aeh5040

I think both of you are wrong here. (Sorry to point it out! ) I do not think the Lego "CV" joint is constant-velocity. Imagine what would happen if it were at an angle of very nearly 90 degrees (if the casing were long enough to allow that). The casing on the left would hardly rotate at all for most of the rotation of the insert on the right. You are probably thinking of something like a Rzeppa constant-velocity joint. But that is very different. It needs free-floating ball bearings that slide in grooves in BOTH axles, not just one of them. On the other hand, it is definitely not equivalent to a (Cardan) U-joint. You can see this immediately from the fact that there are two moving parts, not three. And there are many other differences (sliding versus pivoting, etc...)

In most cases I don't think U-joints are responsible for that. It is more likely backlash plus elasticity of various kinds.

I don't have any actual data, but I suspect that even 1 XL motor, ungeared, can break a 12 tooth bevel gear.

Yes, in fact, I have been "milking" this same basic construction for all it is worth for many years... (I think I need some fresh ideas...)

Indeed - I was partly inspired by these!

I was going to try to make a picture, but then I found it has already been done, on this forum (not a surprise, really). Check out: http://www.eurobrick...showtopic=49999 The first one here is what you usually want (as in my red ring):

And some more kinetic sculpture...

I appreciate the attempt, but this will not work. You will just end up with a bunch of Geneva mechanisms going at different speeds. The slower ones will spend long periods in intermediate states between two bits. If you look at the video it is quite clear that each bit actuates the next one. Indeed, that is a necessity to get a reliable counter of any kind.

A U-joint (or universal joint) is this: It enables rotation to be transferred through an angle. However, it has a disadvantage that the rotation is not quite linear - if the input rotates at a steady speed, then the output will be slightly faster at some points of the cycle and slightly slower at others. With one joint this is barely noticeable, but with several the effects can add up. The two rings each contain eight very large U-joints (they are actually inside out - the colored square is the equivalent of the small piece in the middle of the Lego part above - but this makes no difference to the operation). The only difference between red and blue is how they are connected up. In the blue ring, each grey section contains a twist, so the non-linearity effects add up, to make a big effect at the top. In the red one, the directions alternate, so the effects cancel out. If you you use several U-joint in series, you want to line them up as in the red one!

A demonstration of what happens if you don't align your U-joints correctly...

What a jerk. If you inform Lego Ideas, I expect that they would remove the project and ban the user.

If you make some pictures, I think that there will be people willing to turn them into instructions.... (which would be awesome!)

This is awesome news! Really looking forward to complete instructions coming together!

No worries, and thanks for the interest! Again, this mechanism seems to be different from (and better than) Geneva mechanism one, in that each stage gets its own power, so there is (apparently) no accumulation of friction as you go along the stages. The point is that one could in principle drive something that required lots of torque and power from each bit (e.g. a giant "0" and "1", maybe!), and still have enough power. I'd like to understand exactly how it works!

This is exactly what I wanted to see - many thanks for doing it. It is a brilliant mechanism!

This must win the prize for the most specialized cause of craziness! i can just imagine you trying to explain what it means to a psychologist...

Actually I'm not sure what your conclusion is. You seem to be saying "it is useless except for the thing that it is useful for", which is true of most things...

Wow! This is one of the most awesome models I have ever seen! Both the mechanics and the styling are amazing. I too would very much like to see more "stripped down" pictures of the mechanisms. The "varying amplitude" for turns is extremely interesting and original. I am curious whether this mechanism on its own would be sufficient to make a turn, even without bending the body. Did you try this? Or is it possible to modify it to find out?

Let me again add my voice to the growing chorus (clamour?) begging for instructions for the horse! It is one of the most awesome models ever, and it is rather special in that the appeal goes beyond the usual technic audience. I started trying to reverse engineer it myself, but got stuck. I would also be very willing to make an LDraw file (indeed, it would be a pleasure) from some slightly more detailed pictures. I suspect all it would take would be pictures of the horse broken into 4 or 5 units, each viewed from several angles...

This thread is pretty funny! Our community would provide substantial fodder for a psychology thesis, it seems. Actually, I have changed my "policy" on this. With instructions I typically follow them. In my own building, I used to put the smooth side against the rotating surface, but now I usually put the castellated side against the rotating surface. Almost certainly it makes no difference at all in practice, but my "rationale" is: if the bush is being pushed against the surface then it should not make any difference to friction - a smaller area means higher pressure, and higher friction per unit area, but the total friction should be the same. (Come to think of it, perhaps the torque induced by friction is a bit smaller, because of the average radius is smaller.) When there is no pushing force and the bush just happens to touch, I would expect a smaller area to mean smaller friction. Obviously, an exception to all this is when the "surface" is e.g. a studded technic plate! Orientation of 3L pins can really matter on rare occasions. If you have 3 stacked liftarms connected by a 3L pin, it protrudes significantly more on the "long" side. So much so that certain parts will jam if they are rotating against the surface of the liftarm. I have not felt the need to do this yet - please don't encourage me!

The two axles are still perpendicular. The point is that they are not in the same plane.

Nice! i like the minimalist bracing too!

The hailfire seems to have 92 = 23x4 outer teeth, so some very unusual gear ratios are possible...!

I tried the yellow one, but I think all kinds will work. You probably know that you can fix the gears to an axle using 2x2 round plates. It is advisable to reinforce them in this or some other way, as the plastic is very thin in the centre.

I'm not sure whether mania is the right word, but when I build from instructions I insist on getting the bushes in the same orientation as in the instructions. 99% of the time it of course makes no difference to function, and very little to appearance.