Ultimate 6+N+R Sequential Gearbox

[Anto]

INTRODUCTION

This gearbox is the successor of my previous 8+N+R-gearbox. It is based on the new parts released on the Yamaya MT-10 SP #42159. Thanks to these, I could make a new and better gearbox for 1:8 manual supercars.

Thanks to the 8 positions by rotation, I made a 6+N+R gearbox. The best arrangement I found was to put 2 cylinders between 2 shafts, each cylinder controlling 2 forks. Thus, I would need 2 cylinders to make the 8 positions I wanted.

 

VIRTUAL DESIGN

I started with simulations in Excel when the parts were not available yet. The purpose was to virtually recreate the functioning of the new rotary cylinders to find out a configuration that worked. So, I reproduced the sequence of the cylinders.

Then, I created an offset between the 2 cylinders. To make the best gearbox possible, it was necessary to obtain:

• Reverse gear on an extremity of a shaft
• Not 2 gears engaged at the same time on the same shaft

 

EXCEL TABLE

The Excel table may seem hard to understand, but the main difficulty is because I made written in French!

The first column represents the positions of the rotary cylinders.

The second column is used to create an offset between the 2 cylinders. I tested all the configurations and keeped the ones that respected the criteria above.

Then, I used the supposedly working configurations to calculate the possible gear ratios. The sub-table called “Transitions” corresponds to the ratio between the two main shafts.

Then, on the sub-tables on the right, I put the gears on the primary and secondary shafts to get 3 different ratios on each shaft, placed in the correct sequence, for each remaining possible

configuration.

I got only 2 possible configurations (A and B). Other configurations exist, but they are symmetries of these two configurations, so they are useless.

I finally came out with only one possible configuration, which is the one I realized in Stud.io and in real life.

 

IMPROVEMENTS            

Compared to my previous gearbox, this new version features:

• Way better efficiency
• More compact, easier to integrate into a chassis
• More reliable gear selector
• More realistic and easier to understand
• The shifter can be less robust
• The input and the output are centred, and can easily be moved
• It is reversible, meaning that you can power both the input and the output
• Gears better staged

 

HOW IT WORKS?

The way it works is close to a dual-clutch gearbox - however, the exact same operation cannot be achieved due to the way the sequence of a rotary changeover cylinder is made.

 

The ratios are better staged than the ones of my previous gearbox. The reverse gear is situated between the 1st and the 2nd gear, making it more realistic.

• R: -0.375
• N: 0
• 1st: 0.25
• 2nd: 0.45
• 3rd: 0.50
• 4th: 0.75
• 5th: 0.90
• 6th: 1.50

 

A simple mechanism prevents from shifting from 6th to reverse gear. This mechanism can easily me moved to another place (for instance close to the shifter) though.

 

The centre changeover cylinder is used to create a 45° offset between the two other cylinders. However, to spare a part, you can use one of the following assemblies.

 

VIDEO

 

REBRICKABLE LINK

Here is the Rebrickable link if you want to test this gearbox! https://rebrickable.com/mocs/MOC-171508/Anto/ultimate-6nr-gearbox-with-forks/#details

 

I’m wondering what the gearbox on this summer’s supercar will look like. It should either be pretty similar, or totally different. What do you think?

Edited February 26 by Anto

[NoEXIST]

Really nice design! Actually what I needed for my stopped 1:8 project!

When these parts were released I didn't really think that people will be able to make such compact gearboxes!

[Jundis]

Really cool! Was just about time somebody uses the new components :-)

What does the middle selector do? I ain't attached to anything?

[zoo]

Very cool. Great to see a design using the new gear box elements. I compared the gear ratios to a 911 GT3 transmission and it looks to be quite close with the exception of reverse gear being a bit too fast. Thanks a lot for sharing the instructions. Good job!

[2GodBDGlory]

Very cool! It seems like these parts have been out for a long time without finding much use, so it's nice to see a fully developed gearbox using them! The selector rings are definitely unintuitive in how they should be set up, so good job making that whole spreadsheet to figure it out!

The main thing that stands out to me from this design is the way you were able to use the same selector rings for two different driving rings--you couldn't really do that with the old wave selectors, so I hadn't thought of it as an option! I'll probably be building this today to really see how it works!

