Re: Banebots 42mm gearbox: Recommendations for use
 

There was a BB failure story in the pits next to us at AZ, the arm on the 1212 robot is attatched directly to the output shaft of a BB 56mm gearbox. And powering the 56mm gearbox is the KOP 36mm gearbox! I thought about that a while, and then looked up the numbers, and that little gearmotor puts out about 6 times as much torque as a CIM (according to my flaky unit conversion calculations). So it's no wonder they were ripping up carrier plates and output shafts in the 56mm when the arm was accidently pushed past the end of it's travel.

You need to be careful with the BB transmissions, they are not for high torque applications.

Re: Banebots 42mm gearbox: Recommendations for use
 

I still think the higher gear ratios can be used, provided their function is speed reduction, not torque multiplication.

The torque that the final stage sees is entirely determined by the load. The FP motor can and will supply any amount of torque necessary to move your arm, whether it has to run near stall in a 64:1 gearbox or near full speed in a 256:1 gearbox. (Another way of saying this: If a 256:1 gearbox fails, switching it out to a 125:1 and/or changing to a weaker motor without changing anything else will not reduce the torque on the final stage and will probably fail even faster because of the additional speed of the arm.)

The teams for which the gearboxes have survived seem to be using them to drive light, well-counterweighted mechanisms through significant external reductions. Use your motors as speed sources and everything will be much happier.

I apologize if I am just restating the obvious.

Re: Banebots 42mm gearbox: Recommendations for use
 

You are not stating the obvious, you are stating a point that is rather hard to see. Thanks!

when we were picking the motor/gearbox for our arm we were looking at speed reduction first. 256:1 and 125:1 gearboxes were the only ones that would do what we needed....and we still required another stage of chain reduction. Then, looking at the torque, we though we were ok as far as driving the arm, but the backdriving problem led us to use a gas spring to counterbalance it. It turns out that this also greatly reduced the torque needed to move the arm, which probably is why the gearbox survives.

Re: Banebots 42mm gearbox: Recommendations for use
 

Well, I haven't taken them apart to look. But, it appears that shaft misalignment may have been the issue. If you look at this picture you can see the second plate which supports the shaft moves up and down to tension. They both do, but they are not connected; We think they appear to be aligned but may be slightly off.

We are in a predicament now though. We already bought the AM gearboxes so we will probably go with those anyway. We have added standoffs to the plates to keep them in unison though.

Re: Banebots 42mm gearbox: Recommendations for use
 

Are you stealing the CIM motor for the AM solution from your drive? And are you worried that the reduction of drive power will lower your ability to drive (keep in mind, lower torque not only means less pushing power, but less ability to turn and accelerate as well)? If you are replacing the CIM in your drive with a Large CIM and/or FP, it might be alright, but otherwise I'd suggest looking for an alternative arm solution that does not sacrifice your existing abilities.
Some alternatives for the 42mm Banebots include:
KoP FP gearbox (it may be a bit big and clunky, but it works better than anyone gives it credit for, I believe it is a 128:1 reduction)
AM Planetary then reduced further by an AM single speed (46.793:1 reduction, can accept two motors)
DeWalt XRP transmission (3-speed; 47:1, 15:1, or 12:1 reductions)

Re: Banebots 42mm gearbox: Recommendations for use
 

Our Drive will be just fine.

We are swapping out one small CIM and adding in a large CIM and a FP in the AM planetary. We just slightly above traction limited now.

Re: Banebots 42mm gearbox: Recommendations for use
 

We ran into a similar situation with our arm, we designed it to work with two FP in the 256:1 gearbox's (assisted by a 30lb gas spring). However, during ship night while testing and tuning PID loops one of the gearbox's broke down on us. We had spares of the last plate made for our first regional, but after our first experience didn't feel comfortable with the gearbox's.

So we decided to totally change motors and install a gas spring strong enough to hold all the weight of our arm (90 pounder). This meant we didn't need the torque of the FP anymore and could get away with using on window motor to power the arm. The unexpected advantages of using the window motor is that it made the movement of our arm much smoother and more controllable, plus the advantage of it using a worm gear so no back-driving.

