[time-nuts] Electronically Disciplined Mechanical

Bert, VE2ZAZ ve2zaz at yahoo.ca
Tue Jun 9 10:26:35 EDT 2015


Hi Tom,
Thanks a lot for your feedback and for having been an inspiration for this project.

With a temperature-stabilized setup, it is now concievable to run the system at a constant impulse duration. This would not have been possible without temperature control, as the pendulum arm length varied too much to maintain an oscillation using a constant impulse width. I tried this overnight at a couple of occasions with an aluminium arm, only to find the pendulum dead in the morning...  :-o

So in order to achieve this, I will add a serial command to keep the last impulse duration. That command would be sent once the system is fully stabilized.
Thanks,
Bert, VE2ZAZ

Message: 8

Date: Mon, 8 Jun 2015 14:46:50 -0700
From: "Tom Van Baak" <tvb at LeapSecond.com>
To: "Discussion of precise time and frequency measurement"
    <time-nuts at febo.com>
Subject: Re: [time-nuts] Electronically Disciplined Mechanical
    Pendulum
Message-ID: <ED4D31631FC440A19A71631386F46143 at pc52>
Content-Type: text/plain;    charset="utf-8"

Hi Bert,

Thanks for posting that project. What a wonderful combination of electronic and mechanical timing, of design and measurement, of hardware and software, of PICs and Python.

One side experiment that would be interesting is to collect a couple of days of data using a fixed drive and then compare that with the same number of days using your adaptive FLL drive. The resulting phase or rate or ADEV plots would be amazing.

/tvb

----- Original Message ----- 
From: "Bert, VE2ZAZ" <ve2zaz at yahoo.ca>
To: <time-nuts at febo.com>
Sent: Monday, June 08, 2015 10:11 AM
Subject: [time-nuts] Electronically Disciplined Mechanical Pendulum


> Greetings,
> 
> 
> I just want to let those who are curious about disciplining a mechanical pendulum that I have pretty much wrapped up playing with a 1m-long pendulum, which I control with PIC micro firmware and try to mainain at a constant temperature. Accuracy for 20-second averaging is typically better than 1 ppm. I have documented my work here:
> 
> On my website ( http://ve2zaz.net/Pendulum_Ctl/Pendulum_Ctl.htm ),
> On Youtube ( https://www.youtube.com/watch?v=7NdGX4A8W88 )
> 
> Thanks,
> 
> Bert, VE2ZAZ



   



More information about the time-nuts mailing list