I think you need an escapement....
July 29, 2004 3:57 AM Subscribe
How do you convert the potential energy of a suspended weight into a constant, speed-limited, rotational motion? [more inside]
Before electric motors and light bulbs many lighthouses had large glass Fresnel lenses that were rotated by clockwork mechanisms that were powered by weights. I've looked far and wide online, and I can't find information on how to build such a system. There is plenty of information on horology, but clock escapements tick, and I want to produce a smooth motion. Is it simply a matter of balancing friction against the force the weight provides?
Before electric motors and light bulbs many lighthouses had large glass Fresnel lenses that were rotated by clockwork mechanisms that were powered by weights. I've looked far and wide online, and I can't find information on how to build such a system. There is plenty of information on horology, but clock escapements tick, and I want to produce a smooth motion. Is it simply a matter of balancing friction against the force the weight provides?
You probably want a centrifugal governor. Look at mechanisms for spring-loaded record players (victrola) and such.
Basically, you set up spring-loaded pads inside a drum that is stationary, but the pad springs ride on the rotating shaft. As the shaft speed increases, the centrifugal force pulls on the weighted pads on the ends of the springs, which will then rub against the inside of the drum, slowing down the shaft.
I'm sure there's complex calculations you can do to determine the ideal configuration for a given weight and desired shaft speed, but it would be fun to trial-and-error your way to a solution.
posted by yesster at 5:57 AM on July 29, 2004
Basically, you set up spring-loaded pads inside a drum that is stationary, but the pad springs ride on the rotating shaft. As the shaft speed increases, the centrifugal force pulls on the weighted pads on the ends of the springs, which will then rub against the inside of the drum, slowing down the shaft.
I'm sure there's complex calculations you can do to determine the ideal configuration for a given weight and desired shaft speed, but it would be fun to trial-and-error your way to a solution.
posted by yesster at 5:57 AM on July 29, 2004
This is just a guess at part of the solution, but it's obvious that the weight has to go up and down, and equally obvious how this motion can easily rotate a shaft that's parallel to the earth's surface. You need to change the shaft's orientation by 90 degrees to drive the lamp, but there's no law that says it has to be rotational motion that's carried through the 90 degree turn. Much easier to run the cable carrying the weight over a pulley to change it's motion by 90 degrees from vertical to horizontal before you ever try to rotate a shaft with it. That way the cable can rotate a vertical shaft directly.
posted by jfuller at 6:04 AM on July 29, 2004
posted by jfuller at 6:04 AM on July 29, 2004
Have the lamp on a flywheel that gets driven by clockwork? Doesn't seem that far out or complicated to me.
posted by fvw at 6:12 AM on July 29, 2004
posted by fvw at 6:12 AM on July 29, 2004
i would guess that lighthouses did tick. if there were another way of getting constant motion then you'd find it in clocks (since what you're asking for is a pretty precise description of a clock).
posted by andrew cooke at 6:23 AM on July 29, 2004
posted by andrew cooke at 6:23 AM on July 29, 2004
This page describes an escapement, which regulates rotation/falling. (page may need IE). Here's another one in lego. This would govern the rotation. An the gearing should do the rest.
I remember my grandparents had clocks based on uneven weights, you offset the weights and as they ran back to their original position it drove the clock. I can't remember what that kind of mechanism was called though.
posted by Flat Feet Pete at 7:11 AM on July 29, 2004
I remember my grandparents had clocks based on uneven weights, you offset the weights and as they ran back to their original position it drove the clock. I can't remember what that kind of mechanism was called though.
posted by Flat Feet Pete at 7:11 AM on July 29, 2004
Wow, you engineer-folk just rock. God bless Asperger's syndrome.
posted by leotrotsky at 7:31 AM on July 29, 2004
posted by leotrotsky at 7:31 AM on July 29, 2004
Depending on what you want to do, you've got lots of options. The first decision you have to make is do you care about efficieny or not? Does all the power of the falling weight have to go to doing your task, or do you just want the rod to turn at a certain weight?
If you want to reduce rotational speed without regard to efficiency, then a governor is what you want. There are many types. The simplest is a cetrifugal one, based on the same principle that a figure skater uses to control her spin speed. Viscous ones, an impeller rotating through a viscous liquid, are more compact and reliable. Magnetic ones also work well---essentially these are motors running in reverse, dumping energy through a resistive load. Magnetic ones are very easy to calibrate (just change your load resistor).
If your do want constant power transmission, but a lower speed, then you want a reducing gear system. This works best if you have a constant input speed and output speed, or know the gear ratio you want. Getting a constant speed with just gearing is very difficult---that's why clocks use escapements rather than just gearing everything down. McMaster-Carr sells gear boxes, or you might be able to scavenge bicycle parts, depending on your application.
If you want to limit the force transmitted by the shaft, then you want a torque limiter, a clutch. McMaster-Carr sells them too, but you might find what you're looking for in a replacement drill chuck at your local hardware store.
posted by bonehead at 7:41 AM on July 29, 2004
If you want to reduce rotational speed without regard to efficiency, then a governor is what you want. There are many types. The simplest is a cetrifugal one, based on the same principle that a figure skater uses to control her spin speed. Viscous ones, an impeller rotating through a viscous liquid, are more compact and reliable. Magnetic ones also work well---essentially these are motors running in reverse, dumping energy through a resistive load. Magnetic ones are very easy to calibrate (just change your load resistor).
