pic: Inverted CIM Gearbox V2

So after some feedback, I designed a new gearbox, and its intent was to HOPEFULLY fix the problems from the first one.

Unlike the first design, which had screws, this design has shoulder bolts through idler gears, which are 20T (CIM gears are 12T). Also, bearings were added into the idler gears. For a gear “shaft”.

As many of you might have noticed, the shifting occurs on the upper shaft. This was done by one of the mechanical officers on the team, and I decided to take a stab at it.

The reason for shifting on the upper shaft was due to andymark’s 28T dog gear, and I found this perfect, especially for low gear. But this would require upper shifting.

The main issue with this was mounting the piston, and the shifting mechanism itself. For the piston, it is mounted on a 1/4" aluminum plate so it doesn’t interfere with the bearings on the gearbox, and for the shifting mechanism, I was concerned that off-the-shelf shifters might extend outside the frame perimeter - a big issue, especially with the design (This is true for Vex, but I’m not sure about Andymark).

So I decided to make my own shifter, which is more compact, for it was designed to fit inside the upper shaft.

Problems I have with the gearbox is weight, and I’m going to blame it on the 84T gear in the first stage. This gear is pocketed, but I’m still going over my sweet range, which is around 7-8 lbs. Also, I’m a bit concerned about the design of the shifter itself (if anyone can take a look at that, that will be great). In addition, I didn’t bother to change the material to delrin or polycarbonate. I might as well accept this weight, for as someone said on the discussion thread for the first version, there will most likely be other things on the robot that can use a weight reduction.

Weight: 8.37lbs
Low gear: 6.49fps
High gear: 16.68fps

NOTE: These values were calculated, given the drivetrain wheels are 4.2in and have a coefficient of 1.3

And as always, message me for the STEP file if you want to take a look at it (at some point, I’m going to make a public GrabCAD folder for easier access).

hmmm. That looks like a nice gear box compared to some of the others I’ve seen on here. Where can I find the CAD files? Is there a repository or something for FIRST gearboxes?

I have a grabCAD for pretty much everything I do, but I’m not yet ready to open it just yet for various reasons.

But if you can send me your email address to me via PM, I will be happy to send you the STEP file.

You can make a “Partner Space” in your grabcad project to share with others. That helps keep the secret projects secret, but makes it easy to share what you want.

I see what you did with the low speed :rolleyes:

On to the actual design. It looks like in this design you can actually get rid of the idlers and still have enough room. there seems to be ~1 inch of gap between you bigger dog gear and the body of the CIM. getting rid of that idler gear would be great.
Second, we use FESTO for pneumatics, as they generously give them to us free of charge. So see if you can use a festo piston instead of the one you are currently using.
Overall looks good though. the shifter sticks out of the frame, but it should be well protected by the CIMS around it.

Ill take a closer look at it in person later.

Perhaps in the next version, if I ever do make a third version, I’ll change to a FESTO piston. I’m more use to Bimba’s pistons because I’ve been working with them longer. As for the first stage with the idlers, I think if I want to cut out a significant amount of weight, I will look into belts as an option.

EDIT: The reason why I use idlers is because I find that there is really no possible way to not have it without the CIMs touching the gears on the upper shaft between the plates, if you were using gears.

In reply to @asid61: I just played around with the Partner Space in GrabCAD, and I made one, which contains the first and second versions of the inverted gearbox, and unfortunately a large-sized parts folder. The link to it can be found here (I think I’ll upload pack-and-go zip files and STEP files later so one doesn’t have to have the entire parts folder): https://workbench.grabcad.com/workbench/projects/gcOhfXOehYETdalk4HNJ_3A0HceTxrt6Dv4ytxedJppYkO#/space/gcy-RpuR7fueMhPxpys8KUh7yF1NwcTEWAlStxFm6c3ypj

Definitely possible. Using bigger CIM pinions help, AM has 15T and 17T pinions.

I remember that I was making a sketch of a gearbox 1 or 2 weeks ago, using 17T and 84T for the first stage, and it did work.

After plugging in values into JVN, I’m getting 18.91fps for high gear and 7.88fps for low gear. I think the low gear should be around the 6-range if the robot were to last longer in a pushing match. Also, those gears are coming out to be pretty large (there’s a 84T, a 50T, and a 70T). At this point, I would consider keeping the weight at a minimum and look for another solution.

What if there was something like a large idler gear (singular) above the 84T gear, and the CIM gears go to that idler gear, and instead of shoulder bolts, I can have that gear be mounted by a hex shaft secured on both sides with some sort of retaining ring?

The front plate following the curves of the cims isn’t beneficial. Could take those out and get some more weight savings. Does the output shaft form a live axle in the drive train or is that for sprockets?

I’ve heard that having the plate curve with the cims helps support them.

The weight you’re saving is so marginal (probably less than 0.2 pounds between the two gearboxes) that you might as well shape the plate like that and support the CIMs, even though the gearbox would still be functional without the support. If you need weight that badly, you’d probably be better off lightening all of the gears first.

What about all the gear boxes that run completely cantilevered CIMs?

The OD of the cims isn’t super precise anyway, so the plate either takes some iteration, over constrains and deflects the cims, or has a big air gap and would only contact during huge deflections.

Well we are trying to help teach OP that motors don’t need that that support and they are designed to be cantilevered.

Furthermore the most time efficient way to loose weight from a design is not making feature you don’t need in the first place. Plus this feature is just adding needless time to make the part.