3 hours ago, Jundis said:

Really cool! Was just about time somebody uses the new components :-)

What does the middle selector do? I ain't attached to anything?

He mentioned that the middle selector is only used to get a 45 degree offset between the two shafts, and then showed a couple options with knob wheels to get that offset without using that expensive part.

-One other thing I was realizing is that linear clutches could be used to get any kind of offset desired, as long as there isn't enough resistance to make them skip. Do you think that could open up more options for gear orders, @Anto? It wouldn't fit in this gearbox, though

Edited February 10 by 2GodBDGlory

[Davidz90]

Very impressive! I thought that the new parts will result in a very large gearboxes, but this is remarkably compact.

[Divitis]

Great work!

It looks like a very efficient design in terms of moving parts/attrition. 

I especially like how you tried to make the gear ratios realistic. Too often I see this overlooked.

[Stereo]

Now that I see your calculations, it's making me puzzled why Lego did the 45 degree offset in the two axle holes of the drum.  Cause I see that, to run shifters on the two sides of the drum, each drum has a LR**RL** pattern, which means they have to be 90 degrees offset from each other (+-90) to hit all 8 gear selections at different times.  And flipping the drum end for end doesn't change its pattern, it's a symmetrical part that way.  So you need to undo that 45 rotation between the drums.  The text engraved in the drum means flipping end for end "shouldn't" be a problem for expert builders but maybe they wanted that symmetry.

If you took a different design approach and had the intermediate shafts 90 degrees apart (say left+down from the drum) it changes to a DL*D**L* pattern, which needs a 180 degree offset to hit 8 different gears, still doesn't work with an axle straight between the drums.  I think this kind of right angle gearbox could be fairly compact (seems to fit in a 7x7x13 unit box, vs. this one is 5x11x13, so technically lower volume) but it doesn't fix the issue.

I can't even make it work for 4 drums on a single shaft; the patterns I find are permutations on (1A)4B(2A)1B(3A)2B(4A)3B which uses each drum's "A" engagement 90 degrees later.  For 3 drums of course you get 6 speeds by going 1A 2A 3A 1B 2B 3B so maybe that's the intended "simple" transmission?

[2GodBDGlory]

17 minutes ago, Stereo said:

Now that I see your calculations, it's making me puzzled why Lego did the 45 degree offset in the two axle holes of the drum.  Cause I see that, to run shifters on the two sides of the drum, each drum has a LR**RL** pattern, which means they have to be 90 degrees offset from each other (+-90) to hit all 8 gear selections at different times.  And flipping the drum end for end doesn't change its pattern, it's a symmetrical part that way.  So you need to undo that 45 rotation between the drums.  The text engraved in the drum means flipping end for end "shouldn't" be a problem for expert builders but maybe they wanted that symmetry.

If you took a different design approach and had the intermediate shafts 90 degrees apart (say left+down from the drum) it changes to a DL*D**L* pattern, which needs a 180 degree offset to hit 8 different gears, still doesn't work with an axle straight between the drums.  I think this kind of right angle gearbox could be fairly compact (seems to fit in a 7x7x13 unit box, vs. this one is 5x11x13, so technically lower volume) but it doesn't fix the issue.

I can't even make it work for 4 drums on a single shaft; the patterns I find are permutations on (1A)4B(2A)1B(3A)2B(4A)3B which uses each drum's "A" engagement 90 degrees later.  For 3 drums of course you get 6 speeds by going 1A 2A 3A 1B 2B 3B so maybe that's the intended "simple" transmission?

Yeah, those parts are mind-bending! I've spent the better part of the day designing and trying to design gearboxes with it, but there's so many options it's hard to determine what's even possible!

[Stereo]

1 hour ago, 2GodBDGlory said:

Yeah, those parts are mind-bending! I've spent the better part of the day designing and trying to design gearboxes with it, but there's so many options it's hard to determine what's even possible!

Yeah, I've got a feeling there's some way of getting it to do the right thing, maybe with a new part or just different way of integrating things.

