Tick tock

[ruste...@prototribe.net]

Its alive!

http://www.prototribe.net/vidplay/prototypeclock2_demo1.html

I'll be curious to see if its ticking tomorow morning (its on a older a
somewhat irregular escapement wheel). Currently it ticks 12 times in 10
seconds so is roughly 20% too fast. And unfortunately i noticed that the
longer and heavier the pendulum gets means the drive weight must also
increase.

Working on the hour/minute number plates and still want to find ways of
reducing the weight since we're at around 10lbs currently.

[Eric Hagan]

Wow, that is beautiful. 

How long is your pendulum and how much weight did you put on the pendulum itself? It is hard to tell with your escapement gear how often it goes around since it is spinning so fast, is there anyway to place a stripe of some variety with contrast?. I've mostly been working with pendulums just under 10 inches in length (1 second per period) and I am amazed at how fast yours is.

If've also been attempting to use springs instead of a weight if anyone has any advice to offer.

Thanks,

Eric

[Mathieu Glachant]

Smooth as silk! :-)

Do you have a video with the pendulum installed and the escapement ticking?

On Sun, Feb 26, 2012 at 2:30 AM, <ruste...@prototribe.net> wrote:

[ruste...@prototribe.net]

Ack! A typo! This is what I meant to post!

http://www.prototribe.net/vidplay/prototypeclock2a_demo1.html

[ruste...@prototribe.net]

Arg! I'd marked the drum at the 12'o clock point to see how long the clock managed to tick over the night and now I can't tell how long that was! A failure mode I wasn't expecting, the teeth stripping from the hour gear permitting the drum to spin freely, apparently snapping the end of the line, letting the 10+lbs of weight scatter on the floor!



Fortunately a new drum and hour gear was planned anyway as they need to be able to support a front disc with the numbers for the hours (raised numbering currently colored by sharpie).

[Mathieu Glachant]

Yeah, we saw that failure mode during clockathon 2, which is why I'm working on a better ratchet design. So little time!

[Mathieu Glachant]

Sorry, couldn't watch this on the iPad, so had to wait until I found
an actual computer.

Looking at the escapement, it looks like there is never a dead phase,
one where the pendulum is free the swing with the escapement just
sliding on the tooth, imparting no impulse, with the gear train
immobile. This is probably because the escapement pallets are too
wide, and so the teeth are always landing on one impulse face or the
other, rather than landing on a dead face, and slipping onto the
impulse face on the way back.

This is bad because it means that it is always the escapement that
stops the swing, not the natural period of the pendulum, and that's
why the pendulum's swing looks a little weird... it doesn't really
slow down at the end of the swing. This makes it hard to make the
clock accurate, because the period depends more on how fast the
escapement wheel can spin and hit the other impulse face. In other
words, more weight driving the gear train would speed up the clock.

Normally, this can be solved by making the impulse faces narrower, and
the swing of the pendulum wider, but I'm not sure if there's room in
there for the escapement pallets to slip in between the teeth for the
dead phase, with so many of them on the wheel.

It's hard to describe, but easy to see, I think, if you look at some
of the graham escapements animations on the web.

Or I could be wrong. :-)

[ruste...@prototribe.net]

Ah, I think it should be simple enough to make the pallette faces
shorter. I did notice the teeth hit the palette faces about half way.
Will add that to my list of updated parts for this evening.

Thanks!

[Mathieu Glachant]

De nada! If you get a thirty tooth escapement working, I'm going to steal it for the four gear clock! ;-)

On Sunday, February 26, 2012, wrote:

[ruste...@prototribe.net]

Well, I switched to an equidistant lock design for the escapement and seem to have added the possibility of a dead phase.



The left palette face seems to work great, though the right side causes a little reversal in the wheel so some tweaking may be in order.

I'll try to post video later today. From what i'm seeing so far it does seem like it should be possible to make work.

[Mathieu Glachant]

Yes, we had an issue with our escapement where the arm was hitting the tooth's side during the dead phase, driving the clock backwards for a little bit. This is a bad for a bunch of obvious reasons.

We solved it by increasing the angle of the teeth, which moved their side out of the way of the escapement. The same might work for you?

[ruste...@prototribe.net]

I'm prone to attempt to leave the teeth as-is and start by tweaking the
palette faces. I think if I extend the right palette face (following the
arm's arc) by ~0.5-1mm It should get ahead of the teeth to prevent the
backwards wheel motion. Actually looking at the pallets as I type this I
can see the facets in the arm's arcs. I'm wondering if increasing the
$fn value might help here with more rounded arms.

On 02/27/2012 11:09 AM, Mathieu Glachant wrote:

Yes, we had an issue with our escapement where the arm was hitting the tooth's side during the dead phase, driving the clock backwards for a little bit. This is a bad for a bunch of obvious reasons.

We solved it by increasing the angle of the teeth, which moved their side out of the way of the escapement. The same might work for you?

On Mon, Feb 27, 2012 at 11:00 AM, <ruste...@prototribe.net> wrote:

Well, I switched to an equidistant lock design for the escapement and seem to have added the possibility of a dead phase.