Dear CD,
Finally got around to creating my own gearbox using the Andy Mark Nano as an example. I had access to a water jet to cut the outside plates. The rest are parts from Andy Mark. Final ratio is (505050)/(141414) = 45.6.
The outside dimensions are nominally 2" by 4" by 8". The two, intermediate gear shafts are 3/8 hex supported by 3/8 hex bearings on both sides. The single output shaft is 1/2 hex output.
I could have used this during the FRC2013 season. Now that I have designed one gearbox, the next one doesn’t seem so daunting.
For the idler gears I would recommend using 3/8" round bearings. They are cheaper and if you machine the hex shafts correctly you will not need the BHCS to hold them in place.
Looks like your CIM motor is not mounted flush to the Transmission plate, this will cause a lot of issues down the road with wear. If you are going to use the CIM to retain the bearing I would put the flange on the inside and mount the CIM motor before inserting the bearing. When we do this we some times make a small shim plate for the bearing so that it will be flush to the plates and hole.
On the Output shaft, if you do not need Hex coming out of both sides. I would also recommend a 1/2" round bearing. Tap the end of the round with 1/4-20 and use a BHCS and a washer to retain the shaft instead of the E-Clip.
If you are planning on being at MARC, I would be more then happy to discuss this further with you in person. I can show you some of the things we do on our custom gear boxes.
I will echo some of the same things Clinton has mentioned.
The CIM motor not mounting flush will cause a lot of issues. I would look into that. This is one of the main reasons for using a thicker sidewall, to keep the bearing flush with the outer surface. A quick solution you could do to keep this gearbox/prototype is to create a shim/spacer for the CIM mating surface to move the CIM off the sidewall and bearing. This would require very minimal changes to what has already been fabricated.
The cost of bearings adds up over time, surprisingly. A round bearing vs hex bearing is around 1/2 price. That is a savings of ~$2/bearing; and only requires minimal effort from what is depicted.
How is this being mounted to the robot? Are you using the 4 bolts that hold the transmission together? I would loo into ways for keeping the assembly as a whole unit and pick up a few other points. This makes it easier for hot swapping and changing in the heat of the battle (based on my experiences anyway).
If you want to really look at some weight savings, eliminate all the extra shaft lengths that are not needed. Replace the screws retaining the bearings with an aluminum rivet or two.
Where is the encoder mounting? At the the end of the long turned down hex shaft? That is a lot of added weight for an encoder. Could you integrate the encoder in another spot? Maybe off another shaft?
Hope this helps. This is a good start and has probably taught you many valuable lessons. Keep iterating!
Dear CD,
I appreciate Clinton and Tim-tim comments. Here’s some modifications based on your input. 1) Added rubber shim to better align the CIM gear, 2) Added holes to allow addition of encoder mounting plate, 3) Added numerous holes in side plates to facilitate potential mounting locations. 4) One extra long shaft plate is a piece purchased from Andy Mark that can be shortened by the end user.
I also added 60T Hex Gear (VEX) as last stage rather than 50T. I was able to move the entire assembly 1/4 inch and still have everything remain in the 4 by 8 size limit. The final ratio is now (505060)/(141414) = 54.7 vs 45.6 (50^3)/(14^3)
Comments: 1) Even though the hex bearings are double the cost ($4 vs $2), my design goal was to minimize machining operations associated with the hex shafts (not designed for production). This way, someone can build the entire gearbox without machining the hex shafts, with just the 2 water jet cut side plates and purchase of gears and hex shafts from Andy Mark and/or VEX Robotics in their home garage where mine was built.
You can keep this design constraint and still use cheaper round bearings if you use this AndyMark part for your idler shafts. There’s only enough room for two gears on that shaft, so for large reductions you’ll want to increase the spacing between reductions in order to ensure there’s enough room. There’s a good post Adam Heard made about this style of custom gearbox I’ll try and find.
Using this part and the same spacing as the Toughbox you can make a variety of large yet compact custom gearboxes, as long as the gears between stages don’t intersect.
I think this is an excellent gear box and for the future I would suggest getting comfortable with 1) having a design that is ready to mount encoders 2) start looking into designs incorporating multiple motors around one gear 3) mounting motors already attached to a planitary gearbox to reduce the amount of reductions needed in the custom one. (4 chain and belt drive and what you can do with the reductions they can provide.
A note on how you’re retaining the bearing. It appears like you have the bearing flanges on the inside along with having bolts holding them in. Using bolts is common practice if the bearing flange is on the outside but with it on the inside this is unnecessary. The bearings can be retained simply by sandwiching the gears and spacers between the plates and bearings on either side.
Joey,
Good catch! When the side plates were initially assembled the 1/4-20 BHCS helped to hold the bearings in place during assembly, but, once assembled are no longer necessary. I’ll remove them when I get a chance.
You can keep this design constraint and still use cheaper round bearings if you use this AndyMark part for your idler shafts.
Thanks Chris for your comment. That particular part is too small because the gearbox is nominally 2 inches wide. Andy Mark does also sell the SS Cluster Shaft (am-0536) which is close with a HEX length of 1.895 inches and the overall length of 2.295 inches, but a little short (0.395/2) on bearing support on both ends. The overall length should be perhaps a half inch longer to have full bearing support on both sides. I’ll see if I can get a couple pieces made and re-evaluate the design.
