GearHeads Corner

I am new to this with limited knowledge but would like to make a clock in wood. the question I have is knowing how all the parts go together. I would like to make a mantle clock with a short pendulum
but I would like to be able to see the pendulum swing something like arts tuney clock with a spring winder. with little knowledge I need to know what shafts go with what gears in detail, ware to get parts and what parts and what they do ect. I have seen some sites like Clinton boyer and law clocks but I do not know if there plans are detailed enough for me I would like to see it put together so I can understand what each part is doing, and can not find any place that shows a build in de
tale I think once I have seen it I will have a reference point to confer with when I get stuck. and parts such as bearings rods ect how to find then in usa any help would be greatly appreciated thanks Don

Don:

  Making a clock isn't that hard, making one that works IS. :) , Id advise making it a work of art, wont mean so much when it doesn't run for a long time. Time is the thing that measures how well you built a clock or any ticking mechanism. Its a very delicate balance of input power, vs friction.
  The spring on a clock needs either a ton of power, or a very easy to drive mechanism..so you measure the clock by how long it runs on a wind. 

  The first thing you want to build is an escapement that works. Watch my video on a grasshoppers on youtube for example, you can cut that from wood, and a weight on a  string will make it tick away.. Id advice you to build one of each type of escapement , you can mount them simply enough of a piece of plywood on pegs, and get them to run. Once you make a few working escapements, you'll have a much better idea of what kind of clock you want to make. Small pendulum clocks tick faster than long ones, so they use up more energy, so it increases how well you must build it.. unless you use more advanced pendulum techniques. Best to start it off by building escapements though, its the best starting point, as well as the most difficult part to do in many ways...

Art



 

Don

Art is right regarding friction being the biggest enemy of real clock builders.  I've seen working wooden clocks at Mystic Seaport Museum of America and the Sea.  Quite interesting mechanisms.

My interest in clocks right now is documenting the design of some clocks I have access to by building  simulations of them.

If you are really interested in clocks, it wouldn't hurt if you learn about some of the theory behind clocks and their gearing.  A good, free online source that I've been using to understand clocks can be found at:

https://ia601403.us.archive.org/24/item ... ich_bw.pdf

I knew very little about gears or clocks until trying to understand some of the clocks that I have access to.  I found the following free resource a bit on the technical side...but the terms around the design of gears are good to know:

http://www.bostongear.com/pdf/gear_theory.pdf

Gearotics is a very powerful design tool for building gear related items especially clocks.  Under the circular gear module tool, check out the calculator function.  That is the one that you can either create a gear with a desired diameter or given a pair of gears to define the shaft distance between them.

The calculate diameter function is good for matching a gear sizes knowing the number of teeth in a gear pair.

The calculate shaft distance has a couple of applications.  If I can't measure the diameters of gears, but do know the tooth counts of the wheel & mating pinion and spacing of the gears, I can still create the gear pair.  If you plan to build a clock with concentric shafts it's good to get the spacing approximately correct using the calculate shaft spacing before using the Shaft-to-Shaft function on the Selected Props tab.

Play around with the various gear types and how the settings affect the resulting gears. 

Most of all have fun.  Don't be in a rush to build complicated mechanisms...start simple like Art suggested.

Bill

Hi Don:

Art gave you good advice, it's like a moving train and you have to get on at some point.  Mine started with Clayton Boyer's Number 6 clock.  It has been running for over 3 years but it still stops when the weather gets humid.  Clayton even says that things like that can happen so it doesn't bother me (he even jokes that one of his clocks is so sensitive to changes in humidity that it stops when a sweaty person walks by).  I also built Art's Scimitar ticker and you can learn a great deal by building that as well.  I just finished a Brian Law clock (#24) and I would not recommend that you tackle this as a beginning project because it requires special machinery like a metal lathe.  In contrast, Clayton's Number 6 uses easily obtainable metal rods and no bearings.  It does require a great deal of attention to scroll sawing (better yet CNC) and accurate hole positioning in the frames.  Clayton's plans look complex when you first open the package but careful study several times as you proceed makes things clear.

Like Art said, it's a delicate balance of input power vs. friction.  Subtle changes in how well a gear tooth is sanded can make a world of difference in whether a clock runs or not.  At the very least, you should have a drill press, a scroll saw or CNC device, lots of fine grit sandpaper, small files and a means of cutting small diameter steel and brass.  That and extra hair to pull out when things don't go right.

Good luck!

