Very interesting, thx for sharing. I might have a project after watching your video. I would like to make a pendulum that will give me 1 second. Basically, I want to elevate .500 steel.ball 18 inches high. I'd like to make ball clock without an electric motor Thx.
Very interesting results. There does not appear to be constant rotational velocity. Do you think the gears are deforming under load? Or is the active surface different than expected after printing? The elephant foot in the closeup would not help either. I did an experiment using unloaded gears that would be found closer to the escapement. It used 4 stages of 50 tooth gears driving 10 tooth pinions for a total speedup of 625:1. A weight just large enough to move the gear chain was placed on the rim of the first gear. Cycloid gears and 25 degree pressure angle involute needed similar weights of 56g. A 14.5 degree involute was 20% higher at 68g. A modified cycloid with the teeth tilted so most of the touching occurs after the line of centers was 25% better at 42g.
Yep just opened a big can of worm… I am surprised about the 14.5 degree Bering more than the 25 degree The modified cycloid does make sense I do not thing the gear is deforming much Earlier test showed a module 2 cycloid breaking at around 20 kg In my test I have 1.6 kg at the teeth I need to make a new front frame so I can have a closer look at the meshing under load Also need testing as little backlash as possible Or no backlash and run the gears with some fine grinding compound until free And different number of teeth 6,8,10,16 Then there is various materials Just wondering how brass filled filament would work I ear they polish well I read somewhere can’t remember where that iglidur would not be good for clock application While the friction is low, the adhesion can be quite high
@@jacquesfavre999 Yes, a very big can of worms. There are so many things to try. 25 degree gears have contact over a larger area which should reduce the engaging friction. They probably prefer pinions with 16-20 teeth. They are more likely to push the arbors apart which could enlarge the pivot holes. Would helical gears spread the engaging friction across a larger area? Backlash is never a problem in clocks since pressure is always applied between the weight to the escapement. Grinding compound might work with brass and steel gears where it can be completely washed away. Would it stay embedded in the PLA and wear down quickly? I looked up Iglidur. Yikes, A350-PF is $229 for 750g and it needs a 180C build chamber. They also have other easier to print versions. I151-PF looks a lot better at $58 for 750g. Or I150-PF at $26 for 250g. What does high adhesion mean? Does it stick to the plate like PETG? Or does it have high static friction? Some PLA has low dynamic friction, but high static friction. The start/stop nature of the escapement makes it really bad when the gears have high static friction. Might be worth a try.
@@stevesclocks And there is this The point of contact slides from root to head for the gear and from head to root for the pinion So while there constant angular speed between the gear there is not constant torque With a module 2 gear and the same 12/72 ratio The radius of leverage would go from 35 to 37 for the gear and 7 to 5 for the pinion So the tans mission of torque will vary from 37/5 or 7.4 to 35/7 or 1/5 That is a significant variation Thinking of it there is less variation on the pin pinion I suppose this downfall is minimized with pinion with more theeth
@@jacquesfavre999 Intuitively, a constant velocity gear seems like it should have constant torque. The point of contact does move around, but so does the angle of applied force. Does this cancel out? Or is some of the rotational torque converted into a force trying to push the axles apart? So much to think about. More pinion teeth seems like a good idea. Moving the undercut to the large gear also seems like a good idea. Maybe it was never tried in a traditional clock with cut gears. 3D printing makes it easy.
@@stevesclocks The more I look into it, maybe the variation of torque is not such a big deal, the average is what counts, but while the high is easy to measure , the low not so much... I need to rethink the testing method building a full clock is one, I definitely was going to test the pin gear on a full built clock, maybe compare, pin pinion, cycloid, involute in a full scale clock When you have a 1 to 1000 ratio, things get obvious I am thinking: 1/6 x 1/6/ x 1/5 x 1/5 that is 1/900 , that would allow me to do a 8 day clock with a 1 minute escapement, weight with one pulley, about 1 m drop
I'm became quite curious on the results and waiting for more results. This is exiting. The weight was high and I don't believe the gear efficiency is equal on all weights . Maybe a variation of the weight gives interesting results?
Agree, this is more worst case scenario, and it not quite truly scientific evaluation Just wanted to compare Did a quick test on the involute with a 1/3 of weight, still got roughly about same ratio of weight/force, no dramatic change
@Ralph K If it works with a heavy weight, it will work better with a light weight. @Jacques On one of the close ups, I noticed a bit of elephant foot on the gear teeth. That will definitely make the meshing of the gears less efficient since the load will be concentrated on the tiny elevated part of the tooth. Another factor is the surface finish of the face of the teeth. That can be improved by building a jig and driving the gears with a mild abrasive like valve grinding compound (fine or medium, not the coarse).
@@solarguy6043 Good catch on the elephant foot I saw that too I have noticed many times that new clocks improve after a few days of running I have thought of doing some polishing I have read somewhere using teeth whitening toothpaste? All fun experiences to do
Very nice. Thanks!
Good😄😄😄
Very interesting, thx for sharing.
I might have a project after watching your video.
I would like to make a pendulum that will give me 1 second.
