33 Killer Bees 2019 Offseason Drive Gearbox

[Nick_Coussens]

Below is some renders and a link to the CAD model for the drive gearbox we have developed that is based off of our 2019 season drive gearbox. This version utilizes a 4" wheel instead of the 6" wheel we used in 2019, and thus allows everything to generally be smaller.

Some basic information:
-13.21fps free speed with a final ratio of 7.5:1 (12:60 20DP spur gear stage, followed by 32:48 20DP spur gear stage, 4" wheel)
-3.70lbs projected weight from SolidWorks (based on 2019 we expect this to pretty accurate).
-Design allows for air to be hooked up to the motors for quick cooling. We had this feature on our 2019 robot but never needed to use it. We just designed it in here to be safe.
-Gearbox is fully enclosed to keep any debris out. We were happy with this feature and have decided to do it again in the future, and likely try to do this on other robot features in the future when possible and convenient.
-Removed the mag encoder mount that the 2019 version had. We had this designed in on the 2019 version but never ended up mounting it on as our programmers were happy with the performance of the internal encoders on the NEO. This would always be pretty easy to add back in later if desired.
-Yellow plate is aluminum to allow the NEOs to heat sink through it and the frame. The other plates are nylon plate with the casing and spacers above and below the frame tube printed out of onyx on our Markforged.
-Design could easily utilize many 4" wheel options; VEX Colson wheel used in the model for example.

GrabCAD link here

CD link here to reference our 2019 drivetrain gearbox.

[mpirringer]

Very nice - We (!989) have a couple of gearboxes under development too and other gearing. Wish we could have a markforged but our Prusa clone has to do and our filament is mostly Hobbyking. Will release inventor files if and when it works to satisfaction. We are still running on CIMs as we got them and the matching controllers and are not in a financial position to switch Some of or plans you can see here

And we will release soon (hopefully) a video on a “skateboard” made out of Hips and a 36:1 (orso) Planetary for climbing, a 404:1 Planetary hopefully delivering close to 300ft/lb torque at 6-10rpm (powerd by 2 Cims) Printed out of Plain Nylon (bridge and Hobbyking CX12) the 404:1 is non FRC related project done by some of our members as a “graduation project” and our rack and pinion is being tested but I posted on that repeatedly. Glad to see someone else using 3d printing to do gears here is whats printed so far for the 38:1 (planet carriers and most the gears for the 2nd stage still missing

!
SANY0030|690x388

And an early prototype (dimensional only) for the planetary in the wheel (this is an 8 " we worked out the math for a 5-8:1 in a 6" but have not decided on the final ratio

And the rack we hope will be the core to our climber/lift

Anyway - we are wishing you good luck with your 3d printing efforts in particular and robotics in general

[wgorgen]

I had heard that you guys were plumbing cooling air into you NEO motors through the unused bolt holes. I see that this design has incorporated this as well. Very clever. I have a few questions:

1. Does the pneumatic fitting just screw into the plate that the motor is mounted to such that the hole with the pneumatic fitting is aligned with the bolt hole? Or is the fitting actually connected to the motor somehow?

2. Did you use the 4th hole for exhaust or does the air just exhaust through the gaps between the cap and the base plate?

3. Did you do any experiments to measure the effectiveness of the cooling? Do you have back to back testing with and without the cooling air? How much cooling air did you end up using?

4. Did you turn the cooling air on and off based on the internal temperature measurement, or did you just have it on for the whole match? If you turned it on based on the internal temperature, what temperature did you decide to use?

Hope you don’t mind all the questions. I just think this is so clever and I am wondering how effective it was.

[Murvar]

1. The motor is mounted with a hollow bolt which the pneumatic fitting is then screwed on.
2. Yes, the two unused mounting holes are used for exhaust.
3. We ended up never having to use this with the neos this season iirc. The neos were mounted to an aluminum plate, and we applied thermal paste between the two for optimal heat sinking. In competition, they never got remotely hot enough to warrant using air. In practice, the motors got too hot maybe once after driving for over an hour straight. Usually the motor controllers would get too hot first. When we did use this in the past for CIMs, we used an off board compressor and just released the air through the fittings to cool down the motors, either during practice or between elimination matches when the motors didn’t have enough time to cool themselves. I’m personally not aware of any scientific testing that we did to measure the effectiveness other than “it’s less hot now.”
4. We did not do any active tracking of internal temperatures. In the past it went along the lines of “the motors feel really hot, let’s blow some air through them.”

[Nick_Coussens]

To add on to what Murvar articulated above, the fittings are not connected to the air system on the robot itself. The air fittings are connected to an off-board compressor between matches if need to blow air through the motors to cool them faster. Again, like he said the NEOs stayed so cool this season we never even bothered using it at all anyway.

If you’re looking for specific part info, the screw that threads into the motor is a “vented set screw” from McMaster, specifically 91979A605. We first put a nylock nut on the vented set screw before threading the vented set screw into the motor so that it can double as a way to mount the motor. Then a standard pneumatic L-fitting with #10-32 thread is mounted on. This allows 2 of the 4 NEO mounting holes to be left open for the air to vent through.

With the CIM motors the other screw that mounted the motor had to be hollow for the air to vent out of; the NEO having 4 mounting holes simplifies this so the other screw mounting the motor can just be a standard screw, in our case a #10-32 flat head screw.

It’s a nice feature to have but with the NEOs it seems to not really be needed.

[wgorgen]

Thanks guys. That is really helpful information. I agree that the NEOs never seemed to get all that warm. But I like the idea that we could add this to CIMs or miniCIMs.

[Cole_Nagata]

I know this thread is a little old, but why did you choose to print the casing out of onyx? could you have printed it out of ABS or PLA?

[Ocfrc126]

The HDT of Onyx is stated at 145C, while ABS is only 100C and PLA’s is only 55-60C. I think that it was to just reduce the chance of deforming after alot of use/heat and it’s much stronger than standard ABS and PLA.

[EveryLittleBit]

Why would any of the printed pieces get that hot? The nylon plate is the one in contact with the motors

[Ocfrc126]

I would think not, but not going to rule it out. They did make intergraded cooling ports

Still, I’m gonna go with the strength or they just wanted to use Markforged as their reasoning.