Last night one of our Kit 56mm gearbox failed. Here are the answers to the questions that Dr. Joe asked for.
Approximate weight: 100-120 pounds. (we added weight to the front of the robot to make 100 -120 lbs.) See picture.
7” Diameter Aluminum wheels 1" wide with urethane on the outside with needle bearings on steel hardened shafts. Six wheel drive with the middle wheels down approximately 1/8”. See picture.
30 tooth sprocket on transmission shaft to 40 tooth sprocket on wheels with 25 chain.
2 CIMS adapter.
12:1
supporting outside shaft, gears are approx. ½” from gearbox. See picture.
Do not know.
One gearbox was not able to turn the wheels. We did not tear the 2nd transmission apart.
See pictures below. We removed the weight from the robot to take the transmission off the other side.
We were mainly driving forward and then slamming it in reverse. Doing some turning. We do not have an exact count of how many times we did this. We are guessing 75 – 100 times.
This was a practice frame that we put together to see how the transmissions would work. We ran the robot for a day about a week after after kick-off and then off and on since then.
I checked the backlash on the other transmission. It is approximately 25-30 degrees. I marked one of the teeth and rotated the gearbox until it stopped. It moved about 2-1/2 teeth of a 30 tooth sprocket. 2-1/2 teeth on a 30 tooth sprocket works out to about 30 degrees.
Alright, I got a chance to drive ours. Stock original plates. It has been driven maybe a total of 10-15 minutes. right now it weighs about 55-60 lbs including battery. 6wd, dual CIM adapter kit thing, another bearing supporting the shaft, blue roughtop tread ~0.75" wide. Drives great. about 10 degrees of backlash is apparent right now. I’d estimate perhaps 150-200 normal game play style direction reversals. Will update when more testing is done.
Our team had done some basic testing for about 10 minutes with a low load (50lbs max) and we took apart the gearboxes to take a look at the plates, when we noticed some slight bowtieing on the DD holes. We’ve decided to not risk further damaging the plates, and have replaced the parts with significantly harder metal that we’ve heat treated as well (having a machine shop next door is so helpful) which hopefully should eliminate the bowtieing effect.
Okay, I think we have confirmation that the plan we have in place to address this issue is sufficient.
Special thanks to Alan Fallone & the great folks from Team 1279. These guys took time away from their own robot to work this out for the rest of us. AND they are not even using the KOP transmissions. This is really what Woodie means by GP. Three Cheers and a mighty Hazzah.
Recall that the test buck was a 115lbs 6WD chassis with wheels covered with roughtop conveyor belt. The test set up was a 12:1 gearbox with 2-CIMs. After 500 hard cycles (each cycle was command full forward, get to up to full speed forward, command full reverse, get up to full speed reverse, command full foward again and repeat):
With the shipped carriers, the DD in the carrier was noticeably bowtied (but still far from a total failure).
With the new carriers, this is not an issue.
See the attached great pictures from Alan Fallone.
Like all solutions done in hurry, we had to take some risks. In general this solution has turned out well.
One fly in the ointment is that the carrier pins had loosened a bit by the end of the test. This is not a tragedy. The gearboxes still functioned just fine with the pins less than tight at the end of the test.
Alan reports “…the transmissions ran fine. The pins are a tight fit now that they have cooled off, I probably got the transmissions hotter this time, running the tests in one night with less time to cool off.”
If we had more time to develop the press fit tolerances we could have addressed this but we did not so we are limited in our response. One thing Banebots is doing is pressing in the pins a bit deeper than we did on the first samples we tested. This should help keep the pins in place, but even if it doesn’t the gearboxes still work fine.
I really want to emphasize to teams that this is not a sky-is-falling-in moment. Two important points:
Loose pins do not have a dire outcome with respect to the performance of the gearbox.
This may not even be an issue because this is a very abusive test (more abusive than most teams see in a season) AND Banebots has taken steps to improve the preformance over the tested carriers.
Bottom line: This solution seems to successfully address the DD joint issue. The pins loosening in the carrier is a condition that may occur after a lot of hard driving that doesn’t have any serious consequenses other than that teams have more parts to loose while working in the pits.
Thanks to all who worked so hard on this issue. There are too many thank everyone by name but I have to give props to Ed & Rick at Banebots. These two have gone above and beyond the call of duty in my opinion.
Thanks for the smaller pictures. I get in a hurry and forget about the dialup folks.
As to the pins, yes, they are shorter. This too is a consequence of moving faster than we’d like. Many sources can provide a handful of 4mm pins. It is quite another thing to get 25,000 in a few days. Banebots wisely chose to go with a DIN standard 4mm pin even though it was slightly shorter so that they could buy from a dozen or so sources.
While I am taking about sources, many of who read the Nothing But Dewalt White paper have heard of Lou Oudin and Capital Tool. Capital has gone under but Lou keeps on keepin’ on. He has started a new company, Sonic EDM. Lou was another hero that came to FIRST’s rescue on this carrier issue. I highly highly recommend Lou and his company to anyone who needs work done (especially Wire EDM work – Lou can get almost anything done but his speciality is Wire EDM).
Joe J.
P.S. Again I am forced to say that a have no financial ties to Lou and/or Sonic EDM other than I use his shop from time to time when I need good work done (both for Delphi and for my company Robotic Amusements). Over the years, he has become a friend, but the main reason I recommend him is that his prices are good and his service is off the charts great.
If a team wanted to make sure the pins fit, would they be better off removing pins from the old carrier plates and installing them in the new ones? are the holes in the plates the proper size for a good press fit with the old pins, or are they a tad too large?
For those of you making your own plates, how are you getting the double D in? That’s not a common broach size I don’t think. And I would think laser or waterjet is not accurate enough. Are you wire EDMing it and then filing down the edges on the shaft?
We made both the plate and the shaft, with a square drive, because
the broach for this was cheaper than the broach for the hex drive.
The doubled broach is too hard to find. We then heat treated the plate
and shaft.
Note, banebots is supplying replacement plates. This could be your
best option at this point.
For anyone wondering how to get to and remove / replace the output carrier plate in a 56 mm Banebots Tranny from the KOP:
At the output side of the transmission, remove the four small hex-socket screws in the ‘chrome’ transmission body.
The black outer bearing plate (and the shaft) now will slide off, but it is a tight fit so wiggle it or pry gently.
You’ll see the five planet gears (brass-colored), and the one Sun gear (silver colored - note the ‘double-D’ projection from it!). Reomve all six gears - it is greasy, so have gloves handy!
The carrier plate it the thing that you just pulled those gears from. It also lifts off easily enough.
I do not know if replacement carrier plates are shipped with the 4 mm pins for the planet gears installed or not, but they can transferred if necessary, just use a small hammer and gently!!
Reassemble. Remember the sun gear has that projection facing the center of the trans. Apply good quality grease liberally. You must line up the projection on the sun gear with the indentation on the plate inside the transmission body, or it won’t go in all the way!
Once the black bearing carrier is fully inserted, it can be turned so the screw holes line up. Remember that the mounting screw holes in the bearing carrier also need to be on the same side as those in the rear plate (by the CIM motor) !!
Install the four screws, tighten but don’t crank too hard!
Note that if you installed a BaneBots encoder kit it does not need to be touched in any way for this procedure. If you want the encoder wires to come out from a different corner of the trans assembly, this is the time to do it.
