[GBC] Automatic Reverser - Experimental

[djm]

As part of Splat's To-and-Fro GBC, the mechanism that provides a means to halt the GBC if there is a jam I found to be particularly interesting. RohanBeckett then posed a challenge (half in jest?) to adapt the mechanism to automatically reverse the GBC for a short period of time and then for it to resume normal behaviour.

So I decided to take up the challenge and came up with the following. In the video, my hand simulates the case of a GBC mechanism becoming jammed by holding the output axle. Following the "jam" there is a pause, the output axle momentarily reverses, pauses once more before resuming the original rotational direction.

This mechanism is in the experimental stage and there is a minor defect. The defect, as shown in this second video, is that the differential continues to turn a very small amount even when there is no jam.

The gearing ratios are correct, so I suspect that the rotation is being caused simply by the rotation of the axle.

I'm posting this thread for two reasons;
1) to garner possible solutions to address the defect
2) to see whether anyone else has interest to/is able to improve/alter the mechanism.

I'd be open to any suggestions as to how the incidental rotation of the differential could be eliminated. I'd also be open to seeing other peoples' interpretations to make a mechanism to achieve a similar result.

For convenience, I have made the build instructions (PDF format) and the LDraw file (MPD format) available for download from BrickSafe. There are also a couple of images of the two sides of the build:

Regards,

David

Edited August 5, 2017 by djm
Link amendment.

[Davidz90]

Interesting mechanism! As far as I understand, the differential rotates because even without a jam, there is some resistance in the mechanism so that the torque is split between output axle and the "gearbox". This problem may get even worse when actual load is connected to the axle. Two solutions that come to mind:

-link the differential with another gear that provides some resistance (for example, by clutch).

-add some latch or friction setup to the gear shifting mechanism (specifically, the arm moving on the black, vertical axle between red parts), so that it needs a small push to start moving.

[Lasse D]

Great prototype @djm! I was trying to build a similar mechanism when my GBC ball pump was built earlier this year.

It is really hard to make the "rebound" part of it work well. I gave up and ended up using a much simpler mechanism:

There are blue pins holding the 8 tooth gears on the motor side and the mechanism gives a rebound when it slips, thus undoing the jam. It would be nice to use a more visually impressive mechanism like yours.

[djm]

On 8/6/2017 at 6:38 AM, Davidz90 said:

Interesting mechanism! As far as I understand, the differential rotates because even without a jam, there is some resistance in the mechanism so that the torque is split between output axle and the "gearbox". This problem may get even worse when actual load is connected to the axle. Two solutions that come to mind:

-link the differential with another gear that provides some resistance (for example, by clutch).

-add some latch or friction setup to the gear shifting mechanism (specifically, the arm moving on the black, vertical axle between red parts), so that it needs a small push to start moving.

Thanks for the suggestions. I has tried adding another gear to add some resistance but it tended to cause too much resistance. Similarly friction on the gear shifting mechanism while it would stop the incidental rotation, it caused the sliding arm movement to stick when moving between the two sides.

However, your observation about friction made me think a little harder. I have significantly reduced the gearing "path" required to turn the differential in the opposite direction and this has solved the incidental rotation. 

On 8/6/2017 at 9:46 PM, Lasse D said:

Great prototype @djm! I was trying to build a similar mechanism when my GBC ball pump was built earlier this year.

It is really hard to make the "rebound" part of it work well. I gave up and ended up using a much simpler mechanism:

There are blue pins holding the 8 tooth gears on the motor side and the mechanism gives a rebound when it slips, thus undoing the jam. It would be nice to use a more visually impressive mechanism like yours.

I'm all for simpler mechanisms - it means there is less to go wrong! I'm now working on eliminated the slider mechanism altogether. Instead, I'm thinking of adapting Sariel's 4-speed sequential transmission, changing it to be 2-speed forward and reverse directions only, with the gear changer driven from the differential's gear. All dependent on parts from BrickLink orders arriving.

Regards,

David

[RohanBeckett]

haha.. good to see my comment has borne fruit! ;)

 

Yes.. I agree.. it does look a bit complex... but I might try building it soon to see how it works!

[djm]

I've adapted the auto-reversing mechanism, replacing the sliding mechanism with a variation of Sariel's 4-speed sequential transmission. I'm satisfied that the resulting mechanism is more reliable than the original mechanism. Further, the problem where the differential continued to rotate by a small amount no longer occurs with this mechanism. Here is a short video of the mechanism in operation.

I've got a couple of tweaks to make to it. Once that is done I'll post some building instructions for it.

