pic: Four-Speed Ball-Lock Single-Pneumatic Transmission Design



This transmission concept is meant for a swerve drive, or other system that relies on one power plant, so to speak.

It is powered by two fisher price motors attached to AndyMark planetary transmissions, which have CIM output shafts and bolt hole patterns, so you could substitute two CIMs in there.

The lowest gear should overcome the static friction between roughtop tread and carpet, meaning the wheels would “burn out” instead of stalling under max power. The next three gears are for normal driving.

The shifting method is ball-lock, which you may have to look up for yourself. As the plunger moves down the shifting shaft interior, it pushes ball bearings into grooves in the gears.

The transmission shifts with a single pneumatic, which uses two solenoids to control exhaust, allowing the piston to move to any position while exerting full force on the way (and still acting like a spring in case of jams).

I posted this to get some feedback; I welcome any questions or comments. Thanks in advance!

Here is a very cool demo of a ball lock transmission for FIRST from team 222:

One question: Are you concerned about the precise positioning of the pneumatic cylinder? I know that in my tests with a double solenoid and magnetic reed switch, I had some issues with repeatability.

Yes, we are concerned and we are currently working out a test bed to determine if the precision required is possible. Our proof of concept (actuating the solenoids manually) proved we could get multiple positions but was not nearly as precise as we need. The next step is to use software to do a timed set of actuations.

Looks good! I’m going to reiterate Jared’s caution about using a double solenoid setup. The first ball lock we built in 2004 was a three speed. We experimented with using a double solenoid setup but there was no real repeatability. What we ended up doing instead was using two cylinders for the three selections. We joined our transmissions together through shifting forks and a linkage. One cylinder would shift to first and third gear while the second cylinder acted as a wedge to shift into second gear while we vented the cylinder for shifting between first and third gear. The setup worked well for us, but was alot of weight and alot of work for a third gear that saw very little action. Not to discourage this design in any way, but there are very very few games where you would need more than 2 speeds.

Also another tip on the plunger shaft, using the pneumatics as a spring doesn’t always work. Your best bet if you plan on making a solid plunger shaft would be to get it heat treated. We always run a spring loaded plunger shaft. This year we had a slight alignment issue in one gearbox where the balls were getting jammed and even though we had springs the springs were fully compressed and the cylinder acting as the spring at this point nearly destroyed our plunger shaft.

Anyway take your time on the ball lock shifter and enjoy it. They hold up and function very well. We’ve been running ball lock shifters since 2004. Each year they keep getting smaller and simpler. Best of luck!

We spent about a month working around different solutions the the problem of moving the plunger shaft into more than two shift positions. Personally I would prefer a lead screw but we got to the point that having the screw being .25" in diameter and then having the plunger/spring assembly on the end we weren’t confident it would be possible. Like you said, if something were to bounce around in there chances were it would snap off the plunger.

I know it was discussed in team 222’5 2004 thread but, if anyone has ideas for shifting a ball lock in 3 or more positions it would be greatly appreciated.

Would a servo (or a gang of servos) work, with a decently long lever arm? You’d have to work out the linkage to operate it, but I think you could do it. Many moons ago, my dad and I discussed a ball-lock with servos to shift, but were not having an easy time figuring out how exactly to do it.

I’m not sure the amount of force required to make a shift. Our main concern is that the servos can be painfully slow. Adding a lever arm could potentially increase the time it takes. I hadn’t thought of ganging servos before… that is an interesting thought.

How about multiple cylinders?

If you get a 2" stroke piston and attach a 1" stroke piston to the end of that, you can achieve 0", 1", 2", or 3" of length by firing them in sequence.

Bimba also sells multi-position cylinders that essentially accomplish the same thing in a more compact package. We used one on our Lunacy bot - you can custom order a wide variety of StrokeA/StrokeB combinations that are then controlled by a pair of solenoids.

I didn’t realize those are legal… we’ll look into it, thanks.

I am curious as to the details of your pneumatic setup, especially programming-wise. Were you using sensors, like MRS’s or some such? Or were you timing the shifts?

Also, how were the shifts not predictable/repeatable? We have only experienced “coasting” so far, where the piston keeps moving a bit after solenoids close, as the tubes are filling with air. Did you have any different problems?

Thanks a bunch!

Our shifts were not reliable because of the coasting you mentioned. We actually didn’t even get to the stage of running the cylinders in the transmission because of the coast It is very difficult to get exactly the same amount of pressure at both ports of the cylinder. We didn’t spend alot of time once we couldn’t really get rid of the drift.

Servos would definitely work, but they would be slow to shift. Our transmissions take very little pressure to shift. They will also stay engaged for quite a while without any pressure at all.

You could still use a lead screw setup, but use a linkage to connect it to a spring loaded plunger shaft.

I’m sure there are many other ways to shift this, but I’m drawing a blank right now. It’s too bad multi-position cylinders aren’t FIRST legal.

Why the adherence to the spring-loaded plunger ideal? Do you guys shift at full pressure? I had the idea to pick a spring strength, then set your pressure to provide that spring strength and nix the spring altogether. Have you guys ever tried this low-pressure shifting without a spring?

In 2009 they were legal. We used one to control the deflector angle on our shooter.

<R71> part D states:

In addition to the pneumatic cylinders provided in the Kit Of Parts and the “free” pneumatic cylinders available for order through the Free Pneumatic Components Order Form, additional air cylinders or rotary actuators may be used. Cylinders may be of any configuration, and may be of any size up to a maximum of 24-inch stroke and 2-inch diameter.

We have shifted at low pressure however the problem comes if things get out of alignment a solid plunger shaft pneumatics are going to keep pressing at a constant PSI trying to make this plunger push the balls out to engage the gear. Well the balls will eventually wear the portion of the plunger shaft that pushes the balls out. This is where our use of a spring loader plunger comes into play.
While we are still applying pressure trying to shift, the “floating” spring loaded portion of our plunger shaft is rotating on the plunger shaft and not wearing as bad. Also if the solid plunger is held rigid and not allowed to rotate you will see increased wear either on the plunger shaft or the balls themselves. Spring loaded or not is something we’ve tossed back and forth many times ourselves, but keep going back to spring loaded because it works and I’ve heard of others having issues with solid plunger shafts that are not heat treated. I will say though we tend to over design things, but we’d rather be overbuilt than underbuilt.

That is very useful advice, thank you!