[TC13] Pentagon Clock - Wind-up Pendulum Clock

[mocbuild101]

9 hours ago, Didumos69 said:

Please let me know what you think

I think it's great, and much better than a big blue thing...

[agrof]

I also vote for the clock face and no housing. I like, that the clock mechanism is uncovered, would be pity to hide it.

What I am still not liking, is the pendulum arm. It should be straigth, sleek and functional, as it is, I feel it overdone. It should let the eyes to slide on the main thing: the mechanism. 

Edited January 10, 2018 by agrof

[Didumos69]

Thanks @Appie, @Victor Imaginator, @TechnicRCRacer, @JonathanM, @mocbuild101 and @agrof! This is why I like having a WIP topic, it pushes me further.

16 hours ago, JonathanM said:

I wonder if you could have the slots of the 2L connector on the inner pentagon facing out to line up with the outside pentagon markers (presumably these mark every 3 seconds and half a minute).

I think I understand. I moved the pins with pinhole to the outside.

5 hours ago, agrof said:

I also vote for the clock face and no housing. I like, that the clock mechanism is uncovered, would be pity to hide it.

What I am still not liking, is the pendulum arm. It should be straigth, sleek and functional, as it is, I feel it overdone. It should let the eyes to slide on the main thing: the mechanism. 

Is this better:

Edited January 10, 2018 by Didumos69

[Didumos69]

I built a prototype of the clock face to check how well it keeps its shape. I had to reinforce it with axles for the half minute indicators, but that seems to look quite good actually. I will order the white parts now.

Edited January 10, 2018 by Didumos69

[agrof]

I love the black one, but the main question is: do You? 

[Didumos69]

2 hours ago, agrof said:

I love the black one, but the main question is: do You? 

I like the black one the most too. Makes a nice contrast with the pendulum bob and the clock face.

[mocbuild101]

7 hours ago, Didumos69 said:

I like the black one the most too. Makes a nice contrast with the pendulum bob and the clock face.

+1

[JonathanM]

On 11/01/2018 at 12:08 AM, Didumos69 said:

I think I understand. I moved the pins with pinhole to the outside.

Actually I meant the five, white 2L pin connector with slot that were in the inner ring. Rotate them about the 4L axles so that the slots (on the top of the bricklink image) face forward so you can see the slots, thus denoting the 10ths.

But your new solution is way better anyway! As is the pendulum in black - the white/black combination really pops. Great work!

Edited January 11, 2018 by JonathanM

[Didumos69]

Hanging on a wall with a longer and lighter pendulum. Autonomy now is 18 minutes.

 

[BrickbyBrickTechnic]

Awesome so far! How many rotations does each hand make in 18 minutes?

[lcvisser]

It looks really cool on the wall  Looks a bit like the one I have hanging here (also just has a face with all the gears exposed).

Is it with one or two pullbacks? Can't remember if you proceeded with integrating both.

 

[Aventador2004]

58 minutes ago, Ludo Visser said:

Is it with one or two pullbacks? Can't remember if you proceeded with integrating both.

2, he added a second recently.

[Didumos69]

3 hours ago, BrickbyBrickTechnic said:

Awesome so far! How many rotations does each hand make in 18 minutes?

Thanks! I don't.know exactly, but going by the 1m:32s I filmed it makes about 50 LEGO minutes, 10 LEGO hours and 5 clock cycles (2.5 LEGO days) in 18 minutes.

2 hours ago, Ludo Visser said:

It looks really cool on the wall  Looks a bit like the one I have hanging here (also just has a face with all the gears exposed).

Is it with one or two pullbacks? Can't remember if you proceeded with integrating both.

 

Thanks Ludo! My father also has a (wooden) clock with only a face and no house, which is where the inspiration comes from.

1 hour ago, Aventador2004 said:

2, he added a second recently.

Correct!

Edited January 14, 2018 by Didumos69

[grego18f]

Wow, it is very impressive! I never thought a pullback motor could power a clock and even less for 18 minutes

 

[Davidz90]

Awesome! 18 minutes is a very respectable authonomy. 

[Didumos69]

On 1/14/2018 at 7:26 PM, grego18f said:

Wow, it is very impressive! I never thought a pullback motor could power a clock and even less for 18 minutes

21 hours ago, Davidz90 said:

Awesome! 18 minutes is a very respectable authonomy. 

Thanks!

