Inspired by @Sariel's book I've been trying to incorporate a 2-speed transmission into a very high-speed car I'm building. Here's the essence of how it works:

Now, please believe me when I say that, in the actual car, everything is extremely well braced -- it's impossible for gears to skip, and in fact they don't.

The actual problem is this: using the fast gear, at very high acceleration -- I'm talking 2x BuWizz in Ludicrous mode -- the driving ring  is pushed out of the clutch gear , even when the driving ring is held in place by a rotary catch  which is itself extremely well braced. The result is lots of clicking and little movement.

I can already hear some of you saying "You just shouldn't apply so much force to ABS plastic; of course it'll bend!"

Nonetheless, I'd love to hear if you know of any super robust way to hold a driving ring firmly inside a clutch gear!

Use the lever instead of the wave selector actuated by a worm gear or a mini LA. 

having 2 driving rings in parallel might help to balance the force over 2 connections instead of 1.

Do you have a phone or camera that can do extreme slow motion videos? Mk ight be helpful to see what's actually happening when it skips; who's bending or moving to allow it to miss?

I wonder if using carbon or metal axles may help?

Don't forget, that this is just a toy, not rc car components.

Thanks @Zerobricks and @That_LEGO_Guy, I'll try those in order!

@amorti I've studied it quite a bit and there's really no bending involved. It's just that there are natural tolerances: the thickness of a gear or liftarm or bush is slightly less than 1M. Given the number of components involved (ring, catch, gear, axles, and the surrounding liftarms), the tolerances add up to roughly the length of the "teeth" of the driving ring. It takes quite some force to make all components shift to the end of their natural tolerance, but those BuWizz-powered buggy motors seem to be up to the task...

@Jurss Right, and this is definitely pushing it. But I've seen you guys do things I thought wouldn't be possible either. I'm still trying to learn where that boundary is...

12 hours ago, Zerobricks said:

Use the lever instead of the wave selector actuated by a worm gear or a mini LA. 

This would definitely let you put more force on the driving ring to hold it in, but is that the ideal solution? I think you can then end up with the selector fork and driving gear having too much friction to run smoothly. It would also be slower to shift, which could allow the 'splines' on the driving and clutch gears to be ground down.

Is there space to put a wave selector on both sides of the driving ring?

Would you be willing to use an O-ring to take away any slack between the red and blue clutch gears and the frame they're in?

2 minutes ago, amorti said:

Is there space to put a wave selector on both sides of the driving ring?

Not really, the second one would stick out of the bottom then...

4 minutes ago, amorti said:

Would you be willing to use an O-ring to take away any slack between the red and blue clutch gears and the frame they're in?

Which part is that?

BTW I haven't yet tried whether the old 2L driving ring would hold better. But that one can't be used with the blue 20-tooth clutch gear, so the speed difference between the 2 gears would be less unfortunately.

Just now, astyanax said:

Which part is that?

No Lego part number! Literally a rubber O-ring. 

https://www.amazon.com/1-9mm-4-8mm-Metric-Replacement-ring/dp/B01B3R8U1O?ref_=d6k_applink_bb_marketplace

4.8mm is the equivalent diameter of a Technic axle, this would take up about 1.9mm (~0.25L) of slack across the axle.

No idea if the red selector would grip better, but if changing gear doesn't make a real difference to speed, then it doesn't really matter if shifting is reliable as there's no need to shift.

However, does it maybe make more difference in absolute terms than you are thinking? When you're talking about 1400rpm, a small percentage change in gear ratio could be a big difference to road speed ?

46 minutes ago, amorti said:

However, does it maybe make more difference in absolute terms than you are thinking? When you're talking about 1400rpm, a small percentage change in gear ratio could be a big difference to road speed ?

I'm very happy with the current gear ratios:

1. In low gear (outer buggy motor output -> blue gear) with BuWizz on Slow, the car is still nimble but just slow enough to be driveable indoors without crashing into every single pieces of furniture.
2. In high gear (inner buggy motor output -> red gear) with BuWizz up to Ludicrous mode, it's perfect (outdoors!) for eliciting that "Wow!"-feeling from driver as well as spectators.

So yeah, I'd rather not change those...

Anyhow, I got some new things to try now.

21 hours ago, Zerobricks said:

Use the lever instead of the wave selector actuated by a worm gear

Awesome, this one passes the indoor acceleration test!

Setup: smooth wooden floor, opposite wall covered by a pile of pillows, then full-on acceleration with 2x BuWizz in Ludicrous mode.

Result: zero clicking, crossing the room in a fraction of a second, burrowing deep into the pillows with such force that the front spoiler got unhinged.

This weekend I'll do the outdoor test at the local basketball court. The asphalt will induce a bit more friction, so I wonder what will break next..!

PS: Verily, the gods approve..!

14 hours ago, amorti said:

No Lego part number! Literally a rubber O-ring. 

https://www.amazon.com/1-9mm-4-8mm-Metric-Replacement-ring/dp/B01B3R8U1O?ref_=d6k_applink_bb_marketplace

4.8mm is the equivalent diameter of a Technic axle, this would take up about 1.9mm (~0.25L) of slack across the axle.

No idea if the red selector would grip better, but if changing gear doesn't make a real difference to speed, then it doesn't really matter if shifting is reliable as there's no need to shift.

However, does it maybe make more difference in absolute terms than you are thinking? When you're talking about 1400rpm, a small percentage change in gear ratio could be a big difference to road speed ?

Instead of using an O-ring, I often use small plastic parts I cut out of the bottom of a 4L ice cream tub. I cut a strip, punch a bunch of holes in it, and cut several segments with holes in the middle.