Hi all. This is my custom gearbox with 7 forward speeds, neutral and reverse. This is made with some custom 3D printed components. I basically wanted to make a more realistic gearbox and wondered if it was possible, given the right parts, to make within the Lego "system". Here's a video of the results:

 

Yes it's very noisy/rattley and the parts are very loose with not very good tolerances, but thems the breaks when you have an FDM type 3D printer! This is only a prototype. I'm certain it would run much nicer with proper Lego injection moulded parts, with the much smoother surfaces and tighter tolerances they can achieve. I think it would run much quieter with some grease but wanted to try without any.

I'm really pleased with this result. I like that it's much more realistic and has very low friction. There's no way you could ever push it fast enough to make the gears click while driving a fake engine. I'm also really happy with the sound the engine/motors make when shifting through the gears, it's really satisfying to play with! 

For gear ratios, I decided on 24t, 22t, 20t, 18t, 16t, 14t and 12t as that would allow the main shaft and the lay shaft to fit on a standard 2 module spacing on centre. 2 gear teeth doesn't sound like much of a difference between ratios but remember that their mated gears are also changing size by 2 teeth, so the difference is noticeable, with a fairly large range between 1st gear and 7th gear. 

In terms of torque, it can stall an XL motor without anything skipping in reverse, 3rd, 4th and 7th gears, as those gears are right up against supporting beams. However in 1st, 2nd, 5th and 6th, as those are in the middle of the axles, the axles can bend and gears can slip under the torque of an XL motor. However this can be fixed by adding two more support beams between those gears. This current setup however is fine for medium torque, high speed applications where you are using faster motors, different gearing or simply manually driving a fake engine. I didn't want to test it to destruction however as these parts take hours to print!

The indexing mechanism is also more accurately placed, built right into the gearbox, so no need for that mechanism to take up any space by the steering wheel, so hopefully that should allow for a much cleaner steering wheel construction with a nice, angled up towards the driver steering wheel and flappy paddles. The force required to shift is much less that the UCS Porsche, but being placed far from the steering wheel you'd probably want to use a push/pull linkage rather than a twisting shaft.

Hope you enjoy and any comments/constructive criticisms are of course welcome. Or if you just want to say it sucks you can do that too! 

Edited June 12, 20222 yr by allanp

Wow, this is how it should be. How long does it took to invent the correct shifting channels?

Edited June 12, 20222 yr by efferman

One can only wish for proper Lego parts to be able to build this. Truly a realistic gearbox
And I love your motors, you just can't let them go

Thanks @efferman. The cam piece probably took a couple hours to design. I wanted 8 gear positions but also a small neutral area between each gear to allow a bit of time for the previous gear to disengage before the next one engages, and also to allow for a neutral position without taking up one of the gear positions. So in Fusion 360 I started my making a straight track, using circles that match the shift fork pin diameter in each position. Then wrapped it around the cam.

53 minutes ago, allanp said:

Or if you just want to say it sucks you can do that too! 

I could say that, but I'd be lying, wouldn't I? 

Even with the caveat of using an FDM printer, it doesn't sound out of place at all. You've clearly done a good job with the design of the parts.

 

2 hours ago, Milan said:

One can only wish for proper Lego parts to be able to build this. Truly a realistic gearbox
And I love your motors, you just can't let them go

Thanks Milan. Yup, there's just something more exciting (and I guess in this case authentic) about fast spinning gears and fake engines and the sounds they make! And yeah, if these parts (or something very similar) was available from Lego it would be so much nicer. As it is now everything is loose and wobbly and rattles! I'd rather hear the brraaAAAh click brraaaAAAH of the engine/motors than the rattley gearbox. Works though!

2 hours ago, Jay Psi said:

I could say that, but I'd be lying, wouldn't I? 

Even with the caveat of using an FDM printer, it doesn't sound out of place at all. You've clearly done a good job with the design of the parts.

 

Thank you 

Well done, works very well and sounds good! Hopefully this inspires some people at TLG and otherwise this idea will find its way…

Do you have improvements planned or is this your final version for now?

47 minutes ago, GerritvdG said:

Well done, works very well and sounds good! Hopefully this inspires some people at TLG and otherwise this idea will find its way…

Do you have improvements planned or is this your final version for now?

Thank you 

In terms of improvements, I think I could improve the tolerances by quite a bit. I'm new to 3D printing so not sure how close I can be. Things that should be tight on the axles (like the axle extender pieces that the drive rings sit on) are just completely loose! I could use regular 2 module long axle extenders and leave a cutout on the clutch gears for room, but that would make them need to be asymmetrical. The clutch gears are only asymmetrical right now because they was easier to print that way. But this design does allow them to be symmetric. I could also improve the pegs on the shift forks (that engage with the rotary cams) by not printing them, and instead leaving a hole and gluing in a much smoother, round piece of plastic. I could also step back the drive dogs just slightly more on the drive rings to give some more clearance but that shouldn't be needed if the gears wasn't wobbling about so much. They don't jam but they do cause some rattles. Maybe I could soften their edges a bit. Lots of little improvements like that which I might do over time. But for now it's very reliable and functional.

