pic: Shifting Drivetrain with Transmissions

[Jared]

loading photo ...

[Travis Schuh]

Originally Posted by Jared

It is geared for 6.33 fps in low gear and 13.55 fps in high gear (free speeds).

Why 6 and 13.5? That seems pretty slow. There is more room to make low gear faster while still being traction limited. We ended up with a 6fps low gear in an off-season bot last year, and found how slow it was a limit to how useful low gear was.

[Jared]

Originally Posted by Travis Schuh

Why 6 and 13.5? That seems pretty slow. There is more room to make low gear faster while still being traction limited. We ended up with a 6fps low gear in an off-season bot last year, and found how slow it was a limit to how useful low gear was.

With the 6 fps low gear, we get 48.7 amps per motor during a pushing match, which is why I chose that ratio. If we want to go faster, the CIM pinion could be replaced, increasing the speed to 7.38 feet per second in low gear and 15.81 feet per second in high gear, but we'd have 56 amps per motor in a pushing match.

What do you think ideal free speeds for a 2 CIM gearbox are?

I can also take high gear up to 18.27 feet per second (with low still at 6), but I think this might be too fast. Low gear could also be increased to 9.41 feet per second (71 amps/motor), but this is too high.

I can also do 6.33 low, 17.5 high, or 5.8 low 16.11 high.
If anybody is interested, here is the CAD.
https://drive.google.com/folderview?...&usp=sha ring

Why are the wheel wells indented so much for just 1" wheels? You're greatly decreasing your available bellypan space.

What's with the double set of sprockets on each wheel? Just lazy to make spacers, or do you have a reason for doing so?

What is the reasoning behind not having your gearbox plates backed by the frame? The spacers there make me iffy about the strength. If your goal is to save space with the inverted CIMs, ask yourself if it is really necessary that you save that much space. Because I feel as if lessening your wheel wells will save plenty of extra room and you won't require a design more complex than you need. I'm sure a 2 CIM WCP DS would work perfectly in its place and save you a lot more time and trouble than this custom design is worth. If you still insist on doing this kind of transmission, I suggest you take a look at this one I made.

If you really want that low a low gear, shoot for a high gear around 15 to 16 ft/s.

If you have any more questions about drivetrains or transmissions, feel free to hit me up with a pm. This has a lot of potential, but you need to have more focus on what you want to accomplish with it and how it will fit your team's needs and resources.

[DampRobot]

I wonder if you could get a little more size and efficiency out of your gearboxes by putting both CIMs around one idler gear that went to the cluster shaft, rather than the two idlers both going to the cluster shaft as you have now. I assume you have to use idlers because the cluster gear can't be made big enough for the CIMs to direct drive it.

[Jared]

Originally Posted by Andrew Lawrence

Why are the wheel wells indented so much for just 1" wheels? You're greatly decreasing your available bellypan space.

I can move the wheels out 3/8". The frame was just an example, so I just made the indent an even number.

What's with the double set of sprockets on each wheel? Just lazy to make spacers, or do you have a reason for doing so?

I made the spacer a configuration of the sprocket to save time, but I forgot to make the changes in this assembly.

What is the reasoning behind not having your gearbox plates backed by the frame? The spacers there make me iffy about the strength. If your goal is to save space with the inverted CIMs, ask yourself if it is really necessary that you save that much space. Because I feel as if lessening your wheel wells will save plenty of extra room and you won't require a design more complex than you need.

We used this design with a 3 CIM transmission last year and we saw no problems. Inside those spacers are two 1/4-20" grade 8 bolts that screw into threaded inserts welded into the frame. These same bolts also retain the outer plate too. It lets you take out the transmission by unscrewing two bolts and it just drops straight down out the bottom of the robot. We managed to remove a transmission, disassemble it, replace a dog, put it back together, and install it in under 30 minutes, with no practice doing this repair. This is way faster than any typical transmission, which require that there is room to slide the transmission out (which we usually don't have) in order to do maintenance.

Also, see a size comparison. https://drive.google.com/file/d/0Bzf...s3WFROeTA/view
Moving out the wheel wells saves me .875". Switching to the WCP gearboxes makes me lose over 9".

