[Vinnie] has a wonderful old clock from his grandmother; it’s an exquisite antique with a real mechanical movement and a charming set of bells that ring every hour. Unfortunately, those chimes are a bit of a disturbance to neighbors at 2 o’clock in the morning. Previously, [Vinnie] had been stopping the clock every evening, and hoped he would remember to start the pendulum in motion 12 hours later. This was a chore, so he decided to automate the process.
The build is simple and clever; a small stepper motor is mounted in the clock just underneath the pendulum. Every 12 hours, the stepper motor moves a lever and slowly stops the pendulum over the course of a dozen or so seconds, silencing the clock movement. Twelve hours later, the motor turns again setting the pendulum in motion.
The parts count for this build is very low – basically just an ATmega88, a Darlington array to drive the stepper, and a 32.768kHz crystal. We can think of a few friends and relatives with loud clocks in their house, so we might have to build a few of these to give away.
Take a look at the demo video after the break to see how [Vinnie] stops his grandmother’s clock every night.
[youtube=http://www.youtube.com/watch?v=FZ_Zd7TeMg0&w=470]
Would’t it have been a bit more elegant to use the motor to put a damper on the bells instead of stopping the entire clock?
It has to be restarted exactly 12 hours later… Hmmm…
I second Dax on this one.
I imagine that using a servo to limit the travel of the bell hammer(s) would be far more effective here.
personally i’d prefer this, i cannot sleep with a loud mechanical clock, think its a fantastically elegant solution, and any drift in the time can be easily accounted for in software.
@denis
Any drift in the time can be accounted for in software? Really?
This is not an exact technology by any means, we are talking about an analog device being started and stopped by a digital device.
let the clock run for(say) 10 days with the 12 hour stoppages. compare against an accurate clock, adjust the stoppage time to make up the difference(assuming there will be one). its open loop, and a bit cludgy. so yes. really :)
Presumably the stepper can increase and decrease the speed on the pendulum. Reed switch on the pendulum and we can monitor the rate at which the clock is ticking averaged over 30 seconds or whatever. With the addition of a GPS unit the clock can monitor it’s drift and self correct.
Pendulum swing period is governed by physics. A specific length pendulum swings at a specific period (ignoring air resistance, also a constant.) No microprocessor or stepper-motor intervention will change that. Swing it faster, it swings higher, but the period stays the same.
I am a professional watchmaking and clockmaking student, and I can tell you that he wasn’t the only one with this problem.
Think about the chiming clocks that existed for hundreds of years before this- people then ran into the same problem. Traditionally, this was solved with a silencing lever on many clocks, that simply pushed the hammers away from the clock’s gongs either by a small lever, or pull cord coming from the clock to the master’s bed. They shut off the chiming at night, and restored it in the morning.
Stopping the clock is BAD for the clock’s accuracy. Don’t expect it to keep good time if you do this.
Being what I am, I’d like to see him put a mechanical solution in place, as has been done for hundreds of years, without the normal “let’s throw a microprocessor at it!” mentality.
It’s simple- make a piece that either goes between the gongs and hammers at the pull of a lever or cord, and thus deadens the blows to nothing, or make a piece that pulls the hammers away from the gongs, to where they do not strike.
Many books on striking clocks show this, and any decent striking clock already has this in it.
I also realize few qualified clockmakers exist anymore (I am quite young), so I understand why he went the microprocessor route. Programming is the new mechanics easily learned by my generation- actual levers and gears mechanics is regulated to the dustpile now, and no one knows how to make things by it.
I just wanted to give a mechanical perspective to the problem, and show there are and were ways of solving this period proper WITHOUT arduinos and such.
I forgot- do not attempt to modify the clock if it is a genuine antique worth money. If it is a run of the mill, just old clock, try it.
There may be a silencing mechanism on this already, and he hasn’t noticed it, or it doesn’t work. It may have none.
If you can, use whatever works for you without harming the clock- and see if you can take it to a local clock repairer- or:
There is an organization called the NAWCC, a National Association of Watch and Clock collectors. They meet all over the US. Call them up in Columbia, PA at the national watch museum, and see if they can’t direct you to a local chapter of their members. Those people, a proper repairer among them there should be. Not all of them are, many are enthusiasts that bodge things up. Get in touch with the NAWCC, and they should be able to help you find someone near you that can help you mechanically solve this.
Just remember- stopping the clock this way will harm it’s long term accuracy, and there are tradtional mechanical solutions to your problem!
Last post (promise)- I know he has tried blow deafening items between the gongs and it didn’t work out- the traditional method pulled the hammer or the gong away from each other.
That said, if you were to have no other mechanical recourse, and can’t find someone to fix this, the idea of using a microcontroller this way looked pretty cool to me. It will harm the clock’s accuracy and lead to oils gumming up more quickly, but it’s an interesting solution.
Couldn’t he just use this setup to interpose a piece of metal, for strength, covered in felt, between the hammer(s) and gong(s)? The hammer hits the felt, no sound, but the metal stops any connection with the gong.
It says he tried putting something between the gongs and still heard the muffled hits. If it’s right near his bed, it would wake him.
He could try thick rubber, but honestly, the best solution is a mechanism that pulls the gongs away from the hammer or lets the striking mechanism slip.
As a longstanding(read OLD) clock collector, I admit that the night feature on my Eliott Westminster clock in our bedroom is an elegant solution. It has a Cam rotating in 24 hours that suppresses the strike train for 8 hours currently 11pm to 7am, but could be adjusted to midnite to 8am. This works like a default alarm for us that sometimes gets ignored.
For your purposes, I would suggest a mechanism that would lift the hammers off their lifting pins for the desired time, like many Chime-Silent levers work, then the only noise you might hear the drop of the snail and the whirr of the strike train.
If you are worried about hearing the TAC-TOC all night, let me assure you that eventually it will become a sleep inducing sound.
Regards, Bob
One thing I don’t like about this is that since the clock is stopped at night, you have to restart it at some point, and reset the hands to the current time. I’d put in some kind of circuit board to operate the hands when the clock isn’t using the pendulum.
The article says the motor is set to restart the clock 12 hours later, when the hands would be at the same position.
I am with Drew on this one…
I am an intermediate student clock collector and repair man. To put in a mechanical lever to pull the hammers away is way more logical, and won’t ruin the antique elements within. Just putting a cheap plastic microprocessor inside does not match generations. And also it’s easier just to pull a lever and have accurate time and to keep it in the same generation, I would do a lever.
Sorry for the run on sentence. ;-;
I like the lever idea. Is there someplace that shows how this is done?
I did a similar thing with my mechanical clock except the device that installed interrupted the molecular structure of the clock, causing it to both exist and not exist as the separated atoms were still held in relative proximity to each other via an Ekotin Force Field which was programmed to shut off at 7 a.m., thus allowing the atoms of the clock to reintegrate, with the clock functioning normally. (I made that up.)