Edited June 12, 20222 yr by allanp

It's pretty cool seeing that! Would you be able to share STL files, or better yet, F3D files?

Wow, that's pretty great that you actually printed and built this so quickly and it works so well for the first try. I think you chose the right gear sizes, because this is the only sequence that can still be kept in a 2 module spacing, without getting too bulky because of too big gears (and then every part would have to become bigger and would probably be more wobbly and less precise).

It would be awesome to see only a 4-speed version of this (maybe with 24T, 20T, 16T and 12T gears), but strongly braced, I'd be interested if an XL motor could make its gears skip (which probably also depends on the amount of down-gearig after the gearbox towards the wheels). Do you plan to make some field tests with a more complete chassis?

I also wonder what other layouts would be possible for the whole assembly? Currently it's really tall (even not counting the frame), but would it be possible to move the shifting components lower but to the side for example?

Well done, this is very cool! I think even if LEGO made more gear sizes (clutch and normal) it would go a long way to improving the current system without redesigning the wave selector mechanism. This however takes it to a whole new level. One can only hope someone from LEGO is watching and being inspired in some way. 

4 UCS cars so far... they are 0 from 4 in terms of actually having the correct amount of gears compared to the real car... and 2 from 4 in terms of actually getting the gears in the correct order. Not a very good record 

Very nice. There's probably people that might object to TLG producing even more parts with such specialized applications, but it's safe to say that current official transmissions are unwieldy and inefficient for play. I'm surprised that no clone company has taken this approach yet. Maybe someone can convince CaDA

Edited June 13, 20222 yr by Bartybum

Do you have STL files for the parts? I’d like to print them out and try them for myself.

 

By the way, don’t forget to print a 20x20 test cube and check the dimensions so you can adjust the printer properly. Settings such as “horizontal expansion” helped me adjust my printer so it makes parts that have close tolerances. I’ve done some LEGO monorail track with it recently and it press fits ok.

5 hours ago, Bartybum said:

Very nice. There's probably people that might object to TLG producing even more parts with such specialized applications, but it's safe to say that current official transmissions are unwieldy and inefficient for play. I'm surprised that no clone company has taken this approach yet. Maybe someone can convince CaDA

It's indeed "specialized" as in the parts in question don't have much use beyond gearbox parts... but TLG has made specialized fake engine parts too, so why not gearbox? Sets with gearboxes are released almost every year so it would see a significant amount of use, despite being single purpose.

But yeah, this demonstration proves once and for all that more compact gearbox design is indeed possible within the system and TLG has run out of excuses to not make one.

That reminds me very much of a sequential bike gearbox (cars are probably the same, but i've only worked on bikes). Brilliant build!

18 hours ago, allanp said:

Hope you enjoy and any comments/constructive criticisms are of course welcome.

It's great!

How much force do you need for a gear switch, do you think it would be easy to motorize?

11 hours ago, Bartybum said:

I'm surprised that no clone company has taken this approach yet.

I'm actually surprised that competitor companies (even one that are not simply cloning) use 99.9% the same parts as Lego, and rarely ever invent new ones, not even existing designs in new lengths for example. The only thing they do is some recolors compared to lego. AFAIK only Cada dared to invent a few new parts, but not many.

26 minutes ago, gyenesvi said:

I'm actually surprised that competitor companies (even one that are not simply cloning) use 99.9% the same parts as Lego, and rarely ever invent new ones, not even existing designs in new lengths for example.

One of the reasons for having cheaper products is to spare the R&D costs, they are not interested in innovation, they want profit :)  

But to stay on topic, I'm impressed @allanp how quickly you went from theory to practical testing, and I don't think the sound is much worse then a LEGO gearbox would be :D Will follow the refinements/development for sure! 

 

30 minutes ago, gyenesvi said:

I'm actually surprised that competitor companies (even one that are not simply cloning) use 99.9% the same parts as Lego, and rarely ever invent new ones, not even existing designs in new lengths for example. The only thing they do is some recolors compared to lego. AFAIK only Cada dared to invent a few new parts, but not many.

R&D is expensive, much more so than copying someone else's design. Maybe this will change in the future but at the moment it looks like TLG is the only company with resources to do any innovation.

Brilliant. I'd love to see the different parts in closer detail, just to understand better how this was achieved. Is there anywhere I can go for a look? Thanks

2 hours ago, howitzer said:

R&D is expensive, much more so than copying someone else's design. Maybe this will change in the future but at the moment it looks like TLG is the only company with resources to do any innovation.