These gearbox's can be useful for some purposes.. but driving the main joint of a FIRST arm is not one of them.

Re: Banebots 42mm gearbox: Recommendations for use
 

I think I'm finally ready to put this one to bed.

We ran a BaneBots 42mm 256:1 gearbox at the Boston Regional on our main arm joint. It is followed by a 72:10 chain and sprocket reduction to the arm, which is fairly light, but not counterweighted and has no gas spring. We ran with Victor braking on and ran a position controller. We also used the arm to deploy our not-so-light ramp through a separate linkage, a movement during which the motor sees a significant load. The mounting looks like this:

http://www.chiefdelphi.com/forums/at...2&d=1171338765

I had made hardened tool-steel carriers (see my many posts above), but we ran on annealed tool steel, which is actually slightly SOFTER than the stock plates, to see how long it would last. It lasted the entire regional. There is some backlash, maybe 15-20 degrees, but I don't think it is near failure. I will probably change it out for another soft carrier in Atlanta to be safe.

The bottom line is you absolutely can use a 256:1 gearbox to drive an arm, provided you send it through significant additional reduction. If you are using a FP motor, 256:1 in series with anything less than 4:1 is almost too fast for an arm to rotate anyway. (Ours rotates at about 30-40 degrees per second with the 72:10 additional reduction.) I am now convinced that the ratio of the gearbox itself does not matter nearly as much as what comes after it. The FP torque isn't what kills the final stage, it is the load from the arm. The only way to reduce this is to use additional reduction between the gearbox and the arm.

I am happy that we stayed with the BaneBots gearbox instead of jumping ship, because it was incredibly easy to mount compared to the plastic gearbox and the large reduction was exactly what we needed to slow down the FP enough for our arm. I would use it again, stock, in the future, and if they harden their carriers from now on, I would definitely use it.

Re: Banebots 42mm gearbox: Recommendations for use
 

There are two ways to look at this: either reduce the torque into the transmission (use a less powerfull motor) or reduce the torque load at the output of the transmission (further reduction and/or counterbalancing)

The FP torque IS what kills the final stage, when full load is applied.

Anyways, it's great to hear that the thing worked!

Re: Banebots 42mm gearbox: Recommendations for use
 

Yes, I should have worded it more carefully: the motor itself does not define how much torque is seen on the final stage, the load does too. (Especially in the range where the motor acts more like a speed source than a torque source, less than 1/2 the stall torque, which is where most arms reside on the torque-speed curve.) In the end, though, it is the motor which supplies the torque and breaks the gearbox.

Thanks for everyone's help with troubleshooting and posting your own experiences - good and bad. I said at the beginning that FIRST and Chief Delphi is the biggest, most knowledgeable consulting firm I know of and it definitely showed this year. Hopefully, BaneBots will take all of the data and stories and come up with an improved product for next year.

Re: Banebots 42mm gearbox: Recommendations for use
 

We're also using the 42mm, 256:1 FisherPrice/Banebot combo to drive our arm through a a 54:14 (sprocket/chain/sprocket) reduction. Our arm load is approximately 100 in-lb max static load, with tube. We're using dynamic braking, no counterbalancing and closed loop proportional control. We limit maximum PWM signal to stay within the 100 to 160 command range. Our arm is fairly flexible and has some 'whip' to it - I think that may help to limit peak loads.

After a practice tournament, and two regional events (about 48 matches total), I opened the BaneBot for inspection. The internals look like new, with no observable cam out on the carrier plate. We're crossing our fingers that it holds up through the Championship Event. We do have a 16:1 and a 64:1 that I believe we can use for parts if need be. Another option is to go with a Globe motor-to-'lovejoy' coupling-to-shaft w/ dual outboard bearing set up.

Re: Banebots 42mm gearbox: Recommendations for use
 

It's interesting to hear that some of you have had much better success with the 256:1 F/P combo than we did. We were using about a 4:1 reduction from the BaneBots to our arm, but with very little flexibility anywhere in the system. Also, our closed-loop position control, at times, probably put the motor from full speed forward to near full speed reverse when the arm reaches position. Basically, we were putting extreme shock loads on the B/B. We switched to a home made gear set using a big CIM for our arm. The gear stuff is fine, but we ran into structural problems with the arm itself (now fixed) due to the shock loads. All of this gives me a real appreciation for the tricks they use to soften the shock load in car and motorcycle drive trains to keep from breaking things.

