Drivetrain Gearbox Shifting Without Pneumatics

Hi CD!

So most of us have heard of the famous shifters that allows us to discard the choice between speed or torque in a drivetrain, and switch gear ratios on the fly. Almost all of the shifters I see use pneumatics, whether its a kit from Vex PRO, AM Super Shifter, or whatever. However, I have heard of teams using shifting gearboxes that do not require pneumatics.

Now this leads me on to ask, how do they do it? Is it electrical solenoids? What are the secrets and is it possible to build or purchase a repetitive and reliable one that matches up to the famous Vex PRO ball shifters?

Thanks in advance!

1 Like

Traditionally, the alternative was servos. Problem is, servos don’t have anywhere near the shifting force so performance suffers.

I remember some teams in the mid-2000s using things like window motors to do the shifting with a linkage in the middle of the robot, so it’s not impossible…but you’d need to plan a lot of things around it. Generally, that means pneumatics as the default option wins.

4 Likes

Pneumatics are the standard. I imagine that electric solenoids could be used (albeit possibly not well) so long as they meet the requirements for electric solenoids in the actuators section of the manual (that will be released at kickoff).

Do you have a specific motive for not using pneumatics?

I personally feel that if there are available alternatives to a shifting gearbox on a robot that does not plan to primarily use pneumatics, it’d be nice to look into.

1 Like

You could try using a CVT

1640 are the experts on CVT.

3 Likes

Buyer beware - this is many orders of magnitude harder than using a normal gearbox and replacing the shifter with (something not-air-powered). As designed, the linked model definitely won’t have the reliability or capability of a Vex PRO ball shifter.

1 Like

The alternatives are really just worse for almost everything. If you are at the point where you need a shifting swerve, a CVT might be a solid bet (albeit very hard), but otherwise pneumatics are far superior to any other shifting methods in FRC (there is a good reason everyone does it that way). Also, using a shifting gearbox is questionable in the first place because most teams/drivers wont benefit from two speeds anyway because they wont be able to utilize it well.

TBH, I haven’t heard of many top teams who end up with a non-pneumatic solution. There’s a bit of a design tradeoff to say “we have to have pneumatics for the shifters”. But, once you’ve made that decision, it opens up a lot of possibilities for lightweight and powerful mechanisms.

If you have pneumatics on the robot, it’s definitely the right answer for controlling the shifter gearboxes.

If you don’t have pneumatics on the robot, and are early in the design phases, you can decide to bite the “cost” (dollars, weight, current draw, etc.) of having a pneumatic system, and then design the rest of the mechanisms to take advantage of it.

Most teams who really really really need shifters will make that decision early on, and find themselves in the second bucket. I don’t know of many folks who fall between the cracks of these two scenarios.

2 Likes

I think when you look at the options available, the pneumatic shifter is the standard for a reason. But that doesn’t mean that in the future it won’t be replaced by something else.

As far as drivetrains go, it’s best to go with something you don’t have to worry about.

I believe it was one of either 33 or 67 in the 2010-2013 range that used a non-pneumatic shifting solution. IIRC it was window motors. I know, very vague but I remember it was used more recently than mid-2000s

My hedge wording is purposeful :slight_smile:

I vaguely remember 67 using a vex motor (the green servo motors) shifting design. Perhaps in 2012? I’m pretty fuzzy on it though - it’s been quite a while. I think they had problems with the motors drawing too much current while shifting as well.

1 Like

Looks like they’ve successfully used window motors in the past, but the 393s were a bit underpowered:

The short version is, in the modern era especially with the brushless options available to us, shifting isn’t really mandatory - to the point where if you are not using pneumatics elsewhere I would just not shift. You have to keep in mind that nearly all of the conventional wisdom about shifting drivetrains was developed in the 4 CIM era. You rarely need to build a robot that both goes as fast as the battery / motors allow, while also pushing as hard as the traction devices allow for an unbounded amount of time… you can definitely get close enough with 4 (or 6) brushless motors and good gearing.

That said, your robot should probably have pneumatics on it already, because they are extremely useful, flexible actuators that can serve any number of useful purposes far more easily than motorized solutions. And once you have pneumatics already, the opportunity cost of going to a ball shifter is really not that high.

3 Likes

Yes they are! As a matter of fact they are based not too far from our location!

Hmmm, my engineering sense is tingling:

  • Vex’s 3/4" bore, 1/2" stroke shifting cylinder generates ~35 lbf of force

  • The biggest solenoid I can find quickly generates ~270 ozf (17 lbf - imperial units are dumb y’all) at the beginning of its stroke, but it’s stroke is twice as long as necessary, so with a lever it gets close to the same force ~35 lbf. It uses 60W (5A) doing so apparently, which is well within the bounds of a PDB circuit and relay/motor controller (but too much for the PCM): https://www.mcmaster.com/69905k117-69905K521

  • As designed, that solenoid only pulls with current, then returns with a (too light to shift) spring. If instead we built a bistable spring setup as shown, we could probably use the solenoid to both push and pull, but rely on the spring to hold position without continuously drawing current:

So TLDR - maybe there’s a workable product here? There’s certainly a market - I’m not sure I’d want the complexity and risk of a pneumatic system if shifters were the only pneumatic device on the robot. [edit] OTOH, in that case we’d probably just go single-speed and avoid the extra weight and complexity of shifters altogether.

You could also do something with a rotary or linear motor of course, but those are sloooow

747 had a really cool motor shifting linkage, you can see it close up in their behind the bumpers episode: https://youtu.be/u2rbKC5VvmE

Andymark sells their servo shifter package. The servo they selected (the HiTech HS-422) is not particularly strong with on output of 56 oz-in at 6.0V. HiTech sells much stronger servos. The HS-805MG has 343 oz-in (21.5 in-lb) of torque at 6.0V. If you assume that the shifter adapter horn that is part of the AM servo shifter acts on the shifter shaft at a radius of 1" from the axis of rotation of the servo, that servo would produce 21.5 lbs of pull/push on the shifter shaft. That is pretty similar to the 25 lbs that a 3/4" piston @ 60 psi working pressure produces.

I recognize there are some additional losses due to the friction of the mechanism. But it still seems like that servo would be able to shift under load.

Most of the complaints from servo shifting deal with the speed that they shift at. To get enough torque, the motion is slow to actuate. It’s just worth using pneumatics if you have to shift. Then you have to go back and change everything you can change over to pneumatics, because once you go it’s best to go all the way.

3 Likes