R&D is expensive, yes. But on the other hand, some enthusiasts can develop new Bricks, programm better control apps and other things, without the massive resources of a Big company...

1 hour ago, Gray Gear said:

R&D is expensive, yes. But on the other hand, some enthusiasts can develop new Bricks, programm better control apps and other things, without the massive resources of a Big company...

But those enthusiasts don't really have to answer to anyone but themselves, so they can cut entirely stuff like quality control and all that. They also don't need to make profit (or any revenue at all) and if the project fails, the loss is mostly their own time spent. Then there's of course the scale of the production, it's fairly easy to craft a single, working prototype (like allanp showed here) but there's still a big leap from that to producing millions of these parts. It's basically the same question which arises every time someone wonders why TLG won't produce this or that part.

I'm sure the competitors can also do the math, and that's why we don't see much innovation from them, at least as long as they aren't growing big enough to pose a serious challenge to TLG's almost-monopoly.

28 minutes ago, howitzer said:

*snip*

I'm sure the competitors can also do the math, and that's why we don't see much innovation from them, at least as long as they aren't growing big enough to pose a serious challenge to TLG's almost-monopoly.

Exactly, bringing a one- off to mass production is not easy, but if you look at Effermanns parts for example I doubt it would be that difficult, seing as small scale production is already happening. My point was, developing new bricks isn't hard to a point where competitor brands to TLG can't do it, even private enthusiasts can do it. They just don't want to make any new moulds as that's where the real money is spent.

And for the Control+ failings there is just no excuse, which you noted as well I see

Edited June 13, 20222 yr by Gray Gear

Thanks for all the comments guys . I'm happy to share 3D files and more pics of the individual parts. I will need to clean things up a bit as I have been messing about with things, also parts are split (and later glued together) to make them easier for this 3D printing noob to print them. But let me know what you would like either here or in a pm. I'll probably put the parts together as one piece so you can decide how best to split/print them. Btw let me know if you are able to split the parts or if you want me to leave them already split up, though I wouldn't necessarily trust my judgement on how best to split/print the parts if you are more experienced at 3D printing. 

I would also like to try out other configurations. It was built this way to make it easy to show from the side but I'm pretty sure things could be rearranged to make it fit different spaces. 

@Gimmick not much force to shift. Should be easy to motorise. When I get more time (later this week hopefully) I shall show more of the parts and the indexing mechanism.

6 hours ago, kbalage said:

One of the reasons for having cheaper products is to spare the R&D costs, they are not interested in innovation, they want profit :)  

But to stay on topic, I'm impressed @allanp how quickly you went from theory to practical testing, and I don't think the sound is much worse then a LEGO gearbox would be :D Will follow the refinements/development for sure! 

 

Thanks Kbalage 

3 hours ago, allanp said:

 But let me know what you would like either here or in a pm. I'll probably put the parts together as one piece so you can decide how best to split/print them. Btw let me know if you are able to split the parts or if you want me to leave them already split up, though I wouldn't necessarily trust my judgement on how best to split/print the parts if you are more experienced at 3D printing. 

Thanks! My personal preference would just be for Fusion360 files in the original format, so I could manually edit whatever I need to, but for a lot of people a standard .stl file might be handier.

I don't want to derail this topic, but I don't think this is worth its own topic either, so I'll take my chances and share my version of a 3D-printed gearbox here, I think.

My design is a fair bit more like the original Lego one, but with more flexibility and compactness than the original one.

 The gears are only half a stud thick, and mesh in the standard three stud spacing. The driving gears range from 7 to 17 teeth, with gears at every integer between them, and the clutch gears range from 15 to 25 teeth. The driving rings are essentially the classic 2L driving rings, but with the little cut-outs in the end like the new ones, to allow them to mesh with the newer style of gears. There are a lot of variants of wave selector at different angles for different gear counts, as well as a larger one to allow for more gears to be used. The larger, four-stud diameter of these selectors allows for a lot of gears, and I set it up to allow for up to twelve gears, allowing me to make a 10+R+P gearbox, like the largest gearboxes in cars today. (Neutral kind of got lost, but with some redesigning it could be added.) The park was a custom part that was designed to allow the driving ring to engage with it, but then lock it against the other axle.

I think the whole setup ought to work in theory, but when I printed it, there was way too much friction for smooth operation. It was also a total nightmare to try to get all the gears in the right order for my monster 10+R+P demo gearbox!

Some more refinement is likely necessary to make this usable, but at any rate it should be an interesting counterpart to allanp's very cool design!

https://bricksafe.com/pages/2GodBDGlory/3d-printed-gearbox

Edited June 13, 20222 yr by 2GodBDGlory