[time-nuts] Cesium Mechanical Chronometer

[Scott McGrath]

Remember you ‘beat’ a clock using a audio amplifier and a standard signal there is a screw which adjusts the tension on the escapement spring,   Now you could use a reduction drive to turn the screw or take direct control of the escapement spring using the mechanical ‘ticks’ of the escapement as the input to a control loop

Content by Scott
Typos by Siri

> On Jan 30, 2020, at 7:05 PM, Tom Van Baak <tvb at leapsecond.com> wrote:
> 

That would be a fun project. There are examples of measuring a M21 on Bryan's site:

https://www.bmumford.com/mset/tech/chrono/

Here are phase and ADEV plots for my M21:

http://leapsecond.com/pages/m21/

That page also shows you a typical chronometer rate card, which provides the key "paper clock" advantage over using the clock dial alone.

I use a piezo pickup to extract timing pulses from the clock. The audio waveform isn't pretty but you can form a low jitter 1PPS out of it. Laser sensors give a cleaner signal but are more difficult to use with an M21.

Running a GPS/CSAC + M21 in a master/slave arrangement should be easy, although I don't know how you'll handle the rate card corrections.

Running them in phase lock will be much harder. You can probably discipline the CSAC from the M21 using RS232 commands to the CSAC. But to discipline the M21 from the CSAC requires that you have a way to dynamically adjust the rate of the M21 at ppm levels. That's going to be tricky, given that high-end compensated chronometers like this are specifically designed to be as immune to internal and external changes as possible. One avenue may be the winding interval: notice the slopes of the phase plot.

The biggest problem I had with long-term data collection was re-winding the chronometer. If you design a non-invasive auto-winder as part of your project, please contact me.

/tvb


On 1/30/2020 7:49 AM, Tom Bales wrote:
>> And now for something completely different:  I am working on a quixotic
>> project to control a standard, detent-escapement marine chronometer (e.g.,
>> Hamilton 21) with a CSAC cesium atomic clock module.  Yes, I know this
>> makes no sense--but, then, we're timenuts.  I want the mechanical
>> chronometer to function normally if the CSAC signal, presumably a 1pps
>> pulse, is lost.  The CSAC will be GPS disciplined, so during normal
>> operation, with an operating GPS constellation, the time is referenced to
>> UTC via GPS; if GPS is lost, then the CSAC takes over and its 1pps signal
>> drives the chronometer; if all electronics are lost, the chronometer hangs
>> in as a mechanical chronometer.  Has anyone any experience with
>> electrically controlling (or disciplining) a marine chronometer?
>> 
>> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
> 


_______________________________________________
time-nuts mailing list -- time-nuts at lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.