I tested the thin pendulum arm I showed earlier, but their was too much slack in the connections, which gives some undesired resonance in the arm when the pulback generates the most force. So I will stick to beams for the pendulum arm instead. This will probably be the final design. An all black pendulum arm looks a little too massive, so I opted for a white line in the center of the arm. I'm now waiting for the necessary white parts to arrive...

Edited January 16, 2018 by Didumos69

[Davidz90]

2 hours ago, Didumos69 said:

I tested the thin pendulum arm I showed earlier, but their was too much slack in the connections, which gives some undesired resonance in the arm when the pulback generates the most force

This one looks great! Another stiff solution is to use single line of beams, but connect them with plates:

connection by David_Z1, on Flickr

 

 

[Didumos69]

11 hours ago, Davidz90 said:

This one looks great! Another stiff solution is to use single line of beams, but connect them with plates:

Thanks for the suggestion! I don't really like using bricks, but maybe in this case... They do give a nice stiff connection.

[Didumos69]

It's been quiet here for a while. I received the ordered parts and finished building last friday. It was my sons birthday yesterday. I put the clock on a wall in the livingroom so family and friends could admire my work. Most were very enthousiastic about it, but a few pointed out that there still is room for improvement: The axle that is subject to the escapement mechanism makes a 360 degree rotation every 6 ticks (3 pendulum periods). If I would be able to somehow increase the number of ticks for a complete rotation, I would get a longer autonomy with the same gear ratios.

So, as the number of competitors - and clocks - in this competition grows, I thought it would be worthwhile to make a further investigation into the escapement mechanism. I went back to the drawing table and here are some ideas that might work. In my current implementation, I have 3 pins sticking out of 2 belt wheels, which push the pendulum to the left and right. If one of the saw blade options below would work, it would give me about (8 / 3) x 18 = 43 minutes autonomy. I ordered the saw blade piece, so I will only know later this week.

Edited January 22, 2018 by Didumos69

[Davidz90]

The saw blade is a great idea, the designs definitely look like they would work fine! Obviously, regular gears give the most ticks per revolution, but their fine teeth are hard to work with.

[Didumos69]

17 hours ago, Davidz90 said:

The saw blade is a great idea, the designs definitely look like they would work fine! Obviously, regular gears give the most ticks per revolution, but their fine teeth are hard to work with.

You're right. However, the finer the teeth, the smaller the amplitude it gives to the pendulum. When the amplitude is too small, the pendulum starts oscillating. I hope the saw blade will provide a big enough amplitude...

I tried the version with the knob wheel. It works, but it doesn't give equal tick intervals.

[schraubedrin]

Oh come on.
How on earth do you come up with those mechanisms so quickly?
I was working on an escapement using the saw blade for the last three days with no results. It's the only thing stopping me from entering a competition for the first time

 

I guess i'll just enter my unfinished work

[Davidz90]

 

18 minutes ago, schraubedrin said:

I was working on an escapement using the saw blade for the last three days with no results.

Maybe try grasshopper escapement, like this:

or Galileo escapement

 

I can attest that both of them are easier to get running reliably than typical anchor escapement. Been there, done that 

[Didumos69]

11 hours ago, schraubedrin said:

How on earth do you come up with those mechanisms so quickly?

Well, these are just ideas, I'm still awaiting the saw blade piece, so I can't test it.

10 hours ago, Davidz90 said:

I can attest that both of them are easier to get running reliably than typical anchor escapement.

I believe you instantly.

Meanwhile I have been rearranging (digitally) the back-side of the clock to make it 3 studs shorter. That should bring the whole clock a little closer to the wall.

Before:

After:

The gearing in the back now looks like this. It provides a 20:1 ratio. The rear output of the diff is locked by default, which gives the diff a 2:1 output / input ratio. The rear output slips during wind-up, so the hands of the clock don't move backwards. Bracing everything well was extremely hard. I will build this this afternoon.

Edited January 24, 2018 by Didumos69

[schraubedrin]

Thanks for your advice, I'm sorry that i was so frustrated

11 hours ago, Davidz90 said:

or Galileo escapement

That is actually what i'm in the process of developing, this video actually helped me, thank you! As usual i'm my own worst enemy: i want to build in system with geometrically locked parts.
Currently i have some kind of double-galileo-escapement with an unpowered pendulum. So I'm at least able to show how my machine is meant to be working

 

35 minutes ago, Didumos69 said:

Bracing everything well was extremely hard

And really important. I have the feeling that rigid bracing directly influences the efficiency of a drivetrain, especially at high ratios with low torque.

 

This construction in the Porsche was a good lesson by bad example