Re: Banebots 42mm gearbox: Recommendations for use
 

Just for reference, I'll relate our (on-going) experience with the BB trannies. We're using the FP 1:256 combo to lift our ramps. The peak load at the shaft runs about 250 in-lbs. For those at GLR, our ramps weren't working at all on Friday. They weren't put to use in the first two rounds, in the third round, they failed to work. After disassembly, one carrier was rounded, the other bowtied. Both binding because of it, both clearly showing the shaft was only half engaged. Our quick fix was to have the shaft TIG welded to the carrier. This actually worked surprisingly well, but the weld bead and alignment issues are too much of a concern to run with it permanently. For LSR, we've hardened* 2 more carriers and will machine the ring gear to improve engagement. All that said, I remain convinced that the actual lifting of the robots had less to do with the failure of the plates. My hypothesis is that, instead, the plates were initially damaged during systems integration when the lifts were being tested and were stalled several times. We'll know better after a full regional with the hardened plates and no stalling of the motors.

* This was a 1450*F soak for 30 min, a quench in warm brine, then an immediate temper at 900*F for 2 hours. I was working on the assumption that this was W1 as suggested by ZZII. Yes that's a really hot temper. The chart I found suggested that would give me RC 35-40. I'd really rather have another bowtie than a fracture. I'll be running them by a Rockwell tester on Monday so I'll update with what I've actually managed to do to them.

Re: Banebots 42mm gearbox: Recommendations for use
 

Once we got to a proper design for the gearbox on our arm, we have had no problem at all with our 64:1 banebots gearboxes driving it. We are running a FP motor into this gearbox because the FP motor has better fans to drive air cooling and internal thermal protection that prevents motor burnout.

We are driving our arm with the 64:1 banebots/FP setup through a 40:1 worm gear from Boston Gear, and are using latex tubing to cancel the torque due to gravity on the arm. The result is a very smoothly functioning arm that can be operated in both a manual mode with switches on the motor power and with an on-off feedback setup (not PID) using pots to control position. The BB gearbox drives the worm through a shaft with a hardened pin and the torque is sufficient to snap the pin if the arm gets tangled up in the rack and the driver team miss-handles the power to the arm. This is the only thing that we need to address for future competitons. The worm gear setup has the advantage that no power is applied to the motor in order to hold the arm in a static position, so the motor stays cool. I am not sure that I would want to be running a 256:1 bane bots gearbox and a 10:1 worm setup...

We had no end of trouble before switched to a setup that provided for cancellation of the torque produced by gravity on the arm. If you design your arm drive with a worm gear to get lockup with the power off, and you cancell gravity with latex tubing, you will have no trouble at all with your gearbox. Run current through the motor at stall, or hammer the gearbox with oscillations in the arm and the gear box will die quickly.

We are using the 56mm gearbox, with a 12:1 ratio, and two CIM motors and the two motor adapter for our main drive. We did make our own plate and shaft out of properly heat treated 4130 (with a square drive), and we did carefully check the gear box for any interference problems before breaking it in in both direcitons on the bench. We then tore it down, cleaned it, checked all parts for abnormal wear problmems, and then greased it before putting them back together. These 56mm gear boxes have been through two regionals now without problems, and we are headed to a third regional and the nationals, so I hope that I am not speaking too soon, but we have not had problems with them.

Eugene

Re: Banebots 42mm gearbox: Recommendations for use
 

Update on my hardening results. First, my process:

Soak at 1450*F for 30 min.
Quench in a warm/hot brine. (Hot tap water + rock salt)
Temper at 900*F for 2 hours.

The carriers came out at RC 35. If I were making one more go at it, I'd probably keep the soak and quench the same and drop the temp to 800* - 850*. For the curious, the brine was to even out the quench action out of paranoia over warping or cracking. It was hot due to further cracking paranoia, but mostly to soften the quench somewhat to make sure the carriers weren't too hard. These were the only carriers we had, and I didn't want to botch them.