John Wilding Regulator

[Russell EberhardtParticipant @russelleberhardt48058]

"How did the dial surround/rim change from 'silver' to brass?"

Poor lighting, almost backlit from a window in sunshine!

"Will you add a brass sleeve to the pendulum bob?"

A nice idea but I will have to consider the effect on temperature compensation.

"What wood did you use for the case?"

I managed to find some sapele faced plywood locally for the panels and then ordered sawn sapele from Robbins Timber in the UK. I still have a homemade router table that I made for doing boat repairs many years ago so was able to use that to make the mouldings fairly easily.

Russell

[Stephen BensonParticipant @stephenbenson75261]

Lovely work and and I would also like to know more about how you printed the dial

[NJHParticipant @njh]

Russell

That is beautiful work throughout – something to look at everyday and get pleasure from.

BUT….."the case is now ready for varnishing:"

Please don't VARNISH it! Instead take the time to apply successive coats of button polish, with a rubber, to build up that lovely deep, smooth, shine.

I can imagine many restful hours in the company of your clock as it plays you its peaceful and comforting song!

 ……of course what are you going to do with your spare time now ?

Norman

 

Edited By NJH on 29/04/2016 12:11:11

[Russell EberhardtParticipant @russelleberhardt48058]

Posted by Stephen Benson on 29/04/2016 09:52:01:

I would also like to know more about how you printed the dial

O.K. here goes:

I started by drawing the dial using Draftsight although any 2D CAD system would have done. I then mirrored it to get a reversed print so that the printing on the transparency would be on the underside and directly in contact with the resist. I found that the printer would not print to the defined pen thicknesses so I printe to PDF and then used the PDF viewer to print the master on acetate film using an Epson printer in photo mode in order to get dense blacks.

I thoroughly washed the brass sheet in detergent and hot water to remove any grease and then wet laminated the resist film before passing it through an office laminator. The brass with the photoresist and the master was then sandwiched between an MDF board and a glass sheet with some thin translucent plastic foam to make sure there were no gaps between the brass plate and the master and expose to the sun for 3 seconds. I am in the South of France so I guess slightly longer exposure would be needed in the UK. The resist film came from Mungolux in Germany.

Once exposed the resist was developed in washing soda solution and then washed to give the result shown in my photo. Exposure and development times need to be determined by experiment.

The plate was then etched in ferric chloride for about an hour with occasional agitation. Then washed and dunked in caustic soda solution to remove the resist.

I sprayed it with black acrylic car paint to fill the etched portions and the paint on the unetched parts was removed using fine wet and dry paper and water. I then grained it with 240 wet and dry, fairly coarse but it looks appropriate on a large dial.

Again very careful washing before rubbing with silvering powder and then finishing powder. To protect the silvering I made up a lacquer from shellac dissolved in colourless meths (sold here as burning alcohol).

Throughout I did experiments on scraps of brass to verify each process. I now see why bespoke etched dials are so expensive to buy!

Russell

[Russell EberhardtParticipant @russelleberhardt48058]

Posted by NJH on 29/04/2016 12:10:31:

Please don't VARNISH it! Instead take the time to apply successive coats of button polish, with a rubber, to build up that lovely deep, smooth, shine.

Yes, french polish is very nice but I have never had much success with it. I am using several thin coats of a good quality satin finish varnish (Liberon) and will wax over it. Certainly none of the modern water base varnishes.

Russell.

[Neil WyattModerator @neilwyatt]

Posted by Russell Eberhardt on 29/04/2016 20:05:04:

The brass with the photoresist and the master was then sandwiched between an MDF board and a glass sheet with some thin translucent plastic foam to make sure there were no gaps between the brass plate and the master and expose to the sun for 3 seconds. I am in the South of France so I guess slightly longer exposure would be needed in the UK. The resist film came from Mungolux in Germany.

Three days might do it.

Neil

[Russell EberhardtParticipant @russelleberhardt48058]

Finished at last:

Russell.

[Neil WyattModerator @neilwyatt]

[Michael GilliganParticipant @michaelgilligan61133]

Lovely job, Russell

MichaelG.

[AjohnwParticipant @ajohnw51620]

John

–

[Russell EberhardtParticipant @russelleberhardt48058]

My clock has now been running for over 8 months and after some teething troubles is giving quite good timing. Over the month of December it gained just over one second which is quite pleasing. The following is a plot of the going rate daily change over the month of December:

The green line shows my actual measured values using a stop-watch to compare with an atomic clock app on my tablet. The greater part of the random variation is probably due to my measurement technique.

The black line is a moving four day average to smooth things out and clearly shows an increasing rate, increasing by about 0.25 second/day over the month. My initial thought is that the friction in the clock is slowly reducing with running in.

Now comes my problem in understanding:

Intuitively I would expect reducing friction to speed things up. However reduced friction must result in a greater swing of the pendulum and that should result in a slower rate due to the circular error of the pendulum. So which is right?

Russell.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Russell Eberhardt on 02/01/2017 16:41:23:

… Over the month of December it gained just over one second …

…

The black line is a moving four day average to smooth things out and clearly shows an increasing rate, increasing by about 0.25 second/day over the month.

.

Forgive me, Russell … I have probably missed something significant; but I can't reconcile those two statements.

Could you please clarify ?

Many thanks

MichaelG.

[Russell EberhardtParticipant @russelleberhardt48058]

Michael,

Sorry,I don't think I expressed that very clearly. Must be the after effect of the over-indulgence of the holiday period.

The clock started the month 0.85 s fast. It lost time and was 0.26 s fast by the 18th of the month. It then started to go faster and by the 31st it was 1.98 s fast, all subject to measurement errors of course. The graph shows how much was gained or lost from one day to the next.

At the beginning of the test the average loss over four days was 0.04 s per day and by the end it was gaining a four day average of 0.20 s per day.

So what I meant was that in one month it had increased its daily rate by about 0.25 s.

During the month the mean daily temperature hasn't varied by more than 1 C. I still can't understand the effect of friction on rate.

Russell

[Michael GilliganParticipant @michaelgilligan61133]

Thanks for the clarification, Russell … I'm now pausing for thought !!

The clock looks lovely and is clearly performing well

…. A great start to 2017

MichaelG.

[JAParticipant @ja]

Russell

As mentioned elsewhere I have dug-out my regulator to work on during the cold weather.

I had made a John Wilding breakaway crutch but abandoned it for the conventional two pin crutch. What type have you used?

JA

[Russell EberhardtParticipant @russelleberhardt48058]

I used the two pin type. If the clock stops and the pallet catches on the escape wheel the friction between the pallet frame collet and the arbor is low enough so that the collet just rotates and the escape wheel is undamaged so I saw no advantage in the breakaway crutch.

Russell

[JAParticipant @ja]

Russell

Thanks. I think I will stay with the two pins type if only because I cannot find the bits for the breakaway crutch.

JA

[Russell EberhardtParticipant @russelleberhardt48058]

Just don't tighten the collet screw too much, so that it is able to slip.

Russell

[Author]

Posts