If your do want constant power transmission, but a lower speed, then you want a reducing gear system. This works best if you have a constant input speed and output speed, or know the gear ratio you want. Getting a constant speed with just gearing is very difficult---that's why clocks use escapements rather than just gearing everything down. McMaster-Carr sells gear boxes, or you might be able to scavenge bicycle parts, depending on your application.
If you want to limit the force transmitted by the shaft, then you want a torque limiter, a clutch. McMaster-Carr sells them too, but you might find what you're looking for in a replacement drill chuck at your local hardware store.
posted by bonehead at 7:41 AM on July 29, 2004
Yesster kind of gets at my answer, but what he describes is a brake, not exactly what I had in mind. Here's a description and diagram of the flywheel-style governor that I've seen in action. I admit I don't quite understand the physics involved, but I've seen them work--they're interesting to watch.
posted by adamrice at 7:45 AM on July 29, 2004
posted by adamrice at 7:45 AM on July 29, 2004
Andrew: most cheap modern clocks are simply a constant-winding electric motor run at constant load, geared to the correct speed. There's no mechanical escapement and only a couple moving parts. It's only the more evpensive modern clocks that have an escapement (usually still driven electrically).
posted by bonehead at 7:47 AM on July 29, 2004
posted by bonehead at 7:47 AM on July 29, 2004
adamrice - it wouldn't function as a brake, because if the shaft slowed down, the weighted pads would have less friction with the drum. It really is a governor, not a brake. And it would probably work best on a vertical shaft, rather than a horizontal one. Use jfuller's suggestion for that.
posted by yesster at 8:27 AM on July 29, 2004
posted by yesster at 8:27 AM on July 29, 2004
Response by poster: Thanks a lot everyone. Adamrice, that diagram is pretty much exactly what I had in mind, but I didn't know what it was called. Bonehead, that was an excellent response, thanks. My initial idea was to use a gear system, but I remembered seeing a working clockwork lighthouse many years ago that used a series of what I now know were flywheel governors, and I wanted to see if that would be a better system. Since efficiency definitely is an issue, I might go back to gears and spend some time balancing. Thanks for the record player tip, Yesster, I'll definitely do that.
posted by Nothing at 11:35 PM on July 29, 2004
posted by Nothing at 11:35 PM on July 29, 2004
Response by poster: I also received a great response via email from someone who is not a metafilter member. I'm posting the body of it here (with permission) to add to the usefulness of the archive.
"...
I couldn't find any detailed plans on the internet for clockwork motors with
mechanical speed governors, much less ones specifically intended for use in
lighthouses. However, I know that millwrights developed several devices for
sensing speed of rotation and that these sensors were used later in several
speed regulation devices. In particular, these devices were later applied
for use in the regulation of the steam engine - the most famous being Watt's
centrifugal flyball governor.
"However, I was able to find a link that contains a photograph of a
lighthouse's clockwork motor and governor in operation:
www.hollenback.net/index.php/PointReyesLighthouseInside
"The photographer provides a description: "The blur in the middle of the
picture is the speed governor, which consists of a set of fan blades
attached to a weight. The angle of the fan blades varies with the speed of
the lens mechanism. As the shaft gains speed, the blades turn towards
vertical to bite more air and slow the shaft down. As the shaft slowed, the
blades would then rotate towards the horizontal and reduce the drag on the
shaft, allowing it to speed up." His description sounds like a fan
governor, which is used in some mechancical grandfather clocks.
"Also, in my search I ran across some diagrams of the simple speed governors
used in phonographs that were powered by clockwork motors:
www.clpgs.org.uk/speed_regulating_detail.htm
..."
posted by Nothing at 3:25 PM on July 30, 2004
"...
I couldn't find any detailed plans on the internet for clockwork motors with
mechanical speed governors, much less ones specifically intended for use in
lighthouses. However, I know that millwrights developed several devices for
sensing speed of rotation and that these sensors were used later in several
speed regulation devices. In particular, these devices were later applied
for use in the regulation of the steam engine - the most famous being Watt's
centrifugal flyball governor.
"However, I was able to find a link that contains a photograph of a
lighthouse's clockwork motor and governor in operation:
www.hollenback.net/index.php/PointReyesLighthouseInside
"The photographer provides a description: "The blur in the middle of the
picture is the speed governor, which consists of a set of fan blades
attached to a weight. The angle of the fan blades varies with the speed of
the lens mechanism. As the shaft gains speed, the blades turn towards
vertical to bite more air and slow the shaft down. As the shaft slowed, the
blades would then rotate towards the horizontal and reduce the drag on the
shaft, allowing it to speed up." His description sounds like a fan
governor, which is used in some mechancical grandfather clocks.
"Also, in my search I ran across some diagrams of the simple speed governors
used in phonographs that were powered by clockwork motors:
www.clpgs.org.uk/speed_regulating_detail.htm
..."
posted by Nothing at 3:25 PM on July 30, 2004
This thread is closed to new comments.
If a lighthouse runs a little bit fast or slow probably doesn't matter but I would guess that the period between rewinds needs to be somewhat predictable for the poor lighthouse keeper.
I wouldn't be at all suprised if a search of old patents yielded some actual mechanisms though.
posted by substrate at 5:46 AM on July 29, 2004