For example, bar with towball into a half pin could access the 45 degree points, or if you could place a ball just past the axle the shift fork slides on. (this second one's not really secure enough; the thin liftarm constrains rotation but not completely.  I don't have the parts from the bike to test whether it functions in practice.  A 'dummy' second fork would make it rigid but get in the way)

The second one seems more practical for this gearbox, because the forks would be on the same axle, and thus have 45 degree offset in the 2 forks on the same side of the transmission.  But the first one has the side benefit of +8 gearteeth on one side, so if you have 32 tooth combinations on one side (8-24, 12-20, 16-16) you get 40 on the other (12-28, 16-24)

More offtopic, but I think a manual transmission could be interesting with these forks, something like this.

Edited February 10 by Stereo

[Anto]

I'm happy that you like this gearbox!

 

17 hours ago, 2GodBDGlory said:

One other thing I was realizing is that linear clutches could be used to get any kind of offset desired, as long as there isn't enough resistance to make them skip. Do you think that could open up more options for gear orders, @Anto? It wouldn't fit in this gearbox, though

I didn't check if they can allow 45° offset - anyway, it's not an elegant solution to me, a fixed assembly is still the best in this way. Moreover, a clutch wouldn't allow more shifting options. On the selectors, all the letters have to be aligned, so you need a multiple of 45° offset.

 

13 hours ago, Davidz90 said:

Very impressive! I thought that the new parts will result in a very large gearboxes, but this is remarkably compact.

Thanks! That's what I thought at first, but the new 2L driving rings allow to fit allmost everything inside the frames.

 

12 hours ago, Stereo said:

Now that I see your calculations, it's making me puzzled why Lego did the 45 degree offset in the two axle holes of the drum.  Cause I see that, to run shifters on the two sides of the drum, each drum has a LR**RL** pattern, which means they have to be 90 degrees offset from each other (+-90) to hit all 8 gear selections at different times.  And flipping the drum end for end doesn't change its pattern, it's a symmetrical part that way.  So you need to undo that 45 rotation between the drums.  The text engraved in the drum means flipping end for end "shouldn't" be a problem for expert builders but maybe they wanted that symmetry.

I had the exact same reflexion. The best explanation I came up with is that it was the "less worse" configuration for:

• 4-speed gearbox, like the Yamaha, where the 45° offset is used.
• 8-speed gearbox, when the 45° offset can be cancelled. But it's impossible to make a dual-clutch gearbox structure.
• 6-speed gearbox, where 3 cylinders are on the same axle. But it requires more space than an 8-speed geabox for fewer gears and only 6 positions on 8 of the stepper are used, so this configuration seems useless.

So I can't figue out why they made this pattern neither the 45° offset, this seems to only bring issues. That's why I'm so curious to see the next supercar.

 

11 hours ago, Stereo said:

Yeah, I've got a feeling there's some way of getting it to do the right thing, maybe with a new part or just different way of integrating things.

[...]

For example, bar with towball into a half pin could access the 45 degree points, or if you could place a ball just past the axle the shift fork slides on. (this second one's not really secure enough; the thin liftarm constrains rotation but not completely.  I don't have the parts from the bike to test whether it functions in practice.  A 'dummy' second fork would make it rigid but get in the way)

This is an interesting way to solve the 45° question, but it's a rather complex solution (more than the one I used), which doesn't explain why LEGO chose this design since they started from a scratch.

Like you said, the forks are really interesting to make a realistic manual gearbox.

[gyenesvi]

Great to see the first large scale usage of those parts in a fairly compact gearbox! I really like the systematic approach here, the way you guys are checking all possible configurations. I wonder if Lego also did such an analysis as well. Also curious what they come up with, whether they keep it at 6 speeds like this, or they go for 8 speeds rather, maybe adding a DNR selector before it, and whether we get clutch gears in new sizes. Also, whether they would make something like a 4-speed version of this in the upcoming 4x4, that's probably going to be manual. Can this be used for 4 speeds in a linear build? That would need new clutch gear sizes, right?

I wonder if it would be possible to make this more robust by adding a support wall into the middle (longitudinally) for RC use. I guess the geartrain that connects the two sides inside the frame could be moved outside the frame (so total length would slightly increase a bit). As far as I see, the only thing blocking it is exactly that connector drum that undoes the 45 degree offset.. I really hope they did that for good reason, and not just in a "let's try to make it a bit more universal, may be useful in the future" fashion :)

[SNIPE]

Finally, someone made a gearbox using these new parts. Now I wonder who will make an 8+N+R gearbox with internal centre differential! :P

Edited February 11 by SNIPE

[Stereo]

8 hours ago, gyenesvi said:

Can this be used for 4 speeds in a linear build? That would need new clutch gear sizes, right?