Basically, I want to elevate .500 steel.ball 18 inches high.
I'd like to make ball clock without an electric motor
Thx.
Very interesting results. There does not appear to be constant rotational velocity. Do you think the gears are deforming under load? Or is the active surface different than expected after printing? The elephant foot in the closeup would not help either.
I did an experiment using unloaded gears that would be found closer to the escapement. It used 4 stages of 50 tooth gears driving 10 tooth pinions for a total speedup of 625:1. A weight just large enough to move the gear chain was placed on the rim of the first gear. Cycloid gears and 25 degree pressure angle involute needed similar weights of 56g. A 14.5 degree involute was 20% higher at 68g. A modified cycloid with the teeth tilted so most of the touching occurs after the line of centers was 25% better at 42g.
Yep just opened a big can of worm…
I am surprised about the 14.5 degree
Bering more than the 25 degree
The modified cycloid does make sense
I do not thing the gear is deforming much
Earlier test showed a module 2 cycloid breaking at around 20 kg
In my test I have 1.6 kg at the teeth
I need to make a new front frame so I can have a closer look at the meshing under load
Also need testing as little backlash as possible
Or no backlash and run the gears with some fine grinding compound until free
And different number of teeth 6,8,10,16
Then there is various materials
Just wondering how brass filled filament would work
I ear they polish well
I read somewhere can’t remember where that iglidur would not be good for clock application
While the friction is low, the adhesion can be quite high
@@jacquesfavre999 Yes, a very big can of worms. There are so many things to try.
25 degree gears have contact over a larger area which should reduce the engaging friction. They probably prefer pinions with 16-20 teeth. They are more likely to push the arbors apart which could enlarge the pivot holes.
Would helical gears spread the engaging friction across a larger area?
Backlash is never a problem in clocks since pressure is always applied between the weight to the escapement.
Grinding compound might work with brass and steel gears where it can be completely washed away. Would it stay embedded in the PLA and wear down quickly?
I looked up Iglidur. Yikes, A350-PF is $229 for 750g and it needs a 180C build chamber. They also have other easier to print versions. I151-PF looks a lot better at $58 for 750g. Or I150-PF at $26 for 250g. What does high adhesion mean? Does it stick to the plate like PETG? Or does it have high static friction? Some PLA has low dynamic friction, but high static friction. The start/stop nature of the escapement makes it really bad when the gears have high static friction. Might be worth a try.
@@stevesclocks
And there is this
The point of contact slides from root to head for the gear and from head to root for the pinion
So while there constant angular speed between the gear there is not constant torque
With a module 2 gear and the same 12/72 ratio
The radius of leverage would go from 35 to 37 for the gear and 7 to 5 for the pinion
So the tans mission of torque will vary
from 37/5 or 7.4 to 35/7 or 1/5
That is a significant variation
Thinking of it there is less variation on the pin pinion
I suppose this downfall is minimized with pinion with more theeth
@@jacquesfavre999 Intuitively, a constant velocity gear seems like it should have constant torque. The point of contact does move around, but so does the angle of applied force. Does this cancel out? Or is some of the rotational torque converted into a force trying to push the axles apart? So much to think about.
More pinion teeth seems like a good idea. Moving the undercut to the large gear also seems like a good idea. Maybe it was never tried in a traditional clock with cut gears. 3D printing makes it easy.
@@stevesclocks The more I look into it, maybe the variation of torque is not such a big deal, the average is what counts, but while the high is easy to measure , the low not so much... I need to rethink the testing method
building a full clock is one, I definitely was going to test the pin gear on a full built clock, maybe compare, pin pinion, cycloid, involute in a full scale clock
When you have a 1 to 1000 ratio, things get obvious
I am thinking: 1/6 x 1/6/ x 1/5 x 1/5 that is 1/900 , that would allow me to do a 8 day clock with a 1 minute escapement, weight with one pulley, about 1 m drop
Lantern gear.
I'm became quite curious on the results and waiting for more results. This is exiting. The weight was high and I don't believe the gear efficiency is equal on all weights . Maybe a variation of the weight gives interesting results?
Agree, this is more worst case scenario, and it not quite truly scientific evaluation
Just wanted to compare
Did a quick test on the involute with a 1/3 of weight, still got roughly about same ratio of weight/force, no dramatic change
@Ralph K
If it works with a heavy weight, it will work better with a light weight.
@Jacques On one of the close ups, I noticed a bit of elephant foot on the gear teeth. That will definitely make the meshing of the gears less efficient since the load will be concentrated on the tiny elevated part of the tooth. Another factor is the surface finish of the face of the teeth. That can be improved by building a jig and driving the gears with a mild abrasive like valve grinding compound (fine or medium, not the coarse).
@@solarguy6043 Good catch on the elephant foot I saw that too
I have noticed many times that new clocks improve after a few days of running
I have thought of doing some polishing
I have read somewhere using teeth whitening toothpaste?
All fun experiences to do
👏👏👏👏👏👏👍👍👍👍👍👍
👍educational!! PETG is better, in my opinion!!
😇 P"R"O"M"O"S"M!!