what do you think of this design?
How will you get the pneumatic to go to 3 different positions?
Looks like a sick mad nasty geared transmission. I think 33 had something like this. 494 also had a really cool CVT. (494’s was a demo, though)
I do not know as of now. It is a design in the workings.
Presumably, the output is the rightmost shaft? How do the three center gears interface with the output shaft?
We used 2 single solenoids with a double solenoid to make the actuator on our arm hold multiple positions firmly. FotoPlasma would probably be able to answer any questions you might have about it.
Looks good to me You just need to figure out a way to get the piston to stop mid-stroke… Maybe two pistons in series?? (don’t know if anybody’s every tried that
). There’s probably quite a few clever mechanisms to do the trick.
- Patrick
i think that this might be problematic when it comes to meshing the gears…this is the reason that most teams use shifting “dogs”
it still looks like a sold design but will only be effective if you stop to shift
with this would you have 4 possible positions? Or 3? (While the primary cylinder is in, and the secondary cylinder is in, it is in stage 1. When the primary is in and the secondary is extended it is in stage 2. When the primary is extended and the secondary is in it is also in stage 2. When both are extended it is in stage 3)
The largest gear on the center shaft is always meshed with the long faced gear on the output shaft. When it’s in the extreme position, it acts as an idler gear between the input and output gears.
From what our team found out from using a multiposition actuator, this seems like a very plausible design. The problem of a smaller volume of air taking less time to fill a smaller pneumatic piston can be remedied using flow control valves, but that’s an implementation problem, rather than design.
I would use a single and double solenoid(multi-positioning pnuematics) with three magnectic reed switches to stop you at the correct positions.
I would presume this would work quite well. I’m not as familiar with the programming side of it but it seems to me that this could be done.
If you use hardstops with the actuator, you’d only need one (max 2 for hysteresis) magnetic sensor.
Yeah you could do it with one, with three I think you could know what position you were in easier (maybe not, I know you can still tell with one but three would easily tell you where you are) to make it into a automatic shifting system.
Seems like after all that hassle, it would just be easier to do a 4-speed similar to 33, or 968. (I’m especially partial to 968, but it is more difficult to manufacture than 33’s).
Then again, I’ve found no need for more than 2 speeds.
You can get along fine with just 2, more is overkill.
John
Well If you go for 2 speeds, your probably going to have a manual tranny set up where the driver switches speeds as needed. Going more than 2 implies you are going to have an automatic that switches accoring to speed and load. (See: Team 33)
Either one works, but a 4 speed gives a hungry programmer more to work on
before you take my advice, realize that i’m not an expert (yet… )
it would apearr that there isn’t enough space between the gears that are being shifted. i would say you need at least double the face width of the stationary
gears, otherwise you would end up engaging two different gearsets in mid-shift.
then again, i didn’t see anything saying that this drawing was to scale, so this might be a non issue.
you are right, it is not to scale, but as aI said, a work in progress. I figured a team would use at most three gears, a “speed gear”, a “tourque gear” and a “driving Gear” for normal use. our past robots just used the drill motors. I have been in First for one yaer now and I cannot stop thinking about it
.
The gears that are on the piston are attached to the same rod.
Not at all. It actually works better when it is moving. We played the whole season on our transmission which follows the same principles. Just get the gears hardened and leave a little gap between the gears being meshed and unmeshed. I don’t recall how much ours is, but I can find out if you want.
Very nice transmission, it looks a lot like the one we build, did you ever see it? The nice thing is that it is much lighter than using a shifting dog, and requires much less resources. Instead of one piston, it uses two, so there is not multi positioning of pneumatics. Maybe you could use a servo instead for the multi positioning/shifting. The whitepaper on ours will be here some time soon, check it out when it gets posted.
We used a similar design for a simple 2 speed tranny this year. We also had no problems at all shifting on the fly, but we ground the outsides of the gears to points so that the would mesh better. I will say that you need to be very careful about the size, pitch, strength, etc of your gears, as it’s easier than you think to strip them.
Also, I’m also questioning the need of 3 gears, when you only really need a pushing and driving gear. I’m doubting that a mid-range would really help the acceleration of your robot THAT much, considering how short the field is, and how hard it is to make a turn at full speed.
That being said, I was thinking of a way to accurately shift one of these cylinders half way. I was thinking of using a split system of pneumatics with 60 psi and ~30 psi. shifting full distance would involve flooding one side with 60 psi. to shift half way, you lower the pressure on the extended side to a little more than 30 psi, then flood the other side with 60psi. Then P1V1=P2V2 does all the complicated positioning for you. The valve system would be complicated and/or impossible. I haven’t looked into that yet. but it should work in theory, atleast.