You could use the 1:1 input to output without the secondary shaft as the 4th speed, so it'd be 24:8, 20:12, 16:16 onto a second shaft, then probably 12:20 to bring it back in line:  gear ratios 5.0, 2.78, 1.66, 1.0.  Or use one of the smaller gears on the primary shaft (12:20 or 8:24) so you'd get 3.0, 1.666, 1.0, 0.6 leading out from the secondary shaft.  If you have the end of the primary shaft go into 65414 differential, and clutch engage the same side of it, then the other side is a 1:1 output, so that's a viable option that doesn't really misuse parts.  It does get in the way, though, as the differential's fairly large.

 

65414 differential is effectively a 28T clutch gear too if you don't install the spider gears in it.  I suppose that would pair with 12T so you'd have 28:12, 24:16, 20:20, 16:24.  That's 2.5 stud separation of axles though so not as nice to build with.  And it's 3 studs long without the ability to run an axle through it, not ideal.  

Edited February 11 by Stereo

[2GodBDGlory]

9 hours ago, Anto said:

I didn't check if they can allow 45° offset - anyway, it's not an elegant solution to me, a fixed assembly is still the best in this way. Moreover, a clutch wouldn't allow more shifting options. On the selectors, all the letters have to be aligned, so you need a multiple of 45° offset.

That makes sense!

7 hours ago, gyenesvi said:

 Can this be used for 4 speeds in a linear build? That would need new clutch gear sizes, right?

I've currently got a linear gearbox on my desk with 6 speeds that I'd like to share soon (Though one of them is the "forbidden" 14:20 gearing), so it is possible! The key I found is that it's actually very easy with these fork pieces to flip one of the driving rings from the top shaft down to the bottom, so you can have 24:8, 20:12, 20:14, 16:16, 12:20, and 8:24 all in the same gearbox. The challenge then is to figure out a workable shift pattern for it, which I concluded was impossible with doing 45 degree rotations of the wheels (Because the two gears with the flipped driving ring have to be next to each other sequentially, because of the requirements of where the gears have to go, and then the fact that the wheels require two spaces between the two engagement points). I believe I've solved this by basing a single shift on a 135 degree rotation, or three positions, which allows me to cycle through each driving ring's gears one at a time, which is quite predictable and easy to set up. I still need to design a viable 135 degree stepper, though!

7 hours ago, gyenesvi said:

I wonder if it would be possible to make this more robust by adding a support wall into the middle (longitudinally) for RC use. I guess the geartrain that connects the two sides inside the frame could be moved outside the frame (so total length would slightly increase a bit). As far as I see, the only thing blocking it is exactly that connector drum that undoes the 45 degree offset.. I really hope they did that for good reason, and not just in a "let's try to make it a bit more universal, may be useful in the future" fashion :)

That was my exact thought too! It should be possible, but it'd likely be bulkier.

6 hours ago, SNIPE said:

Finally, someone made a gearbox using these new parts. Now I wonder who will make an 8+N+R gearbox with internal centre differential! :P

Well, that one's not going to be possible without adding some kind of multiplier or secondary gearbox, because these wheels only have eight different positions! It'd be cool to see a more advanced one like that, though!

18 minutes ago, Stereo said:

You could use the 1:1 input to output without the secondary shaft as the 4th speed, where you just slide the clutch collar from one axle joiner to connecting 2 axle joiners, so it'd be 24:8, 20:12, 16:16 onto a second shaft, then probably 12:20 to bring it back in line:  gear ratios 5.0, 2.78, 1.66, 1.0.  Or use one of the smaller gears on the primary shaft (12:20 or 8:24) so you'd get 3.0, 1.666, 1.0, 0.6 leading out from the secondary shaft.

 

65414 differential is effectively a 28T clutch gear too if you don't install the spider gears in it.  I suppose that would pair with 12T so you'd have 28:12, 24:16, 20:20, 16:24.

Oh, are you suggesting sliding the ring itself to cause it to lock onto a second axle extender, to create a sort of linear clutch that you could then put any gear you want onto? If so, that's a very interesting idea, but it seems like the engagement point would be too specific to work well. I might want to try that, though!