Mark

I'm in the process of building my first clock and learning what some of the problems are. The wheels were designed in Gearotic and cut on my newly constructed CNC router. The rest of the structure has been trial and error. Mostly error  :)

I'm still trying to get a sufficiently friction-free running train but have learned one important lesson already: Don't cut the wheels out of MDF! The mess is horrendous, the fine dust is all pervasive and toxic and the finished surface of the teeth looks suspiciously like glass-paper. Smoothing them by hand then creates even more micro-fine dust.

I've been experimenting with using arbors and bearings made from K&S brass tubing, the sizes of which are designed to telescope, but am beginning to understand why the professionals allow the arbors to run in holes cut in the wood of the frames instead. My next iteration will be the same set of wheels and arbors running in holes drilled in a hardwood (Jarrah) frame. I may try using ball bearings for the shaft carrying the weight-driven drum. This shaft has much higher forces on it's bearings than the others and is therefore much more likely to bind.

Good luck with your own project.

Kit

Hi guys thanks for the reply's I will take arts advise and go slow even though it is killing me lol. But I am still wondering even though I am not there yet, do you know of any videos on a complete build would love to see finished parts and ware they go. Art I see on another post on here you said something about your clock 2 I think it is I see I can download all the files and cnc them I am not at this point  looking for a perfect build would just like to complete one like you said for an art form but once again even in your videos it does not help me know how to put all the parts together. Also one of the limitations with v_carve desktop is I can only add one stl at a time machine it and then I have to load the second and so on in your clock 2 I would have to do this 20 some times what a pain in the butt that is one reason why I asked you the other day about software I can't combine multiple imiges in this soft ware. I know I am putting the cart before the horse but would love to build something cnc is a drug and I am hooked need a fix lol. Just to let you know the way I have been learning in v-carve is watching there videos and then writing down what they set thing at and copying them it is a bit of a pain but I find after I do it things become clear to me and then I can do it myself so that is why I am asking to see complete builds something I can follow and copy and build weather this is the write way or wrong way to learn it has been working for me once again thanks guys for your input  Don. 

Don:

 

KK, lets start with the STL problem, to make gears and such you don't use the
STL output, you use GCode output, or, if you wish to path them in Vectric, you load
DXF's put out by gearotic. To make things like a clock you can cut many gears at once,
and gears cut very quickly as they are all 2.5D.

 You machine STL's only for 3d objects , for things like Gears and such, stay with DXF's , or have
gearotic put out the Gcode for your machine.

 As to how to put a clock together, there is no real right way..or wrong way. A clock is basically
two gear trains. An escapement sits between the two. The first gear train takes a weight or a spring,
and then ratio's that spring way way down till it reaches the escapement.

   When you design your escapement, you will decide how many turns it takes per minute. The
second hand of many clocks sits on the escapement wheel, such an escapement then would
have to rotate once per minute. Some may put the second hand on a gear off an escapement
that rotates 2 times a minute to slow it back down to 1/minute.

  So lets say you build an escapement, ( always the best starting point ), you can then determine
just how much weight it takes to drive it on ..say a pulley on the escapement shaft..

  From there you can calculate how much weight or spring you need to drive such a thing for 1 week.
This is why you need to start with the escapement with your first clock. You need to know how
well you did. So you build such an escapement, put it on a plywood board, then hang a weight
on it and let it tick. So lets say it takes 3grams of weight to make it drive reliably, thats the minimum
you find that will work. You now know that if you wish to drive it for a week, a pulley with 3 grams will
have to turn about 10000 times before the spring needs rewinding. (minutes in a week.)

  Well, obviously you cant have a pulley with 10,000 turns of string on it so you need to put the
pulley on a gear at say 5:1 ratio, now you need only 2000 turns at 15grams.. , hmm, still bad,
lets add another say 8:1 on that.. 250 turns with a weight of 120 grams.. Still too much..
etc...etc.. etc.. till you find a point at which you can drive a pulley by spring or weight and the escapement
has the proper weight on it to drive.. You have to add weight to the original 3grams above to
accommodate the fact your weight will also have the drive the second gear train that comes off
the escapement to drive the hands of the clock.

  Now this is all backwards, any clock designer would I think make the second part first, then figure out the
force required to drive it, then make the weight gear train.. BUT, for a first clock, Id go in the order I describe, firstly because then you get used to how an escapement works, secondly, its more fun to see it tick as you go and fudge your weights and such around..

 Now, finally, you ask where do you put things.. the answer is wherever you wish. When I do such a thing,
I will put the gears on separate shafts ( dowels in plywood ) so each one meshes with the next gear..

