pic: 2 speed V.3



Here is V.3. Personally I am pretty confident about V.3. I am going to submit a whitepaper soon.

What has been changed- As suggested by Andy Baker, The pockets for gears on the output shaft are .2" deep now. There is a "wiggle" now Andy. The dog has 3 tooth now instead of 4. The pin is much smaller so it won't interfere with the engagement of the gears to the dog. Everything is fastened with nuts and bolts. As many other pointed out that 3 fps might be too slow. Now low gear has a new speed. It goes 6.5 fps on low gear. Took more weights out of the gears. If there is one thing I have failed to do on this design is to be creative on the plates (the cutouts). The gearbox weighs 4 pounds. Thank you to every single person for their input on V.1 and V.2 thread.

It looks great! You’ve come a long way from v1 and it keeps getting better. Congrats

even more awesomer than before. you continue to amaze me with your 2 speed, i’ve learned a lot from it.

My only question, are you gearing it again after the sprocket that you put on the output shaft?

For Reference:
2 Speed Shifter V.1
2 Speed Shifter V.2

I just skimmed the other two threads, and didn’t see anything about this(correct me if I’m wrong), but instead of putting 8 lightening holes in each plate, why don’t you just use thinner plates? I’d suggest 3/16 for the CIM side, and 1/8 for the sprocket side.

That will take out more weight than all your lightening holes, and save you time in the process.

Minor nitpick, but in the gear closest to us, when you try to machine what you’ve shown, unless you use an incredibly small endmill and make about 500 passes, you’re going to end up with a round hole, and not a square pocket like you’ve drawn (even doing all that, it’d still be a filleted square)

The same is true for all the cutouts in the plates (assuming the material will be removed with an endmill).

Also, I still think this can be made a good bit smaller/lighter/cheaper (about $40 in two gearboxes). Here’s how.

Keep the 12T gears on the CIMs and have them both meshing with a 40T gear (instead of the 60T). Then the 40T gear meshes with the current 30T gear (no change). This concludes high gear. The CIMs will just barely fit side by side but they should definitely fit (I calculate .080 between them).

Then on the same shaft as the new 40T gear, put a 25T gear (to replace the current 45T gear there). It meshes with the current 45T dog gear. (no change) this concludes low gear.

So basically, replace the 60T with a 40T and replace the 45T non-dog-gear with a 25T.

This will give you a difference of 2.4:1 between high and low gear as opposed to your current Version Three 2:1 difference. The 2.4:1 difference could give you something like 13fps and 5.4fps.

Of course you will have to redesign your plates for the new center to center distances, but you seem to whip out new versions very quickly so I don’t think it would be much of a problem.

I know the AndyMark transmission uses .090 plates and several people have reported no problems, but in previous years I have seen dog shifting transmissions with 1/8" plates all wavy at the end of the competition. I think a 3/16" plate would be fine, but if you use traditional bearings (like the highly-recommended-by-me SPI BRF-06 and BRF-08) they will stick out beyond the plates. For Arefin’s design, this would be an issue on the CIM plate but the outer plate probably wouldn’t matter.

Bearings that are thicker than the plate they are in seem to work just fine (as shown by the AM and 2004 968 transmissions) but to me it just doesn’t seem right.

Of course you could always start with a 1/4" plate and mill it down thinner where there are no bearings, but that is a lot of extra work.

hmmm i guess i’ll comment on it now even though i’ve seen it first and already said a few things…

it’s looking really great for being sick and tired…feel better :slight_smile:

like some have said above…your gonna have fun making your sorta squares…

too bad you can’t make it so we can see it all nicely mounted up…

keep up the good work… i’m really proud of you

JVN’s Quick Lesson on Making Your Pockets CNC Machinable:
Step 1 - Put a .0625"R, .125"R, or .1875"R Fillet in the corner of all your pockets.
Step 2 - Done.

It is impossibly to cut sharp corners. In the corner of each of your pockets, there needs to be a radius of at LEAST the radius of your cutter.

(If you are using a 1/4" dia end-mill, you need a 1/8" Radius in each corner… If you are using a waterjet with a minimum radius of .02", you need to put that in each corner.)

I would throw down a 1/8" Radius, so it can be cut with a 1/4" Mill.
But then again… I usually just ask my machinists what they want to use, then throw that in there. (I am a sub-par machinist at best.)

Since everyone seems to be picking on your “square holes” you might as well fix this, and get 'em all to shut up! :wink:

Remember Design Principle #1138 - “Design for Manufacture”

Keep going Ary.
-John

Actually, our machinist prefers to put in a radius slightly larger for the cutter like if you have a 1/4" cutter then use maybe a .130" radius. He says this will make for a better cut because it is conturing the corner instead of just running into the corner and then switching directions. Also, it eliminates that little “squeak” sometimes when it hits the corner during machining.

I love it Arefin!! Definitely could get more creative on the cut outs like you said but it looks great. Maybe this one will be on a 1251 bot and everyone will get to see the final made product in the 2006 season =). Looks great man thank you for explaning to me all the gear ratios and mechanisms. Very cool Arefin great job. Listen to John about the pockets, John knows best! hehe :]

-Drew

Good work.

With all this Inventor experience, no you can draw up a whole robot, maybe you already have??