 

Also, we're getting pretty off-topic. Would it be worth starting a separate topic for discussion of gearbox design with Yamaha selectors?

[Stereo]

26 minutes ago, 2GodBDGlory said:

Oh, are you suggesting sliding the ring itself to cause it to lock onto a second axle extender, to create a sort of linear clutch that you could then put any gear you want onto? If so, that's a very interesting idea, but it seems like the engagement point would be too specific to work well. I might want to try that, though!

If I'm understanding the pictures right, anywhere from 0.5 to 1.5 stud (ideally 1) works; which does mean it's not compatible with the shift drum.

Shifting on the secondary is a good idea though, if you only need the 1 extra higher ratio for a 4 speed that's probably fine.

Do the forks clip onto the previous 18947 3 length driving ring?  Stud.io doesn't have the new one to test with.

Edited February 11 by Stereo

[2GodBDGlory]

1 hour ago, Stereo said:

Do the forks clip onto the previous 18947 3 length driving ring?  Stud.io doesn't have the new one to test with.

Yep!  They work on all three driving ring variants, as far as I can tell

[SNIPE]

On 2/10/2024 at 10:47 PM, Stereo said:

One thing I found that works is this:

The white plate round with handle on underside is slightly too short, but because the stud goes into a pinhole its hard for the driving ring to slip out of place anyway.

Note that you must use this part and not this one

[Anto]

23 hours ago, gyenesvi said:

[...] or they go for 8 speeds rather, maybe adding a DNR selector before it, and whether we get clutch gears in new sizes.

It is the most probable option in my opinion, to stick with the previous 8+DNR gearbox architecture. But an 8-speed gearbox would be harder to make, because you can't easily get 4 different speeds on a same axle, like @Stereo explained.

 

15 hours ago, 2GodBDGlory said:

Well, that one's not going to be possible without adding some kind of multiplier or secondary gearbox, because these wheels only have eight different positions! It'd be cool to see a more advanced one like that, though! 

Yes, it's possible, but too complex for only 2 more gears. However, it's interesting to notice that the new 8 teeth stepper gear perfectly fits pins, so you can make something similar to my 8+N+R gearbox using a pin instead of a microphone, to have 1/8 of a turn in output for each turn in input - or more.

 

28 minutes ago, SNIPE said:

One thing I found that works is this:

This is interesting, but what does this solution add compared to the forks, except the price of the parts?

[gyenesvi]

15 hours ago, Stereo said:

You could use the 1:1 input to output without the secondary shaft as the 4th speed, so it'd be 24:8, 20:12, 16:16 onto a second shaft, then probably 12:20 to bring it back in line:  gear ratios 5.0, 2.78, 1.66, 1.0.  Or use one of the smaller gears on the primary shaft (12:20 or 8:24) so you'd get 3.0, 1.666, 1.0, 0.6 leading out from the secondary shaft.  If you have the end of the primary shaft go into 65414 differential, and clutch engage the same side of it, then the other side is a 1:1 output, so that's a viable option that doesn't really misuse parts.  It does get in the way, though, as the differential's fairly large.

As far as I understand this would not use only 2 shafts; that would be my goal for simplicity and space saving. Also, I know more gear ratios are possible with differentials, but that's also something I'm trying to avoid here, as they are bulky (too long).

15 hours ago, 2GodBDGlory said:

The key I found is that it's actually very easy with these fork pieces to flip one of the driving rings from the top shaft down to the bottom, so you can have 24:8, 20:12, 20:14, 16:16, 12:20, and 8:24 all in the same gearbox.

That sounds interesting, so you are saying that some clutch gear engagement could happen on the upper shaft, while the others could happen on the lower shaft? I guess that seems logical, as the fork is 3 studs tall, and the ball connection is in the middle, so can simply be flipped around. And then leaving out the 20:14 gearing, and one more, we could get a simple 4-speed version, using for example 24:8, 20:12, 16:16 and 12:20 for the ratios?

Wonder if the same could be done using two wave selectors as well on two shafts.. Would become more compact, as the forks/drums take up a lot of space sideways if we want to keep the main driveshaft in the middle. Maybe it is possible to sync two wave selectors in the right order using the new 45 degree knob gear on one shafts and the old one on the other..