Regards,

David

[Davidz90]

There is something hypnotizing in all the little sounds it makes after stopping 

[djm]

As promised, building instructions are now available for this mechanism. The LDraw file is also available.

Feel free to Improve/alter this mechanism as you see fit. If you do, please post the results back on this thread.

Regards,

David

[RohanBeckett]

might try to build this, this weekend.. and see if it helps out the steering cup GBC that likes jamming every once in a while...

[RohanBeckett]

ok.. I built this last night... and it seeeems to work... but in reality... I had a lot of slipping with the two opposing crown gears... clearly being used in an unusual way! :)

I made it drive the steering cup module, which was a regular jammer, when it had a full hopper of balls... and unfortunately, it just didn't react quick enough, or drive in reverse for long enough, to unjam..

very impressed with the build of this little thing - so compact and sturdy!!  and excellent instructions! 

I think if we can find a way to get rid of those crown gears.. and delay the return-to-forward time.. it should work better...

Ultimately, I did just go and improve on the Steer Cup pickup.. and now it appears to be perfect - not one jam in 30mins of running... Will post improvements to the GBC General thread...

[djm]

3 hours ago, RohanBeckett said:

<snip>

I think if we can find a way to get rid of those crown gears..

</snip>

I just tried replacing the two crown gears with 3x 16 tooth gears and that works. The opposite gearbox position is then used but that is neither here nor there. Further, it removes an undesirable "feature" of the first build where the output axle was rotating in the opposite direction to the input axle.

I can see some slippage between two 8 tooth gears (close to the 40 tooth gear) as they are on physically separate technic parts. I suspect I could improve this and eliminate that slippage.

3 hours ago, RohanBeckett said:

<snip>.

. and delay the return-to-forward time.. it should work better...

</snip>

I'll have to think some on this one. Nothing immediately comes to mind.

Regards,

David

[RohanBeckett]

yes.. noticed the 2 x 8t's and thought that area could do with a bit more bracing ;)

... will try swapping out the crowns, and see how that behaves

as for the delay.. I guess changing the gearing on the small arm that flips the knob gear, so it runs slower would be the only option?

[djm]

Something like this, which eliminates the crown gears, tweaks the bracing and changes the gearing post the 40 tooth gear.

Build instructions

Caveat: only built digitally right now. I have not tried the changes as a physical build.

David

 

[Captainowie]

6 hours ago, RohanBeckett said:

Ultimately, I did just go and improve on the Steer Cup pickup.. and now it appears to be perfect - not one jam in 30mins of running...

Pfft, 30 mins. You know perfectly well that's nowhere near long enough to declare something "perfect" :-)

[RohanBeckett]

30mins is a big improvement on 1 min... which is how it used to behave, when a dump of 30 balls was dropped in :)

a dump of over 50balls = the odd one flicked out... but a small spill is better than a jam!

[djm]

On 9/26/2017 at 8:22 PM, djm said:

Caveat: only built digitally right now. I have not tried the changes as a physical build.

I'm glad I put that caveat in. The modification wasn't buildable (a clash between a pin and axle which I hadn't spotted). I've corrected that and at the same time improved the support for one of the 8T gears, replacing the tan Technic, Axle Pin without Friction Ridges Lengthwise with a 3L axle. This required a rejig of some of the other parts but I have built it to the newer instructions and it now seems more reliable. I have replaced last night's building instructions with newer ones but retained the same link.

One thing that puzzles me slightly is that even when in neutral, the gears and output axle still rotate. They rotate in the "forward" direction when my expectation was that they wouldn't rotate at all. You can feel there is significantly less force behind the rotation when it is in neutral than when in the forward direction but nonetheless the rotation occurs. If anyone could work out a way to make it rotate in the "reverse" direction when it is in neutral, that could address@RohanBeckett's desire for a quicker switch between forward and reverse.

Regards,

David

[djm]

18 hours ago, djm said:

One thing that puzzles me slightly is that even when in neutral, the gears and output axle still rotate.

Having slept on it, here's an experimental path for me to explore further:

The motor would be attached on the green axle on the right hand end. The left hand end of the green axle would run through to the differential. The gear shift would be on the (red) output axle with appropriate gearing to reverse the axle direction. I need to explore the reverse gearing a bit - if I can make it have the same ratio as the forward gearing, that should mean that any rotation from the clutch gears when in neutral should cancel out.

I built this physically and it appears worthwhile exploring. However, further exploration my me will have to wait until the end of October, post our LUG's Brick Show.

Regards,

David

Edit: et voilà. Bricks rather than liftarms to provide the same ratio.

 

Edited September 28, 2017 by djm
Et voilà