 You end up with a long train of gears.. THEN, if it all works and ticks and such, thats the fun part,
because gears will fit in any place you like , design a set of plates to hold them folded in on themselves
to make a compact train.

 Now there's lots more to it than that, you have to make sure pairs of gears have the same shaft
distance so you can share shafts and such, but when you have them all on a piece of
plywood running between pegs, its easy to redesign and replace sets of gears..in various ways.,

  You will always end up cutting 3 times the number of gears you will use in the design of your own clock,
and there is tons of experimenting as you go .. but its fun and rewarding to do it.. and you learn alot about
clocks and the respect due those who do it well, like Boyer.
 Its a totally different process from cutting someone else's proven design.

So get a 2'x2' piece of plywood, some wooden dowels, and use the DXF's from gearotic to cut
the gears. ( A gear usually takes about 5 minutes to cut..and youll end up cutting dozens..lol)

 Place the escapement on pegs, the only hard one is the pendulum shaft as if has to be able to switch a pendulum.. Once your comfortable with it running, you'll begin to see how adding a plate to the front to hold shafts between the front and the back plates, will allow you to fit the gears many different ways.

  Many who built their first clock do not do things like have the second, minute and hour hands on the same shafts, that takes more experience, so long as you can make one gear in the train rotate once per minute,
another once per hour, and another once per 12 hours, you can be quite happy that looking at the three
hands, even if their 6" apart, tell you the time. If it will run longer that a day, then your pride quotient
goes up, hit a week and Ill be asking YOU for advice..

Thats my advice..
Art



Art



 




   

I think you've got some really good advice in this thread.  There are a couple of points that I might add. 

For a first clock I would aim for would be one that runs only 12 hours and I say that because it will only have an escapement and as few as 2 or 3 gears in the  drive train.  To build one that runs a week requires at least 4 gears in the drive train and a lot more weight and friction problems.

  Friction is your major challenge and the fewer actual "rubbing" parts the better.  For me that even includes bearings - certainly they can be added but don't necessarily make the build any easier or reliable.  I have a grandfather wooden clock that uses wooden dowels and wooden holes in hardwood and it is one of my most reliable clocks. 

Clayton Boyer's designs are great and they work as described - I've built about 6 of his designs and they all work.  (except for high humidity)

I've also found that the most successful clocks for me, have a lot of backlash i.e. they feel sloppy.  And for me the best tooth form is epicycloid as it is most forgiving as to "depthing"  (that is making sure your shaft distances are most accurate).

The advice for the driving weight for your clock that was the best for me was from a woodworking magazine.  This builder says to get your clock running with the least weight possible and then add 30% to beat some of the humidity.

No matter what - Make some gears and let them run!

Some great advice appearing here, years of experience condensed into a few paragraphs. This is what the internet is all about.

Kit

Once again thanks for all the input from everyone that really helps. OK Art I am going to do your plywood dowel pin test but still have a few questions. I went on gerotig and chose recoil I watched some of your videos and chose a 6 inch with a 1.5 relief, and drop angle of 0.5 on a 30 tooth escapement gear. So I take it that 30 tooth will give me 60 seconds one second per side of swing on pendulum is this correct. Now I saved them in dfx to be cut in vectric. When I am in vectric if I just select the shafts and make a drill path it should give me the proper distance between pallet shaft and escapement gear is that correct. If I am correct on all of this so far when I cut this to the plywood and put dowel pins in I should be lined up. Now what I am trying to understand is the pulley will give the escapement gear energy so I think I would glue pulley on escapement gear. Is escapement shaft made loose on escapement gear in other words escapement gear spin freely on shaft and if so dowel on plywood base would be tite.  Now for the pallet it sits snug on its shaft but not tite so would the shaft move freely in the plywood base, so would the hole have to be loose enough for the shaft to spin in it. next is the pendulum first I would have to make one, and how is it located I take it that it is connected on the pallet shaft but how loose hole on pendulum and pendulum connected to pallet so they move as one or tite fit so when the pendulum swings so does the shaft. So I hope you are following all of this, bottom line is when I said in earlier post ware things go I meant what gears move with the shaft and what gears move freely on the shaft and placement of the pully , pallet, pendulum ect. I hope this is making sense before I confuse myself an you lol, If my escapement assembly is correct how do I make it run my thinking is pulley connected to escapement, weight on pulley gives energy pallet, pallet holds energy until pendulum swings to release energy. how is that for an explanation. Thanks for the patients Art.  Don       

Don:

Your final statement is correct.

The pallet is tight on a shaft which is loosely suspended between a front and back plate. The shaft protrudes out the back..

