Re: Banebot Transmission Issues
 

While we're on the topic, I'll write out our notes about the big planetaries (included in the KOP) and the dual motor adapters. The "step key" included with the dual motor kit was about 1/16" too long. Also, While it does keep them on, I'm not fond of the pointed set screws in the shaft collars. Also, the screws that hold the planetary body onto the backplate seem weak We broke one. The gears are alright. One of them had a slight bind and is pretty noisy. Not sure where the noise is coming from. When we first ran them, we drew 7.5 amps (at 12V) between two motors. After 15 minutes or so, that was down to 5 amps (for the both).

Use blue loctite on everything.

Re: Banebot Transmission Issues
 

Whoa, let me get a grip on what you're saying here Joe.

You're saying that the FIRST Supplied Kit gearbox, the 56mm, small CIM, primary drive 12:1 ratio gearbox is barely adequate (about 2x factor of safety) for a single CIM install under cyclic loading. Yet, FIRST and Banebots offer a 2x CIM adapter for use in these exact transmissions? Doesn't this 2x CIM adapter effectively reduce my factor of safety on that double-D to 1x?
(I won't even get into the gear change option that brings this gearbox to a 16:1 ratio...)

Is this a good thing? Am I supposed to feel confident in the fact that FIRST and Banebots have given us a product to hang ourselves with? And, to discover this at week 3 into our build? I'm a little disturbed here!

I can understand doing my own Due Diligence on a COTS part that I purchased for my robot. But I should't expect to have to do the same to a FIRST supplied and mechanically approved part(s.) Correctly me if I'm wrong here, Joe. (Please...)

BEN

Re: Banebot Transmission Issues
 

Is there any easy way to avoid this problem?

Such as supporting the drive shaft on both sides? I already knew that sideloads are never good for planetary gearboxes, but could they really ruin them like this?

Re: Banebot Transmission Issues
 

Ben and Joe,

You're both making me nervous. We are in the same boat. Having trusted FIRST to do some of the "due diligence" for us, we are using the 2 into 1 adaptor as well.
I do not have any experience with the set up yet as we have not powered up the drive train yet. Although, Ben, check the run out of the 2 spur gears that mesh the CIMs into the BB 'top hat'. One of the sets out of three we bought is out by a lot, probably 5-10 degrees. Too much for my taste, BB has an email to answer.

Re: Banebot Transmission Issues
 

While the possible weakness in the 56mm gearboxes leaves me with some concern, the "don't stall the little gearboxes" warning was made last year.

http://www.chiefdelphi.com/forums/sh...light=banebots

In the right application, with proper design the small BB transmissions work great. We used a 5:1 reduction running off an FP on our shooter last year with no problems... and have a couple 256:1 ones sitting on the shelf waiting for a job to come along for them. Since they will run off FP's, which let out the "magic smoke" pretty quickly as they approach stall, we try to keep the FP's down to 10A or less and within the torque limits.

I am sure that as I type, however, there are people working on clarifying any problems with the 56mm gear box and developing a solution.

I'll keep my fingers crossed,

Jason

Re: Banebot Transmission Issues
 

Many folks imagine that FIRST has more resources that they really have. As a practical matter, there is nobody in Manchester with team experience. There are great people working for FIRST, but they just don't have the team experience.

The 56mm Banebot transmission was designed specifically for FIRST. Banebots asked me to help them with the requirements, which I gladly did. At this point, I wish I would have spent more time analizing the gearbox. It was not until after the kickoff (when I saw the picture of the 4:1 gearstage from Banebots' website that allows the 56mm gearbox to convert to 16:1) that the double D joint came to my attention. I should have noticed it sooner, but it just didn't come to my mind.

FIRST insisted on the 2 motor solution. I was not a big fan of it, but I did not oppose it either. After seeing the double D joint, I was very worried.

Seeing a picture of a FIRST test mule made me think that perhaps FIRST did not do enough testing. The picture has been removed. For those who did not see it, it had a long unsupported shaft that simply could not have stalled 2 CIM motors.

Initial calculations did not provide comforting answers. I have spent 3 weeks getting the data I need to make a reasonable recommendation. Yelling the sky is falling when it is only an acorn is just as damaging to FIRST as failing kit gearboxes.

What I still don't know is what is the dynamic load that a single motor CIM gearbox with 12:1 ratio. We really need someone to build up a 120lbs test mule and just cycle it Full forward, full reverse for about 500 cycles and then open the gearboxes up to see how much the double d had turned into a bow tie*. This is a big question. Is this just a problem for the 2 motor CIM adaptor or is it a problem that is more widespread. At this point, we don't know.

The fix is not that hard, but it is not free either.

As to whether we should have caught this earlier, yes, I believe we should have. But, mistakes are part of the human condition. Me, FIRST, Banebots, the subcontractor that builds the parts, there is plenty of blame to go around.

Right now, we need to understand the extent of the problem and come up with the best solution possible.

Joe J.

* see the attached photo for an idea of what happens to the joint. This was done with 100 cycles of +/-40in-lbs torque (approx. 2 CIMs). During the test, it was clear that each cycle is opening up the joint. I started with about 180 degrees of backlash at the input to the transmission. By the 100th cycle, the doulbe d had opened to the bowtie shown in the photo and the backlash was over 270 degrees.

Re: Banebot Transmission Issues
 

To help those of us who have not disassembled one of these gearboxes understand the situation, am I correct in assuming the output shaft has two flats on it, which fit into the Double D hole?

Re: Banebot Transmission Issues
 

Yes, see photo below.

The shaft is essentially 2X as hard as the carrier. I think the easiest fix will be to get harder carriers made.

Joe J.

Re: Banebot Transmission Issues
 

Joe,

What are the odds of getting a cad drawing of the carrier(s) to facilitate getting new ones EDMed or otherwise fabricated? Seems to me that enough teams are facing this issue that going in together on an order from eMachineshop or an EDM house would be fairly economical.

Re: Banebot Transmission Issues
 

Well, I've been reading through this thread, and I'm about to throw up.

We bought the 2-CIM adaptor for the 56mm planetaries.
We bought the 16:1 conversion kits.

Somewhere, some mechanical engineer reading this post is laughing his head off.

We've basically done what everyone has been saying NOT to do.

What still bothers me is this:
Why would FIRST supply us with products that clearly were not properly engineered? Or, look at it this way. Maybe these things are substantial. FIRST usually does things the right way, wouldn't they have faith in the company that is supplying such a critical component in the KOP?

I...I don't know what to say. I am at a toal loss for words. I want to have faith in the 56mm's, but everyone is basically saying "ehh...notsomuch!"

Re: Banebot Transmission Issues
 

I cannot disagree with you more strongly.

It is not clear at all that we will see failures in the field of this joint.

Looking at the design, my gut told me that perhaps there might be a problem. I have done quite a lot of work to try to get the answer to the questions.

Again, you image that FIRST has a Great Oz working behind the scenes to ensure that there is never a mistake made. Trust me, there is no Great Oz. There are only frail human beings trying their best to get great things in the KOP without breaking the bank.

Now back to the problem at hand. Contrary to my initial gut feeling, my analysis up to this point leads me to believe that if the carriers were hardened to the same spec as the shaft, the gearbox will likely perform without failures, even for the 2 CIM + 16:1 set up. This is a pretty straightforward fix that should not cause an undue hardship on any team.

Again, the big unknow is the dynamic loading we expect. Is there a team availalble that can quickly adapt a chassis as a test buck for me? I want to do 4 tests each with a 120lbs robot, 500 cycles of

1) apply full forward,
2) reach full speed forward,
3) apply full reverse,
4) reach full speed in reverse
5) repeat

Test 1: 12:1 gearbox 1 CIM per side
Test 2: 12:1 gearbox W/2 CIM adaptor
Test 3: 16:1 gearbox 1 CIM per side
Test 4: 16:1 gearbox W/2 CIM adaptor

The motors/gearboxes needed to do these tests will be provided.

Are there any takers? PM me ASAP.

Joe J.

Re: Banebot Transmission Issues
 

For a controlled test I'd think you'd want to put the test article in an enclosed track and let the RC run it back and forth autonomously.

Ramp up/down or jump right to full forward and immediately switch to full reverse? Could also run a variety of both to simulate different driver styles.

We can do that for you if you don't get a closer volunteer team. We have spare kitbots.

Re: Banebot Transmission Issues
 

O.K. I have kept quiet long enough.

The bottom line is that FIRST had a solution for the kit drive train that could also be used for an arm. They used it for two years with the only reported failures being from teams who did not lubricate the gearbox. It was designed and TESTED before it was placed in the kit. A 120 lb. prototype mule was created, beat against a wall, durability tested and shipped to Manchester in August of 2004 so they could do more testing. It was already in the kit and no more work had to be done.

The Banebot solution was a cost down effort by FIRST, plain and simple. This is O.K., but testing needed to be completed.

My question is why wasn't the same rigor a MINIMUM requirement for the BaneBots solution? If this was not tested on an actual robot before delivery to FIRST, then that is a huge oversight ... HUGE!

Now Dr. Joe,

You and I agree about 99.5% on things engineering, but you and I are polar opposites on this one. You made this comment:

C'mon. You get the torque increase as a result of the increased ratio. It is there whether you want it or not. While you can design your arm to not need it it will always have the ability. If your arm gets stuck, wedged, etc. then the motor will output that torque and you will have almost nothing to say about it. Gearboxes must be designed so the motor is the weak link. If you are reducing the FP (or Banebot) motor to get to a reasonable speed, then you better design the gearbox to handle the torque.

If BB offers a 256:1 ratio that is meant for a specific motor, then it better be able to handle the torque. If they offer a conversion kit to add a second CIM motor, then the gearbox better be able to handle the maximum possible torque the CIM motors can output.

If the gearboxes are not designed this way, then we will have a lot of robots that can make a perfect circle with their drive base.

I am now officially worried about teams using the new kit transmission with 2 motors. My calculations do'nt look good ....

-Paul

Re: Banebot Transmission Issues
 

Sorry Mike, although I understand your frustration, there will always be minor setbacks in life. A valuable lesson can be learned from this group working together to arrive at a solution. Remember, until 2 years ago, we didn't even get a frame or transmission in the kit.

Joe, you have a PM. Team 1279 has the transmission pieces available, and a prior year robot that we should be able to use as a test bed.

Re: Banebot Transmission Issues
 

You have some good points. I am not happy that this is where the world ended up. There were supposed to be testing done. In retrospect, it was not enough. There is plenty of room for blame and finger pointing.

Believe me, I realize the damage this does to FIRST.

Me and others are trying to work out a solution.

Joe J.