I would also argue that this design is no more complicated than the WCP gearbox, is significantly cheaper to our team, and lets us use 3.25" wheels because it has more ground clearance.

I'm sure a 2 CIM WCP DS would work perfectly in its place and save you a lot more time and trouble than this custom design is worth.

I don't want to run 3.25" wheel that wears down on the same shaft as a 3.1" OD gear. Also, this design is cheaper, and meets the needs of our team better.

If you have any more questions about drivetrains or transmissions, feel free to hit me up with a pm. This has a lot of potential, but you need to have more focus on what you want to accomplish with it and how it will fit your team's needs and resources.

The goal is for the transmission to be cheap, to have two speeds, and to be maintainable. These two transmissions would cost our team $321.71. Two WCP DS transmissions cost $520, and require that we spend $50 more on wheels. That is a $500 savings on two robots. It is also significantly easier to remove than a WCP transmission, and weighs almost the same.

I like your design, but I'm not a fan of cantilevered gears, especially with questionably fitting hex bearings. Last year, we had cantilevered gears, and we saw accelerated wear and eventually failure. Also, it seems that your design would be slightly thicker, as you do not have the cluster gear over the shifting shaft.

Originally Posted by DampRobot

I wonder if you could get a little more size and efficiency out of your gearboxes by putting both CIMs around one idler gear that went to the cluster shaft, rather than the two idlers both going to the cluster shaft as you have now. I assume you have to use idlers because the cluster gear can't be made big enough for the CIMs to direct drive it.

Yep. If I make the cluster gear bigger, it hits the dog. I've played with the gear ratios a lot, and it just isn't possible with vex gears.

Originally Posted by Jared

I like your design, but I'm not a fan of cantilevered gears, especially with questionably fitting hex bearings. Last year, we had cantilevered gears, and we saw accelerated wear and eventually failure. Also, it seems that your design would be slightly thicker, as you do not have the cluster gear over the shifting shaft.

How a cantilevered gear right next to the plate frame any worse than the cantilever of the pinion on your CIM output shaft? Also the cluster gear fits over the area where the pneumatic cylinder is, so no space is lost. You are right that flipping the CIMs takes out CIM area, but you are forgetting that the pneumatic cylinder is there and you still need a cutout in your bellypan large enough to service with. I think you'll find that in the end you're not going to be saving as much room as you think you will be.

[Jared]

Originally Posted by Andrew Lawrence

How a cantilevered gear right next to the plate frame any worse than the cantilever of the pinion on your CIM output shaft? Also the cluster gear fits over the area where the pneumatic cylinder is, so no space is lost. You are right that flipping the CIMs takes out CIM area, but you are forgetting that the pneumatic cylinder is there and you still need a cutout in your bellypan large enough to service with. I think you'll find that in the end you're not going to be saving as much room as you think you will be.

Good point about the pneumatic cylinder. I didn't even notice it on yours. For space, I think it all comes down to what you're putting in the middle of the robot. For us last year, the pneumatic cylinder was low enough that it didn't get in the way of the ball, but the year before we wouldn't have been able to fit our climber in if we went with that design.

Since my first post on your gearbox thread (where I shared that cantilevered gears worked well for us), we had multiple failures of these gears.

From my experience with hex bearings and hex shafts, there is a small amount of play between the shaft and bearing. This isn't really noticeable when the gear is in between the bearings, but it becomes significant once you have the gear hanging off the side. This makes it so that the gear is always slightly in the wrong spot, and you end up with a wear pattern like the one on the remaining teeth in this gear (http://www.chiefdelphi.com/forums/at...9&d=1395019667).

The cantilevered CIM shafts are steel as opposed to aluminum, and can bend under heavy use.

Originally Posted by Jared

Good point about the pneumatic cylinder. I didn't even notice it on yours. For space, I think it all comes down to what you're putting in the middle of the robot. For us last year, the pneumatic cylinder was low enough that it didn't get in the way of the ball, but the year before we wouldn't have been able to fit our climber in if we went with that design.

