pic: Team 948 2-Speed Ball-Shifting Gearbox CAD Drawing

[987HighRoller]

Has anybody run into major problems so far with their ball-lock shifters?

[David Guzman]

Originally Posted by jfelser

Yes, that is true about the model. The model is not quite correct in that the larger end piece of the inner shaft (the one attached to the cylinder) will be removable. The reason the shaft is two pieces is so that the larger spool will be centered with springs on the smaller inner shaft, which will allow it to lag a little when it pushes the balls outwards which make shifting easier.

Ok. Thaks for explaining.

[antoineft]

Originally Posted by 987HighRoller

Has anybody run into major problems so far with their ball-lock shifters?

We used a 2 speed ball lock shifter last season and thought it worked out great. Assembly was a little annoying until we figured out that using grease to hold the balls in place while sliding the parts together works very well!

We also just built the plunger as 1 piece instead of using springs. This seemed to work pretty well because the air in the cylinder basically acts as your spring. I had the rookie kids this year pull apart the gearboxes to "get their feet wet" plus I wanted to see what they looked after a full season. The insides held up even better than I would have thought. The plunger definitely showed lots of tiny "dings" but is far from failure. We put it all back together and it still works great.

--Antoine

[camtunkpa]

Three years strong and I don't think we'll go to anything else....at least not in the near future. The concern Sanddrag has expressed about shaft wear and such is NOT a problem at all. We do measure sizes before and after season when looking at what we could improve and we have not discovered wear. This is simply because we harden the pieces we need to and keep a good amount of lubrication. So rather than posting something you DON'T know about from experience, maybe you should let the people who have built these talk about the down sides. I don't mean to start a fight with anyone, but I just can't understand how people who have not used these transmissions can discuss any problems with them. The biggest negative we've seen is these transmissions take time to machine with our team capabilities. You can make them with just a lathe and a mill. They are that simple, but having access to a wire EDM or a water jet would greatly help reduce the amount of machine time.

Great to see 948 investigating this design. Hope to see it on your robot this season! Good luck

[Jeremiah H]

Very interesting, indeed.
I am not familiar with the concept of ball lockup shifting gearboxes, indeed this is the first I have ever seem of them. I am in charge of developing a godd two-speed gearbox for '07, so here are my questions (more will follow after this) Your expertise and help is greatly appreciated.
I understand how the balls are pushed into slots in the gears, but what is it that causes the thus engaged gears to turn the output shaft? In a dog shifter gearbox, the dog is positively engaged on the shaft (still allowed to slide) and so whatever gear the dog is locked into thus turns the shaft. I cannot find any such mechanism in this drawing. Perhaps I am missing something, but I think this is an important consideration.
And, following that, what happens to the gear (with its ball slots) when it is disengaged and must freewheel on the output shaft? My understanding of the design at present leads me to believe that it would bounce and chatter considerably, possibly even locking the entire gearbox up if it doesn't properly disengage.
Just my thoughts, if someone can answer them, I would greatly appreciate it. Thank you in advance. -JH

[camtunkpa]

Originally Posted by Jeremiah H

Very interesting, indeed.
I am not familiar with the concept of ball lockup shifting gearboxes, indeed this is the first I have ever seem of them. I am in charge of developing a godd two-speed gearbox for '07, so here are my questions (more will follow after this) Your expertise and help is greatly appreciated.
I understand how the balls are pushed into slots in the gears, but what is it that causes the thus engaged gears to turn the output shaft? In a dog shifter gearbox, the dog is positively engaged on the shaft (still allowed to slide) and so whatever gear the dog is locked into thus turns the shaft. I cannot find any such mechanism in this drawing. Perhaps I am missing something, but I think this is an important consideration.
And, following that, what happens to the gear (with its ball slots) when it is disengaged and must freewheel on the output shaft? My understanding of the design at present leads me to believe that it would bounce and chatter considerably, possibly even locking the entire gearbox up if it doesn't properly disengage.
Just my thoughts, if someone can answer them, I would greatly appreciate it. Thank you in advance. -JH

The balls are what cause the engaged gears to turn the output shaft. Comparing this to a "dog shifter" the balls are your dogs. Half or more of the ball remains in the shaft and the other half goes into the slots in the gear. two things can happen when the balls are disengaged. They depend on the design. The original hodaka design had a small groove in the gear and used the balls to act like a ball bearing when not engaged this means the balls stick out from the shaft a couple thousandthsand the gear rides on the balls. For simplcity we have not put this groove in our design. As Brad stated we leave between 4-10 thousandths of air gap between the shaft and the gear. In our design the balls fall flush into the shaft. Then all we do is use a thin oil/grease mixture to lubricate the gears and shafting. The gears have some space to move on the shaft, but not as much space as it seems you are getting at. Also we have never had the transmissions jam up. I don't have time to look for it, but we have produced a white paper on this type of shifting and feel free to PM myself or team222badbrad with any questions.

[team222badbrad]

How The Ball Lock Transmission Works:

Ball Lock

[jfelser]

Awesome video badbrad!! It explains the transmission very nicely. One question: How deep do the slots for the balls in the shifting gears go? A full 5/32? Or less than half of the bearing. It seems that slightly less would be the way to go, maybe 1/4 inch deep into the gear with a 5/16" EM (our design uses 5/16" bearings as well). I would think that the balls would disengage easier.

[team222badbrad]

Originally Posted by jfelser

Awesome video badbrad!! It explains the transmission very nicely. One question: How deep do the slots for the balls in the shifting gears go? A full 5/32? Or less than half of the bearing. It seems that slightly less would be the way to go, maybe 1/4 inch deep into the gear with a 5/16" EM (our design uses 5/16" bearings as well). I would think that the balls would disengage easier.

Sorry for taking so long to get back to you.


How far the balls will protrude out of the shaft depends on the diameter of your "Ball Plunger" + two times the diameter of your balls - the diameter of your shaft.

If your plunger is .515" and your balls are .3125", then your slots must be atleast 1.14" from tip to tip. Each ball upon being "Locked" would travel .07" above the shaft if your shaft is 1".

1/4'' = VERY BAD!!! If your balls were to travel past halfway above the shaft your shifting transmission would then become a single speed or perhaps shred to pieces. The balls would become trapted between the gears slots and shaft because of the rotational forces you put on them.

This attachment should help you out.

77 thousands is plenty as thats what the motorcycle transmission I used in the video had.

[jfelser]

Ah, thanks very much for the warning!!!

There is a very good chance you'll be seeing a transmission like this on NRG's robot this year... we talked to our machine shop about it a week or so ago, and they were very optimistic about the design.