Adjustable Spirograph

[PG52]

Hum, you are using a 2L pinaxle without friction to hold the 40t, right ? On the original model, I use long axles which gives a lot of stability to the arm. The problem could be from here.

But it is also very hard to obtain good drawings with the technique you are using ; maybe you should synchronize the rotations of the turntable and the 40t gear ; because the ratio 8t/TT is 1/7, you can't obtain it with classic gears so eversything is completely unsychronized... I have to say that I never tried to do that. Another french technic MOCer, who gave me the inspiration for spirographs, has done it : http://forum.brictec...hp?f=114&t=5425 (in french)

Edited February 26, 2014 by PG52

[1963maniac]

Yup, just a 2L axlepin holds the 40t gear on top of the turntable. A long axle coming through the turntable drives it. I need to think of a better way to attatch the 40t gear to the turntable. That is the problem.

[1963maniac]

Guys & Dolls,

I rebuilt my mod of pg52's spirograph version 4. It works much better now. It goes a long time before it starts to retrace the graph. It can also be irregular as you can see here. I also took away the two extra M motors and made it all run off the one large motor. I put some weight on the arms to help the pen draw better. I'm recharging my batteries so I can try some more drawings.

[isopyl]

Hi,

I recently built this using the excellent instructions created here (well done everyone) but I've noticed that its very hard to reproduce a pattern. I initially thought that noting down the position of all the gearboxes would be sufficient but as you can see in the image below, you can get very different patterns using the same gear settings.

Has anyone figured out the best way to align the gears or align the arms so that the patterns are reproducible using the same gear ratios?

Cheers,

Stephen

[Blakbird]

Hi,

I recently built this using the excellent instructions created here (well done everyone) but I've noticed that its very hard to reproduce a pattern. I initially thought that noting down the position of all the gearboxes would be sufficient but as you can see in the image below, you can get very different patterns using the same gear settings.

Has anyone figured out the best way to align the gears or align the arms so that the patterns are reproducible using the same gear ratios?

I see the increased variability as a benefit rather than a detriment. However, if you want to reproduce results you'll need to record more than just the gearbox settings. You also have to record the relative initial phase of the output arms. If you start the arms in the same relative position, you'll get the same result repeatedly.

[PG52]

Exactly. To have a good precision when you place the arms, you could maybe use the mechanism with a screw on a differential, and input and output the entries of the diff'.

[bartneck]

I build an extension of this design using EV3. Now the Spirograph Automaton knows when the pattern is complete and lifts the pen.

[legolijntje]

That's also a nice version.

How does it know when to stop? Or doesn't it and just stops at a certain time?

[Boxerlego]

That is amazing bartnect. Though I'm a big fan of the mechanical versions, your coin operated EV3 version is incredible. I clearly see that the core shape of the drawing is a heptagon.

I'm guessing that the EV3 is programmed to stop based on the platform rotation, because this machine is capable of generating wide variety of unique patterns and if all those pattern can be drawn completely with in one revolution of the platform. But I not sure on this because of the special part I see with the drawing arms that have a website on it, that is probably the true way on how it stops.

Edited April 14, 2014 by Boxerlego

[bartneck]

The EV3 calculates the right time to stop simply through math. When you know the gear ratios of both arms and the table, then you can calculate how many rotation the motor needs to make However, the user can select four different gears for each of the arm and the table and hence it is necessary to use sensors to detect in what gear each of the three components is. This is achieved by rotation sensors (the special parts, Boxerlego) that are attached to all three main components. The table (or platform if you like) turns as often as necessary. It is not set to just one rotation.

[Blakbird]

The EV3 calculates the right time to stop simply through math. When you know the gear ratios of both arms and the table, then you can calculate how many rotation the motor needs to make However, the user can select four different gears for each of the arm and the table and hence it is necessary to use sensors to detect in what gear each of the three components is. This is achieved by rotation sensors (the special parts, Boxerlego) that are attached to all three main components. The table (or platform if you like) turns as often as necessary. It is not set to just one rotation.

So you measure the rotation rates to determine the gears and then just calculate the least common multiple?

[Rishab N]

I like the EV3 transformation!

[bartneck]

So you measure the rotation rates to determine the gears and then just calculate the least common multiple?

Yes, that is pretty much it. It is a bit tricky to determine the gears since they can be pretty close to each other, so I also wrote a program to calibrate the system.

[dvsntt]

The most sophisticated Spirograph ever! That is awesome.