Since my first post on your gearbox thread (where I shared that cantilevered gears worked well for us), we had multiple failures of these gears.

From my experience with hex bearings and hex shafts, there is a small amount of play between the shaft and bearing. This isn't really noticeable when the gear is in between the bearings, but it becomes significant once you have the gear hanging off the side. This makes it so that the gear is always slightly in the wrong spot, and you end up with a wear pattern like the one on the remaining teeth in this gear (http://www.chiefdelphi.com/forums/at...9&d=1395019667).

The cantilevered CIM shafts are steel as opposed to aluminum, and can bend under heavy use.

Could you provide pictures of your setup? What you seem to be experiencing is very different than what I have experienced.

[Oblarg]

Seems similar to what I was toying around with a few months ago.

Any reason why you appear to have the WCP cams with the milled-slot bearing blocks? I can't see how those work together.

Is there any support for those gearboxes other than the two visible bolts with the spacers on them? If those are it, I'd be worried about that. Not having the plate in contact with the framing member removes a lot of rigidity.

Originally Posted by Oblarg

Seems similar to what I was toying around with a few months ago.

Any reason why you appear to have the WCP cams with the milled-slot bearing blocks? I can't see how those work together.

Is there any support for those gearboxes other than the two visible bolts with the spacers on them? If those are it, I'd be worried about that. Not having the plate in contact with the framing member removes a lot of rigidity.

The cams were designed to be used with the milled slot blocks.

[Oblarg]

Originally Posted by Andrew Lawrence

The cams were designed to be used with the milled slot blocks.

They were? I was under the impression they were designed for the clamp-style versa-blocks. We used the "milled-slot" blocks on 449 in a previous offseason project, and they couldn't really slide at all. Do you oversize the bolt holes to give them wiggle room?

[EricH]

Originally Posted by Andrew Lawrence

How a cantilevered gear right next to the plate frame any worse than the cantilever of the pinion on your CIM output shaft?

Where is the other end of that CIM output shaft? Right, securely in the motor, at the other end of the motor, where it's held securely so it just about can't move around and flex, just rotate.

If you really wanted to, you could model this as a statics problem. Draw a long beam, with a support at one end, a support near but not at the other end, and then a load at the end without the support. Now draw a short beam with one support, and the same load at the end. Apply all relevant beam equations. For bonus points, translate those loads into actual shaft loadings and see which shaft needs to be beefier to prevent bending (and uneven wear, etc.). I'd be willing to bet that the cantilever shaft supports need to be stronger, and it's likely to need a tougher shaft even though it's shorter. (general-case solution, obviously dependent on specifics of design)


Please also note carefully that this team did use cantilever last year and "saw accelerated wear and eventually failure". That statement right there indicates that for whatever reason, a cantilevered gear (presumably right next to a frame plate) is not holding up the same as a CIM pinion gear.

[Travis Schuh]

Originally Posted by Jared

With the 6 fps low gear, we get 48.7 amps per motor during a pushing match, which is why I chose that ratio. If we want to go faster, the CIM pinion could be replaced, increasing the speed to 7.38 feet per second in low gear and 15.81 feet per second in high gear, but we'd have 56 amps per motor in a pushing match.

What do you think ideal free speeds for a 2 CIM gearbox are?

I can also take high gear up to 18.27 feet per second (with low still at 6), but I think this might be too fast. Low gear could also be increased to 9.41 feet per second (71 amps/motor), but this is too high.

I can also do 6.33 low, 17.5 high, or 5.8 low 16.11 high.
If anybody is interested, here is the CAD.
https://drive.google.com/folderview?...&usp=sha ring

We use 5500 as our CIM free speed. Last year we ran 9FPS as our low (4 CIMs) and found this to be fine (no blown breakers, and I am pretty sure we had a couple pushing matches in there), although we didn't do a current analysis like you have done.

I personally would go with the 7.3 low, and 15.8 high given the options you have listed (assuming you are using a similar free speed to what we use), but then again we don't like slow robots. I should add that faster robots require more driver practice and good drive code to get the full advantage of them, so I would recommend picking speeds that fit your strategy/resources best.