Help Understanding Transmission

Hi. Could anyone knowledgeable enough explain one bit about torque with multi-speed transmissions systems to me.

First, I fully understand why a transmission designed like a planetary gear by wrapping a chain around in place of the orbit gear doesn’t really work. Here is a nice animation by the awesome jpearman:


I’ll label the high-speed version as T=1, v=1. The high-torque version ends up as T=1 and v<1 instead of getting the trade-off you’re looking for. This is due to back-driving a motor because of nothing locking the gears in place. Although I expect the torque might be minimally above 1 if some friction is able to help against back-driving a motor, it’s certainly nowhere near what you’d want from the loss in speed and application of two motors for torque.

Second, I fully understand drive-line windup such as on H-drives and how a differential can drive two sides at different rates to avoid drive-line windup (which I’m sure we don’t really care about here like we do with military vehicles). So, yes, I understand the relative motions of the gears within a differential.

Onto my problem, which is trying to understand how a bunch of the transmission systems I’ve seen people design with differentials don’t suffer from the same back-driving issue as the planetary gear/chain models I’ve seen people design. As far as I’ve been able to tell, you have the same torque limit in the high-torque mode as in the high-speed mode due to back-driving a motor. Here is one of several examples I’ve looked at between the VEX EDR and VEX IQ forums:
http://www.vexiqforum.com/forum/main-category/vex-iq-robot-showcase/11329-apalrd-s-cvt-evt
Here is a different design, and I totally get why this works:


Here we have a worm gear, which is usually self-locking and so quite hard to back-drive. Either in this video or on the discussion somewhere on this forum they mention too much lubrication stopping the self-locking becoming an issue, just as I would expect.

So what is it about the version without the worm gear that actually makes it anything more than a fancy way to do less when you move to high-torque mode? Why is back-driving the motors not an issue? The comments on the worm-gear version being over-lubricated match what I expect so well, so maybe I actually do understand correctly?

Thanks for your time!

@callen your understanding of the limitations of the planetary transmissions is correct.

If both input sides have motors attached to them then speeds add up, but torques do not. The output torque corresponds to the weaker of either the first or the second input. Even though automotive transmissions are usually some sort of planetary gear system, they connect single power output with a single power input. At any given moment of time most of the planetary gears are locked down such that there is only one gear path available for the power to flow through.

In order not to lose any torque due to either … or constraint, you would need to lock (break) or shift the gear path to ensure only a single power path exists at a time. Here are a couple of examples:

Shifting:

Locking / Braking:

If the there is a non-shifting, non-locking transmission that doesn’t “waste” any torque - I am not aware of it. And, according to my understanding of physics, it is not possible, if power could freely flow between output and both inputs.

The non-shifting transmission on in the video linked in OP is the locking type, by the nature of the worm gear. It is not a magik solution however, because this comes at the expense of some power loses due to the friction between worm and worm gear.

I have more details about the graph showing friction losses in this thread:[https://vexforum.com/t/non-shifting-multi-speed-transmission/34201/1):
https://vexforum.com/index.php/attachment/586348f070727_TransmissionPowerCurvesGraph1.jpg

But the gist of it is that if worm gear is less lubricated and has more friction (violet line) then low gear mode delivers more torque. And if worm is well lubricated (less friction/green line) then high gear is more efficient at the expense of lower peak torque in the low gear mode. Blue line shows shifting transmission for comparison.

If you need to mostly operate in the slow low gear mode with the short bursts of high speed, then worm gear based transmission might work, because you would waste extra power (to friction) rarely and mostly drive in an efficient low gear mode, when the worm gear is self-locked.

The biggest issue with worm gear based transmission is that the only worm, worm gear, and bevel gears available in VEX are not that strong and may not last long due to the shock loads and/or wear and tear. You would need to experiment to see if it is feasible for the specific situation. What would be the vehicle weight, driving pattern, max required acceleration, etc…

So, before discussing a shifting transmission, which is many times proven and relatively reliable option, I would like to remind everybody that “direct drive” is kind of a transmission too. It is an electric continuously variable transmission (CVT) that uses a fixed voltage battery as its power input and, thanks to the motor controllers, could provide variable voltage/power to the motors.

However, if you are sure that direct drive doesn’t give you enough speed and torque range then you have to consider a shifting or locking transmission. One of the biggest consideration would be how to do the shifting or braking. Do you use pneumatics or an additional motor for that?

Is there a solution to do the shifting, without using extra motors or pistons?

While leaving that question open, lets take a look at this video (from this thread):

In this transmission @Mr_L_On_Yoshi is providing symmetrical power inputs. This lets it avoid issues related to the backdriving of the weaker inputs. The above transmission is not designed to extend speed or torque range beyond what direct drive could offer. Instead, it lets you have an additional power output depending on the relative direction of the motors (and, yes, it is a hint for the question above).

Thanks. That makes so much more sense. I couldn’t figure out how they weren’t just “wasting” torque. I’ve found enough things like these that I started doubting myself and was trying to search for how they could possibly be self-locking.

And thanks for all the extras!

Ooh! Clever! Yes, I see the backward (for wont of a better term at the moment) use of the differentials, which in this case essentially uses the perpendicular inner gear to lock the two inner parallel gears together while the large outer gear is locked between the motors that can drive the strafing wheel. Then, when running the strafing wheel, the chain prevents the two sprockets from rotating against each other, locking things down the other way to allow the large outer gear to engage. Well, at least I think I see its functioning well.

Hopefully they managed to get the center of mass right over that strafing wheel. After all the beautiful work on the drive system, it would be a shame to not be able to use it as intended. But with such nice design work here, I bet they could handle the center of mass well of the external constraints didn’t prevent it.