28 minutes ago, Anto said:

It is the most probable option in my opinion, to stick with the previous 8+DNR gearbox architecture. But an 8-speed gearbox would be harder to make, because you can't easily get 4 different speeds on a same axle, like @Stereo explained.

How about the way above?

[SNIPE]

39 minutes ago, Anto said:

This is interesting, but what does this solution add compared to the forks, except the price of the parts?

It allows the drive ring to be on the same plane as the rotary shifter rathern than 1L higher/lower like with the orange shifter fork from the yamaha motorcycle.

Edited February 12 by SNIPE

[danielhstahl]

I've been assuming (perhaps incorrectly) that https://www.bricklink.com/v2/catalog/catalogitem.page?id=235271&idColor=86&ccName=6469530#T=S&C=86&O={"color":86,"iconly":0} will be a major player in the next Lego gearbox.  Which would imply two axles of cylinders still, but may explain the 45 degree offset.  

[Stereo]

5 hours ago, danielhstahl said:

I've been assuming (perhaps incorrectly) that https://www.bricklink.com/v2/catalog/catalogitem.page?id=235271&idColor=86&ccName=6469530#T=S&C=86&O={"color":86,"iconly":0} will be a major player in the next Lego gearbox.  Which would imply two axles of cylinders still, but may explain the 45 degree offset.  

Yeah, I played with that to try to add 2 speeds to the bike's 4 speed, with a relatively small lump added to the transmission (+3 length - 2 for the driving ring, 1 for the gears to get power to the 3rd shaft, +2 stud width):

If you name the outputs (left to right) abCDef, the shift patterns that are easy to do are aeDbfC (how it's set up; it currently has 'b' selected), eaDfbC, DaeCbf, DeaCfb.  Some of which use two of the black knob gears, some use one of each type. With the actual specific gears I'm using, it's -1.67, 1.67, 1.33, 2.78, 5.0, 2.22, which is six different ratios but not yet in a good order.  Would need a spreadsheet to work out the actual sequence needed, I'm thinking the 8:24 and 12:20 need to be the ones that get reused, otherwise 8:24 (5.0) is too far away from all the other ratios as it's a pretty big jump.

The 16T gears on the left are either reverse (-1.67) or 1:1 depending which 20T dark azure gear is connected, and fit in-line with the shift forks, so it's only adding 1 to the length of the transmission to use it as 1:1 instead of reverse, and width/height are unaffected.

Edited February 12 by Stereo

[2GodBDGlory]

13 hours ago, gyenesvi said:

That sounds interesting, so you are saying that some clutch gear engagement could happen on the upper shaft, while the others could happen on the lower shaft? I guess that seems logical, as the fork is 3 studs tall, and the ball connection is in the middle, so can simply be flipped around. And then leaving out the 20:14 gearing, and one more, we could get a simple 4-speed version, using for example 24:8, 20:12, 16:16 and 12:20 for the ratios?

Wonder if the same could be done using two wave selectors as well on two shafts.. Would become more compact, as the forks/drums take up a lot of space sideways if we want to keep the main driveshaft in the middle. Maybe it is possible to sync two wave selectors in the right order using the new 45 degree knob gear on one shafts and the old one on the other..

That's the idea! The big issue for any design, though, is that that driving setup requires gears 1 and 2 to be on one driving ring, and 3 and 4 to be on the other. Then you run into the issue that neither old nor new wave selectors let you select the two sides of the driving ring in order without switching some other gear in between. The result is that the actual gear ratios more or less need to be in a 1-2-3-4 order, but the shift order has to be something like a 1-3-2-4 one. If you're shifting the gearbox with a PU motor, though, you can easily just program it to go to the right gears in order, so it doesn't really matter then!

[1963maniac]

On 2/12/2024 at 5:11 AM, SNIPE said:

One thing I found that works is this:

The white plate round with handle on underside is slightly too short, but because the stud goes into a pinhole its hard for the driving ring to slip out of place anyway.

Note that you must use this part and not this one

What part is used as a towball and does part # 25893 share the same pinhole?

STEREO, That gear gearbox looks nice and compact, keep going with that idea.

BTW, Philo has these parts for Studio: https://www.philohome.com/studio/packs.htm 

Edited February 14 by 1963maniac