature correcting a clock
that should take significantly more power than would be used
in a normal watch chip.  Measuring the temperature would be
the most power hungry operation, I would suspect... but
fortunately that doesn't have to be done all that often,
perhaps once per minute.

After the temperature is measured, a lookup table can be used
to find a second by second correction value to be added to
the seconds counter.

-Chuck Harris

Dan Rae wrote:
> On 12/19/2011 5:34 AM, Chuck Harris wrote:
>> It is not clear to me that a 32KHz xtal is any less stable
>> than a 262KHz xtal, though. I would think there would be a lot
>> more to be gained by using a microprocessor/thermistor to
>> measure the temperature within the watch, and provide an
>> adjustment to compensate for the xtal's natural temperature
>> sensitivity.
> Chuck, I have recently made a couple of Real Time Clocks for my homebrew radios using
> the Maxim DS3232 IC which has a built in TCXO at 32+ kHz doing what you describe, the
> temperature is checked at intervals and the corrections applied. The performance is
> excellent, once set, getting to what I term "Harrison Level", i.e. less than a second
> a week error. The downside is that the standby battery demand is pretty high and my
> first builds using a Lithium button cell ran for only around a month max, so I had to
> go over to NiMH rechargeables. I suspect using this technology in a watch one would
> have the same problem.
> The best crystals for room temperature use are around 4 MHz with temperature
> inflections around 20C, and this is what was used in an early Braun alarm clock I had
> which also had this kind of performance. Long gone, alas.
> Dan
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

More information about the time-nuts mailing list