Created 3 June 2016

This is a side note to the Bürk Mutteruhr / Hauptuhr / master clock characterization project.

I have made some rough manual measurements of the Bürk clock’s stability.

I have used the HP 5335A Universal Counter and my photo-interrupter setup to aid in adjusting the rate regulation nut on the bottom of the clock’s pendulum.

The weather has been quite warm over the last week or so here in Northern Germany. Over the last year or so I have noted how the change from winter to summer does make quite a difference to the rate of the clock. I have a BME280 temperature, humidity and air pressure sensor inside the clock but sadly I have not had the time to set up logging of its data on the connected Raspberry Pi.

Rough observations
I am taking my time with this project. For now, after resetting the clock rate with the counter setup, I have been manually and roughly tracking the difference between the Bürk and the time reported by a Trimble Thunderbolt GPS Disciplined Oscillator. This is very rough. I’m simply looking at the time, to the second, being sent over the Thunderbolt’s serial port and counting the difference between it and the Bürk when the Bürk hits the top of a minute.

Over the last year the Bürk has been a bit all over the place from day to day when I have used this manual method of comparison. I think that is the result of the gas powered heating system in the apartment. I haven’t checked but I would say that the variation in indoor ambient temperature is greater in the colder months than it is in the summer in this apartment. In summer this apartment gets hot and stays hot.

At at 21h00 on 30 May 2016 I made a final adjustment to the regulation nut and restarted the pendulum. From the start the Bürk was 1s fast (early) compared with the reference Thunderbolt. About 12 hours later it was the same. On the second day at 22h00, the Bürk had slowed down by 1s and was in sync with the Thunderbolt. Today, the fourth day, it is still in sync. I don’t think I have observed such stability since I fixed up the clock.

What I would like to do in due course is add a small temperature controlled heating device to maintain the temperature in the clock case at the highest temperature observed by the BME280 throughout the year. From what I have seen over the colder months, this should significantly improve stability. Maybe I could just use a small 25 watt oven light bulb in there.