Just curious did you have Inventor taught to you or did you learn it by yourself/tutorials?

I am also wondering about the gears, maybe you have already answered this in another thread/post, but did you draw the gears?

If so how, I know making gears is not fun! :ahh:

I would like to keep both plates the SAME. It’s 4 pounds each gearbox. I am not worried.

Yes there will be a 24 tooth sprocket on the 4" wheels. If you would like to use a different diameter of the wheel, you will need to play with the numbers.

I only asked for suggestions how to make “THIS” gearbox look better. Not how to make a WHOLE NEW GEARBOX. I am not planning on changing the gears. If I mesh a 40 tooth gear with the 12s on the CIM (both)… I am concerned that the CIM motor cases will be conflicting with the bearing (on the plate) on the 40 tooth gear.

Andy uses .090 steel plate and then he bends it so you can mount it on the chassis.

That’s the only other change I am going to make in this transmission. The plates. Make it easier to machine. Add radius to every single corner. After the plates are changed I am going replace current picture with the new one, because I don’t want the forum to loaded up with all my transmission pictures. Thanks John for the suggestion.

It’s called not having the best time in your life for the past few days, and you want to keep your mind off of things.

Been there done that. Team 108’s 2005 robot was drawn in inventor by me. I also designed another robot that I proposed as a design. No I didn’t draw the gears. It’s VERY hard to draw each gear. I got them from Bostongear and Firstcadlibrary. I got it out of there and then made modifications. Actually Mike (punkrawker303) told me how to make holes and and draw/sketch/extrude in Inventor. Then I pretty much taught myself.

I apologize if any part of this post offended anyone, I am not in the best mood today (didn’t sleep for more than 35 hours). I know I shouldn’t be taking it out here, but I do have to answer the questions.

Nice design, how come I didn’t get a sneak preview? :wink: I’d like to point out that David, Cory and even Tiffany all touched on the radius issue though, credit where deserved and all that. But maybe I’m just whiny…

Also, I understand not wanting to redo the wheels, but those ratios worry me, I’d like that on record, the low gear is ok at 6.5 fps…but the high gear seems too high…in the hard to control range, making it useless, I think you’d be better off toning it down a bit further, say 4-4.5 fps for low and no higher than 8 or 9 for high, just to maintain control. I can’t see a use a for a wildly speeding robot and 6.5 isn’t that “low” really. Or maybe I’m crazy. You tell me.

Well that’s my input, probably not wanted as usual but that’s yet to shut me up.

P.S. I think you singed a few coattails by accident, might want to watch that until you get better.

2 thumbs way up!

Thanks for your input Matt. Even John told me that high gear was too fast. But if there has been robots and transmissions that has been designed and was faster and the drivers were able to pull it off on the field, every other driver in FIRST can pull it off too. There were bunch of robots out there which were fast. Why not give it a try?

I understand your concern about high speed. It maybe too high. But I believe that if one driver in FIRST can pull it off, every single other can too.

There is a very easy way to solve this problem without changing the inside of the gearbox. I can play with the sprocket, in the configuration I have right now… if I replace the 10 tooth sprocket wtih a 8 tooth sprocket (on the output shaft of the gearbox), the robot is going 5 fps on low and 10 fps on high. If I change the wheel size to 6 and change the wheel sprocket to 40 tooth, the robot is going 11 fps on high and 5.8 on low.

(I might be missing something)

I need someone to tell me and prove it to me saying that this transmission doesn’t work and the robot won’t move with this transmission, not the fact that drivers will need a lot of practice.

I am calling out the engineers out there. Please help me figure this out. Thanks.

I think running a 8-tooth sprocket on #35 chain is not a good thing, or so I’ve heard.

Did someone say it was going to be an 8? Anyway, some would say (me) running #35 chain on an FRC drivetrain is not a good thing (overkill, too heavy).

While the mechanics of a transmission are obviously important, I still think it’s very (if not equally) important to make sure your drivers can work with it. I can design a 6 speed crab-omni-maks belgian waffles drive (ok maybe not, but you get my point) but if it leaves the drivers in the station scratching their heads, its completely useless. Equally useless if its just too fast/slow/whatever. I just wanted to make that point.

I think I made all the points I wanted to, so I don’t plan on revisiting this thread anymore.

Back to fixing my amnesiac computers…

Arefin,
First off this is very good progress. I have a comment about the high speed. I would suggest you leave it at 13fps because although the drivers will rarely use it at full speed, for some games it can be very useful to be able to run around the field quickly. This year our robot was able to pull of a practical(not theoretical) 14.4 fps but we almost NEVER used that sort of speed on the field. It came in useful at the end of the match when we needed to get back behind the line, probably because the acceleration rate was high. If it was not that fast, I know we would have not made it back and probably lost the match.
-Bharat

I would concur that if the drivers can handle it, go for 13-14 fps. Our drivers rarely used all that speed, but they certainly could have. I’ve driven it at that speed and it is great for me.

The way I see it is greater speed = greater productivity. Just make sure you don’t have to downshift to turn, or you may lose that productivity.

Anyway, you can always adjust the sprocket ratios in the end. I haven’t found a 8T #35 (only in #25) but a 9T is available and you could always go slightly bigger on the wheel one (I don’t know how close you are to the 4" diameter though). :slight_smile: