Our Banebots Experience

Technical Motors
1

We purchased three Banebots 256:1, 42mm planetary gearboxes to control our main arm joint. Two were being used while one was meant for backup.

After seeing several threads on CD about the reliability of the carrier plates we called Banebots to see if we could return them. We were told that as long as we did not exceed 10 N of force on the gearboxes the carrier plates would not distort, and that if they did he would provide a full refund.

After mounting two (along with a gas spring) the carrier plates quickly distorted after regular use at 10 N. The gearboxes experienced no “shocking” due to the gas spring that was assisting.

We called Banebots and they are refusing to provide us with a refund, saying that other teams are running them with loads in excess of 10 N. They obviously have no concept of customer service and were demanding that we sent them pictures of our robot to prove that we were using their gearboxes appropriately.

This is all after spending thousands of dollars at Banebots over the years and I’m happy to say they will never see any of our money again; and I’m hoping not through FIRST either.

http://img260.imageshack.us/img260/8141/cimg2477cf7.th.jpg](http://img260.imageshack.us/my.php?image=cimg2477cf7.jpg)

2

It seems like alot of teams have had customer service issues with BaneBots. We were thinking about buying a gearbox, but decided making our own would be worth it down the line in case the BB blew up or something.

Thanks for the post, I hope it helps some other teams out with their design decisions in seasons to come.

3

My thoughts- Banebot transmissions are not up to snuff to withstand the kind of abuse recieved in many FIRST robotics applications- especially not in the quickly and often poorly designed systems we put on the field. However, Banebots is *not *an engineering consultation company, and they are not responsible to tell you whether their product will work for you, or even fail under a certain load.

I am very sorry if they have told you explicitly one thing, offered a full refund literally, and then refused it. That is very, very bad business. However, they simply cannot refund every team who breaks their product, especially seeing as they cannot guarantee that you *actually *used it totally within the proper bounds. Mcmaster-Carr would be way in the hole right now if that was how it went.

Now, my suggestion to you to retrofit your robot, if you need to, is to use the Dewalt gearbox’s mated to a Fisher-Price. I feel that they will offer a comparable weight, and far better reliability. Search for NBD in the whitepapers. Excellent gearbox IMHO.

4

I LOOOOOOVE BANEBOTS!!!

Jacob

– At least they have a fix for it. Getting the plates in the mail tmrw… I hope they work!!

Picture 046.jpg


Picture 046.jpg

5

While we haven’t experienced any severe distortion of the carrier plate double D hole, we did have one 56mm gearbox bind up so bad it couldn’t be turned by hand with a 5" wheel on it. And this was in a 50 lb robot single CIM configuration on a slippery surface. And the instant releasal of magic smoke by the small banebots motor wasn’t so great either. I like their products and prices, but due to reliability and durability concerns, BaneBots products will not appear on 696’s 2007 competition robot.

6

Let me add that after hearing the experiences here on CD that teams such as 1114 had mid season with these gearbox’s, we did some major redesigning and and re-thinking of our arm design. We designed our arm using two FP to power it, both into 256:1 gearbox’s, 1:4 sprocket ratio after that, then added a 30 pound gas spring. We designed the arm to use 1/5 - 1/6 of the FP’s stall torque, and even added heat-sinks and fans to keep them cooled. During our testing we were only drawing 8 amps between the two FP’s… which is a small fraction of the motors tourque. Lets just say we babied these motors and gearbox’s… and they still failed.

It looks like we will try and move away from them now, using a much larger gas spring and a different motor to power the joint. I had high hopes for these gearbox’s, they fit our needs perfectly… but ended up failing us. But these are the challenges we have to deal with in this competition, I just wish we could have spent more time on the programing challenges then spending ship night on poorly made gearbox’s.

Here is a picture of our arm setup, our three other joints were working really well (turret, extension, and manipulator) with PID controls.

http://img47.imageshack.us/img47/556/img0035ch6.th.jpg](http://img47.imageshack.us/my.php?image=img0035ch6.jpg)

7

Yeah… Take that gearbox apart, and it will be one of two things…

The carrier plate or
The motor shaft.

The one awesome fix we found for binding up gear boxes is to put a washer between the CIM and the mounting plate for the gearbox. Some of our CIM shafts were too long or the gearbox hole was too short, and putting in just regular flat washers fixed 'em. If you need me to explain more, I will be happy to do so.

Jacob

8

They are sending out replacement plates for the drive gearbox’s but I haven’t heard anything about the the multitude of teams that bought out their stock of 42mm gearbox’s for arm applications. These gearbox’s are suffering the exact same design problems. We got their “word” that these gearbox’s would work for our application, and they did not. I am sure FIRST got their “word” as well for the drive gearbox’s, now the only option we have to use these gearbox’s is to spend $100’s to get the last stage machined out of a harder steel ourselves.

9

10 N with what moment arm? Or at what point in the gearbox? I would have guessed that they’d give you a torque specification, rather than a force specification. (Force times moment arm equals torque, which causes torsional stress, which breaks gearboxes.)

Didn’t Joe Johnson report that BaneBots was out over $30 000 (USD) in dealing with this issue. It may not be great customer service from your point of view, but I can imagine that with unexpected short-term expenses like that, they’re pretty exasperated about it as well. I’m not saying that you’re wrong to stop dealing with them—indeed, that’s what disgruntled customers are supposed to do—but it might be worth reconsidering the situation in a few months, after this issue has blown over, and everyone is under a little less pressure to perform (BaneBots and FIRST teams, alike).

My guess is, they want our business, and will know to perform and publish a more detailed engineering analysis for next time, to ease fears that this might reoccur.

10

Yeah… We were going to machine our 56mm plates because we had ours sent to our regional, but we called and were able to order more to be sent here. According to tracking, they will be here tmrw.

As for the 42mm, that really sucks. They should at least provide a CNC file or a CAD drawing at least. If you call, they say that that info is proprietary. The best we have is stuff put up here on CD (thanks guys who did this…)

Jacob

11

Anyone using the 42mm gearbox should carefully read the (now three page) recommendations for use thread (http://www.chiefdelphi.com/forums/showthread.php?t=53010&page=1), in which several people including myself have been evaluating the problem and offering suggestions ranging from software to heat treatment.

The CAD files may be proprietary, but nobody can stop you from measuring the thing. I can save you the effort:

Thickness: 4mm (precision-ground 5/32" stock works well)
OD: 25mm
Shaft Diameter: 10mm
Shaft Flats: 8mm
Pin Hole Radius: 9mm, 4 places, evenly spaced
Pin Size: 3mm, maybe undersized 3mm. Doesn’t matter much, use a #32 drill for a workable press fit.

Also, as I have posted in the recommendations thread, I would be willing to rough cut a (small) batch of the 42mm plates on a waterjet THIS WEEK if there is interest. So far only one person has replied yes. They would be O1 tool steel, similar to the stock plates, NOT hardened and NOT finish-machined (the pin holes would still need to be reamed and the double-D would have to be filed to take out the taper from the waterjet).

They can be hardened and tempered to a C40. I can provide a procedure for that if you have a local shop that can do it. I did this for our 56mm plates and they have not deformed or fractured. We also used a non-hardened (annealed) plate in our 256:1 42mm gearbox for over a week with only minimal deformation, so they can be used for cheap spares if you don’t beat them up. (Until hardened, the tool steel is actually slightly softer than the stock plates.) For what it’s worth, our arm is not counterweighted and we have no gas spring. The only thing protecting the gearbox is a 7.2:1 reduction. And thus far I have yet to see anything even close to failure of either the stock plate or our new soft plates.

BaneBots had to take care of the larger problem, the 56mm gearboxes, so I’m not surprised they didn’t get around to the 42mm non-kit stuff. But it seems like there are only a limited number of teams still using them and if that’s the case, I can produce a small quantity of rough cut 42mm spares (in about an hour). If there is interest, please let me know as soon as possible. I would just ask for payment to cover my materials/machining (waterjet is cheap…$5 per part maybe)/shipping costs.

12

I realize that you can measure it. It just seems like BaneBots should give you the drawing for you to be able to fix their gearbox if they don’t provide the means to do so for themselves.

Jacob

13

Many manufacturers offer planetary gear boxes for industrial motion control. The are beefy and will take years of hard use in manufacturing. The problem with the banes bot planetary is that they compromised to get them in at a low price. You can’t have it both ways. Take what you paid for the 42 Banes and multiply by 5. That will get you a robust planetary. I use allot of power tools and over the years have rebuilt allot of tools. The tools I beat the c##p out of and they survive year after year do not have planetary gears. There are some things to like about parallel shaft spur gear transmissions.

14

OTOH, the automotive transmissions that I put a lot of power thru and they survive year after year have planetary gears, not parallel shaft spur gears. Being planetary or conventional doesn’t really matter as much as using a transmission within it’s design limits.

The BB transmissions that we are using for arms have a LOT of gear reduction, and really do need to be stronger for the use we’re giving them. If you want to use a low cost transmission, then you have to realize that you are using a low cost transmission, and design the robot to be easy on it. BB doesn’t have a torque limit or shock load limit warning on these things…maybe they should.

15

Well, when I was growing up, my mom told me “if you can not say something nice, don’t say anything at all.” As such, I should probably keep my fingers OFF the keyboard, but in spirit of contributing to helping others avoid issues, I will keep typing.

Sorry this is kinda long, didn’t realize it until I had finished typing it :frowning:

I was VERY disappointed with the drivetrain transmission problem. Having read through things, seems like something that should have been caught before hand.

I WAS impressed with BB getting replacement parts to solve the problem and FIRST allowing teams to hold onto the parts after the robot was supposed to ship. This did not help teams that had not left their transmissions in an easily accessible location. After receiving the replacement parts, then tearing down the transmissions to replace the parts, I was surprised to see the distortion that had already begin in ours. We had very little “live” time on the robot, so was somewhat caught off guard. At least we had solid replacement parts.

We also have used the RS-540 motors with transmissions on our robot as well. These ones, I am LESS than pleased with. We started with a pair of 64:1 mounted to a common shaft that is used to raise & lower a set up that very much resembles a fork lift (multi-stage sliders). During testing (pre-ship) the 64:1 motor set up did not seem to provide the needed OOMMFFF to give us what we wanted, even with 2 of them mounted and acting together.

So, after we shipped the robot, we decided to get 2 more 64:1, along with 2 pairs of the 125:1. We decided first thing Thurs. to give the 125:1 a try, since we had seen what the others already provided.

Things STARTED out OK. The elevator mechanism went up and down, albeit, naturally slower (due to the difference in gear ratios [which was NOT unexpected]). We thought we were set and in business, but alas, that was NOT the case.

Towards the end of Friday, after 4 matches, we had a severe mechanical failure in the elevator motor/transmission. The original thought was that we had let the “magic smoke” escape from the motors, thus rendering them useless. We had anticipated this some and purchased some spare motors.

After tearing down the transmissions, I discovered that the motors were just fine. They powered up normally and spun at the appropriate rpm, even under load. So I started looking further and discovered that the planetary gear set suffered the SAME failure as the drivetrain transmissions. The final stage where the shaft meets the plate used the same double-d setup AND was nearly completely rounded through.

Sadly, seems like we are between a rock and a hard place. I do not suspect that we can replace the setup with some other motor/transmission setup (mostly due to weight and configuration). So, our other alternative is to simply buy multiple replacement sets and deal with it when the failures occur.

I am considering sending the failed transmission back to BB, if nothing else for them to do some failure anaylsis for their own benefit.

In the end, I am disappointed with our experience with the “product” that BB has produced. Hopefully, they might be able to improve as a FIRST supplier and make products that geared towards our application (i.e. rapid development with sometimes rushed or not completely analyzed designs).

At this point, if I had it to do again with the knowledge I had now, would probably NOT use the BB products.

16

I will add our experiance with Banebots and their transmissions at the BAE GSR in New Hampshire.

First off, our drivetrain setup.

4 wheel drive, Skid steer, direct drive, 1 CIM motor per wheel, high traction tires, 16:1 upgrade kit. We are a ramp bot and play solid defense.

When we sent in for our 16:1 hardened plates, Banebots sent us 3 (not 4) 12:1 plates. Since we assumed these were the 16:1 (4:1 secondary stage) plates we just figured we’d pick up 1 more at the regional (since they said they would have spares there). Well needless to say, they did not have any stocked at the regional. The spare parts guys called Banebots and their rep brough us 1 16:1 hardened carrier plate (which is what we asked for) in about an hour.

A couple of hours later (still on Thursday) we took apart our transmissions and found out that they had shipped us the wrong type. At this point we had no choice but to go with the unhardened 16:1 transmissions. The only thing we did to ‘protect’ these transmissions was to have a software ramp up / ramp down.

8 compitition rounds, and 7 elimination rounds later we (and our wonderful alliance partners 501 and 175) won the BAE GSR. We played each match running hard defense and pushing those transmissions to the limit and they never failed us. They now have more that 30 degrees of backlash and our 'bot doesn’t really want to drive straight, but they held out.

So, just to back up what Dr. Joe has been saying all along, These transmissions are not bad. They are inexpensive and we (as robotics teams) tend to push these things beyond their tolerances. We will have to replace them before the Championship event, but we knew, going in, that these may not survive.

All, in all, I believe that these transmissions are good as long as you understand their limitations.

JMHO
–Dan L.

17

Perhaps someone can help me out here then (now that it’s after the fact for us). I thought I had done the calculations correctly, but apparently not.

I had a 3.5 foot arm, total weight about 6 pounds. Using the 540 motor into the 64:1 gearbox, we then stepped down 28:10 and 70:10. The few times before shipping that we were able to test the arm it worked fine except for the backdrive. Then on ship day we fried the motor while driving it to the crate.

On Thursday we replaced the motor and swapped the 64:1 with a 125:1 gearbox. We proceded to fry 2 more motors before scrapping motors and switching to pneumatics (which were also abandoned for lack of time and a better strategy).

Also of note–we had a 20 amp breaker on the motor (stall current 42 amps) and in one practice match we completely drained a battery and the motor smoked with only 2 seconds left in the match.

Have I just messed up my calcs and completly abused these poor motors, or is there another issue here?

18

Not sure if I can help, but have a question…

Are you sure that you fried the motors??? We had some problems (you can read in a previous post in this thread) with these motors, thought we had fried the motors, but ended up a transmission problem.

You might want to take the transmissions apart and check the final stage of the planetary gears, where it meets the drive shaft. Also, check the motor when it is disconnected from the transmission.

19

The biggest issue with the 540 motors (from what I’ve heard) is running them to ‘hold’ position because they have poor heat dissapation when not moving (I believe that their cooling is directly related to the shaft moving). As such, when you are powering the motor to hold a position, it heats up very quickly and then lets the magic smoke out. This is compounded even more if you are at stall current.

Example of ‘hold position’: An arm that is extended and would backdrive quickly (due to gravity, etc) if the motor did not have current running through it. This can be checked by putting an ammeter inline while running the arm in ‘hold’ position.

The 20 amp snap-action breakers only trip if current is ABOVE 20 amps for considerable time (the amount of time is dependent on the current since the breaker is a thermal device). They also quickly reset and reapply the current to the device, not giving the motors enough time to cool. Spikes can eliminate part of this problem because they have an integrated fuse, but if that fuse goes you lose all functionality of the device it controls.

20

Suprisingly, the BaneBots seem to be working well for us. We’re using a 36:1 coupled with a FP motor to drive our elevator - it’s been working fine without any problems so far.