[bartneck]

It is Easter Time and would it not be great to use a spirograph to draw on eggs? I thought it would be sufficient to rotate the egg instead of the table, but the results were dissapointing. The patterns are not particulary pretty and I am not even sure if they repeat accoding to my calculations. Any help is welcome.

[Captainowie]

It is Easter Time and would it not be great to use a spirograph to draw on eggs? I thought it would be sufficient to rotate the egg instead of the table, but the results were dissapointing. The patterns are not particulary pretty and I am not even sure if they repeat accoding to my calculations. Any help is welcome.

Hi.

Why would the pattern not repeat? If it repeats with a turntable after a certain amount of time, surely if there are no changes in gearing it should repeat with an egg after the same amount of time?

It seems to me that the problem you're having is that in changing the axis of rotation from vertical to horizontal, you've made left-to-right movement of the pen across the egg the same as rotation of the egg itself. I don't know how easy it is to make modifications, but if you could decouple the left-to-right motion of the pen from the front-to-back motion, and had the left-to-right motion much smaller than the front-to-back motion, you might end up with a prettier pattern.

Does that make sense?

Owen.

[1963maniac]

Doe anyone still have a copy of the LDD file. I made a MOD and would like to use in to record how I made my MOD. ML CAD software is a little bit tough for me. I'm finally getting a little handy with LDD.

[PG52]

Hi !

The LDD file (the previous link is broken) : https://www.dropbox.com/s/qskfqojh2usf05n/Spirographe%204-2.lxf?dl=0

[DrJB]

It is Easter Time and would it not be great to use a spirograph to draw on eggs? I thought it would be sufficient to rotate the egg instead of the table, but the results were dissapointing. The patterns are not particulary pretty and I am not even sure if they repeat accoding to my calculations. Any help is welcome.

The math might be easier to formulate (and the motions to replicate) if the pen moves only along the axis of the egg, and the egg is made to undergo 'complex' rotation about its axis (not continuous rotation). The challenge there is that the gears driving the rotation of the egg need to be 'preloaded' so that any reversal or slowing down of the egg's rotation can be 'controlled' so as not to get jagged lines. Of course, for simple patterns (such as sinusoids), the egg will be rotating one way only and no need to control the backlash. I think it might be easier to build this 'simple' one first (continuous one-way rotation of the egg), then we can move on to more complex egg motion. We can elaborate a bit more on this if you're still interested.

On a different note, PG52 mentioned he was working on a fifth version of the spirographe, any update on this?

[PG52]

The fifth is done for a long time, see the first page ;) I've already built the versions 6, 7 and 8 but I have really no time nowadays to publish videos or reviews (like any student in a mechanical engineering school, I don't have so much time for me and I prefer spending some time to build than to publish... However I still take some pics and vids of my MOCs to publish them later :) ). Something will probably come in almost a month and a half :)

[legolijntje]

What are the differences (either positive or negative) of versions 6, 7 and 8 compared to 4, for which we created the instructions?

[TeClem8]

Super Pierre ! Your Spirograph made a real success ! Quite simply because it's great !

Here is a video I realized with PG52 at a meeting in my house.

My black spirograph, and two of PG52's.

I think that the right one is the V6. It was extremely fast !

My first spirograph, original but not stable.

and the second, présent in the first movie.

Edited January 6, 2015 by TeClem8

[PG52]

Salut Clément ! Happy to see you there and thanks for your answer ^^

In fact, the right one was only the V3. It wasn't able to make the drawings that the V5 and later can create, but yes it was quite fast.

The main differencies between V5 to 8 are the pure performances of the versions. V6 had the same than V5 with some additional features (which failed because the vibrations were too important).

It had plenty of possibilities, it was fast but was vibrating constantly and needed too much power to work properly (anyway the plastic axles and beams did'nt like this episode ^^).

The V7 solved the vibrations problem, had a lot of drawing possibilities but still required a lot of power (1*XL) to work well. (My goal is a spirograph working perfectly with only 1*M motor)

The V8 finally solved the power problem, but sacrifying the number of possibilities (you can only choose the number of "petals" of the drawing, 1, 2, 3, 4, 5, 6, 7, 8, 9 or 12 ; that means only 10 different drawings. It also has an interesting new technology of drawing board, with a big intern gear made with 16 gear racks that increases precision and reduces the torque in the mechanism.

I'm always looking for a version having the qualities of the V6, 7, 8 and not their problems :D

Edited January 9, 2015 by PG52

[Rishab N]

38.5 percent of the parts hmm I might build this. What gear combinations produce the best pictures?

[jeffy1021]

Thanks PG52 and those who made the instructions. I just finished building the V4 posted on rebrickable, lots of fun!