Author Topic: Hello  (Read 15798 times)

Offline pcstru

  • Newbie
  • *
  • Posts: 22
Hello
« on: November 28, 2007, 08:55:56 AM »
Hi All,

I've been interested in clocks for quite a while and recently started mucking about trying to create some.

Grasshopper Escapement (Prototype)

Gravity Escapement, compound pendulum thing.

Egg Timer with a verge and foliot escapement.

Gary's 2003 clock with re-arranged geartrain layout.


I've been struggling a bit with the grasshopper. I've read through the PDF "anatomy of a grasshopper" which someone here linked to/posted. The prototype very nearly works for short periods! I design using pen/paper for quick throwaway sketches and then draw the layout using VCarve Pro, fit the components onto a sheet, set up toolpaths then generate a file for a CNC router. All but Gary's clock are made with 4mm 3 ply because it's cheap! All arbours are 6mm dowel and bearings are nylon, ptfe or 10mm/6mm roller bearings.

Offline bobledoux

  • Newbie
  • *
  • Posts: 42
Re: Hello
« Reply #1 on: November 29, 2007, 07:10:24 AM »
I've spent several months watching my grasshopper and getting it to work so I appreciate your questions.

Your pallet assembly has lots of mass.  This might contribute to some reluctance to work.

Make certain the mechanism is actually spanning the number of scape wheel teeth in the design.  Its easy to set up the mechanism to be a tooth too wide or narrow.  That adjustment is made by twisting the pallet assembly, relative to the pendulum.

In setting my clock I played with a lot of adjustments between the pendulum and pallet assembly joint.

I also increased the driving weight until the mechanism worked.  Then I reduced the weight until the mechanism stopped and added 25% for good measure.  You might try adding weight to the pendulum. 

Start by swinging the pendulum to the left to engage the right pallet.  Then swing the pendulum to the right.  As the pendulum approaches the right swing limit the left pallet will engage, causing recoil.  This recoil will release the right pallet.  As the  pendulum begins swinging left make certain the right pallet is adjusted so it just misses the end of the tooth it just released.

As the pendulum approaches the end of the left swing the left pallet will engage causing recoil.  This will release the right pallet.  Adjust the left pallet so it just misses the scape tooth above as it engages. Then adjust this pallet to achieve an even tic-toc.

Using this procedure check that the swing right and left of the pendulum are equal angles. If you can rotate the whole clockwork you should find a position where the escapement works. 

When properly adjusted it should not be possible to release both pallets from the scape wheel with the pendulum hanging centered.

Adjusting the grasshopper requires playing with a number of variables.  Its an efficient escapement but start with excess weight in the drive and pendulum.  Then play with placement of the pallets relative to the pendulum.  Finally, work on the individual pallet adjusters. 
 



« Last Edit: November 29, 2007, 10:39:03 AM by bobledoux »

Offline pcstru

  • Newbie
  • *
  • Posts: 22
Re: Hello
« Reply #2 on: November 29, 2007, 10:13:49 AM »
I've spent several months watching my grasshopper and getting it to work so I understand your questions.

Very useful. Many thanks.

I did get it sort of working driving with my url=http://www.youtube.com/watch?v=1V5QzY2uB2I]hand[/url] but will look to incorporate your suggestions for the next cut. My other question would be, if R2 is bigger than R1 according to calcs, such that AH, OG or both is not perfectly vertical, what constrains the position of the circle R2.


Offline bobledoux

  • Newbie
  • *
  • Posts: 42
Re: Hello
« Reply #3 on: November 29, 2007, 10:52:17 AM »
See my revised edits above. 

I built Clayton Boyer's Bird of Paradise:

http://lisaboyer.com/Claytonsite/Claytonsite1.htm

The Aydlett article show the two circles placed over the pendulum centerline.  Boyer placed the pendulum to the side, with the two circles not in vertical placement.  There may not be a constraint on circle R2. 

Rick from Manchester

  • Guest
Re: Hello
« Reply #4 on: November 30, 2007, 07:10:15 AM »
Pcstru,

I love your enthusiasm.  Looks like you have the geometry on the right track, just needs some refinement.  Your grasshopper is built like a champion weight lifter, he needs to be more like a delicate butterfly (I did the same thing in my early grasshopper prototypes).  The forces at the top of the clock train are very small, and the escapement can be made very light without fear of anything breaking.  You might want to try to reduce the overall mass to 1/4 of what it is now. Also, reduce the pull of your elastic band, again, think delicate, and very small forces.

From what I see in your photo and video, it appears that recoil is not being allowed to occur.  In your video http://www.youtube.com/watch?v=1V5QzY2uB2I, notice how the long pallet leg is sliding on the tooth (it's being forced back into the escape wheel by the adjustment screw), when it should be allowed to recoil a bit without any sliding at all, remember, sliding is friction, and when Harrison invented the grasshopper, his goal was to reduce friction to zero (I think the same thing is occurring on the small pallet leg too, but it's hard to see).  Where you have the stop screws, you need to replace them with very very delicate springs, or very small counterweights with their own stops.  Here's a video that shows the recoil taking place http://www.youtube.com/watch?v=rluQpoV8638, and this one http://www.youtube.com/watch?v=lDJvzg3J1mM&feature=related which shows the lighter construction. Yes, these two examples are of a different design, they use a coaxial escape wheel, but a single wheel design will work too, and the recoil and delicateness concepts apply regardless of the design type.

Here's an animation; http://en.wikipedia.org/wiki/Grasshopper_escapement however, the recoil is not really shown, but the thin pieces that the pallets drop back onto after leaving the escape wheel would be delicate springs which would allow recoil without causing the pallet to slide on the escape tooth.

I hopes this helps.

Rick.
« Last Edit: November 30, 2007, 08:08:52 AM by Rick from Manchester »

Offline pcstru

  • Newbie
  • *
  • Posts: 22
Re: Hello
« Reply #5 on: December 01, 2007, 12:44:32 AM »
Hi Rick,

Many thanks for those tips. I've recut the escape wheel and palletts and it runs a little better - even for a time being driven by a weight. I'll give it another go and try for more butterfly than sumo!

Piers



Offline pcstru

  • Newbie
  • *
  • Posts: 22
Re: Hello
« Reply #6 on: December 03, 2007, 04:28:57 AM »

Based on feedback Updated MKII Grasshopper. While the whole thing is much bigger than the last version, the palletts, arms and frame are more 'delicate' (although some more material could come off the frame). Accuracy has been improved a bit by using a vbit on the surface of the ply before cutting (the 4mm ply might be cheap but it's a pain to work). Still some work to do before I can test it with a pendulum.

Offline bobledoux

  • Newbie
  • *
  • Posts: 42
Re: Hello
« Reply #7 on: December 03, 2007, 07:03:00 AM »
Look at Clayton Boyer's "Bird of Paradise" or "Swoopy" Plans.  These use a grasshopper with fairly heavy construction and no recoil springs.  There is a utube movie of Swoopy running.

http://lisaboyer.com/Claytonsite/swoopypage1.htm