In the simplest system you can make, ( though not very efficient, you can hang the pendulum right on that shaft in the back.
( You ll want it snug, but adjustable as to position to "beat" the clock in. Beating the clock is the process of centering the pendulum to the pallet so that its even on each side. ( The Tick should be the same as the Tock...). You wont hang the pendulum on a final clock that way, youll instead hook a crutch and yoke to that shaft and it will nudge the pendulum,which will be balanced like a pivot on its own shaft for frictionless motion.. but I wouldnt worry about that until you get preliminary ticking.

  So yes, attach a pulley to the escapement wheel, and loop string on it to a weight. then swing the pendulum and if there enough weight, it will tick.. Id then switch to a proper crutch and yoke to make the pendulum go.. but again, only after
you experience the first ticking from a set pendulum, which is going to take more weight to drive is all..

The length of the pendulum will control the actual time period. Once you have one ticking, you'll begin to more fully understand
what you have to do to build a complete mechanism..

Art

Thanks Art I made a escapement and pallet yesterday and hung a brass 1/4" rod with a weight on the pendulum shaft when I swing the rod it works well moves the escapement gear in clockwise direction no binding very happy at what I did. Now I have to move on to the pulley. You are write when I see it working I get a better understanding on how it works. I will start with the pulley sounds pretty easy, and then I will need help to make the proper pendulum arrangement. I have another question when I watch your video on gear trains you skip right past how you came up with the number of teeth, is that important. I would like to start making the first gear train in gerotic. can you help me with some numbers I take it I will be going from escapement to one min gear and that the escapement gear with 30 teeth is doing 30 sec in one rev so it has to go around twice to get a min. I would like to cut those gear to see how they interact with what I have built so far, i do not think this would take long to do just need some help with what numbers to put in and ware to put them in gerotic. I love watching this thing tic now I will see what the pulley will do thanks for all the help you are giving me. Don 

Don:

Here are a couple of photos showing the pendulum-crutch arrangement that Art mentioned.  In this case (Clayton Boyer's Clock #6), the crutch is attached to the pallet shaft and is a tight fit so it's adjustable.  The pin extending from the crutch goes through a slot in the pendulum with only about .010" slop between the slot and pin.  The driving force on the escape wheel rubs against the pallet faces and gives a tiny push to the crutch pin with each beat, thereby keeping the pendulum going.  A lot of new clock builders think that the pendulum drives the escape wheel when, in fact, it's the other way around.  I highly recommend that you purchase a proven design plan and build it.  Consider the money spent as tuition on learning how all the parts go together.

Mark

Mark:

  That IS truly nice work, makes me want to drop everything and build another clock. :)

Harnden:

As to tooth counts, you dont worry about them, you worry about ratio's. All thats important is ratios..

SO lets say you know the escapement wheel goes around once in 1 minute. You then know you need
a gear somewhere to hang a minute hand on that is 60:1 ratio..it has to turn once per 60 minutes.

  SO using the gear train calculator, you can tell it how many gears you wish to use, and what ratio to get..
it will then give you all the gears you need to make that 60:1, you can then place a minute hand on the
shaft of the final gear. Make another at 12:1, and you have the final train leading to the hour hand..

  Mark makes a good point, a professional design is one where the designer has gone though all this
in hundreds of permutations and working various placements so its a great clock..

On the other hand, if your like me, you wont go that sensible, smart and learned way, lol, you
will instead design a gear train and place it on to see how hard it is to drive a minute hand.. Youll
be amazed to see it seems to work well and proceed to change your pendulum to a crutch as Mark showed..
  at which point youll begin to see why Mark suggested a full design.

  At some point in this process you will begin to see just how sensitive a device a clock is, and just how
much its a marriage between art and physics. Most have absolutely no idea. Until you've built one,
and experienced it, I dont think one can appreciate just what a miracle they are,moreso when made from wood.
. But your doing it right, and its a great journey..

  In the end, of all the things Ive seen built since I started Mach3, nothing has impressed me more with
a persons craftmanship than clocks. Make a 6 hour and you did well, 12 hour and one should be proud,
24 Im envious, and a week , well, I bow in your direction and welcome our new overlord...

Art

Hi Mark
I totally agree with Art.  In my case, I build with wood, nearly always with Baltic Birch because it is dimensionally the most stable I can get. The one thing I always watch is the mod. I want biggish teeth because they are more forgiving to build. I like about 4 teeth per inch because its just plain friendly.    Having said that. As Art said make some and see what you like.

John