Banebot 56mm gearbox - double D - RELOADED

[eugenebrooks]

At this point, we have machined carrier plates and axles out of the 4130 steel that we have on hand, using a square broach for the carrier plate. We have hardened at 1600F with an oil quench, and tempered at 800F.

A link to a table showing the strength characteristics of 4130 as a function of the temperature for temper is shown below:

http://www.eng-tips.com/viewthread.c...=116052&page=4

<<<<<
Following is a listing for tempered 4130 steel:

UTS YS

C F MPa psi MPa psi El RA HB

205 400 1627 236 1462 212 10 41 467
315 600 1496 217 1379 200 11 43 435
425 800 1282 186 1193 173 13 49 380
540 1000 1034 150 910 132 17 57 315
650 1200 814 118 703 102 22 64 245
>>>>>



The yield strength at this temper is 173,000 psi and the elongation at break is 13%, not too brittle.

We have assembled and broken in the gear boxes using no load and 8 volt power, running both ways until the gear box "sped up" from the polish established on all the parts. This was about a half hour in each direction. We then inspected for any evidence of metal chips, and installed them in the robot.

We should get a day or two of testing by Monday. After some agressive testing, we will report the results.

The material is from McMaster and you can make these parts easily if you have access to a lathe, a mill, and a heat treating furnace. This approach to the problem is best described as Dr Joe's plan C. If you don't have this equipment, I would suggest that you head to a local machine shop that can heat treat with suitable material and the existing parts to copy with the modified square hole setup.

If I bought material just for this task, I would use 4340 steel and suitably adjusted hardening and tempering temperatures in order to get a little more strength while maintaining toughness.

Well, I would really use maraging steel from http://www.onlinemetals.com if I ordered material just for this, but I won't get into that.

Eugene

[ZZII 527]

Update on the A2 tool steel carrier plates:

After cutting them on the waterjet, finishing the pin holes with a 5/32" drill bit, and filing the double-D to be a close fit, but not a press fit, I tempered the steel today. Here was my process:

1. Heat at 950C (1742F) for 20 min.
2. Air cool (A2 is air-hardening tool steel) until warm enough to touch.
3. Temper immediately at 650C (1202F) for 2 hours. Air cool. Repeat temper for two hours.

The high tempering temperature brought the hardness down to a C40 on one plate and a C39 on the other. (Tensile strength around 182 ksi.) I am hoping that this won't be too brittle.

A2 holds tolerance well in heat treatment, so the press fit pin holes worked fine. The double-D was a bit tight, but after a bit of sanding they seemed to go on easy. One gearbox assembly went smoothly. Putting the two stages in first and then fitting in the shaft later is much easier than messing around with a press fit shaft and carrier. It ran smoothly with the motor, drawing about 2.5A. (Is this normal? Good? Bad?)

The second gearbox is still having alignment issues, binding slightly in particular positions. I think the carrier plate may need some more filing to get it on straight. Maybe a fresh look at it tomorrow morning will do the trick.

So it is indeed possible to make a working HRC40 carrier plate out of A2 tool steel. It's a nice steel, because it can cool slowly, not deforming much, and still harden. McMaster also carries 5/32" (4mm) stock, saving one machining step. But it is somewhat expensive and needs to be heat-treated to get to the right hardness, so it probably isn't the ultimate solution.

The most important thing I learned so far is that a press fit makes assembly (disassembly) and alignment VERY difficult, and so I wouldn't recommend it. If I cut more, I'll cut them exactly 9mmx12mm. If the plate can self-align in the motor, it will run much more smoothly. But your results may vary, of course.

For now, I think I'm done. As long as the second gearbox goes together okay and nothing breaks this weekend when we drive the heck out of it, I think I'll be happy. We are using 12:1, 1 CIM per side, so I don't think I need to make a sqare-hole version in anticipation of needing the harder shafts. But I might cut some just in case..

[sanddrag]

Originally Posted by ZZII 527

It ran smoothly with the motor, drawing about 2.5A. (Is this normal? Good? Bad?)

We draw about 5 amps with two motors, so yeah, I'd say you're good there.

[Jon Jack]

Originally Posted by AcesPease

I would like to add that we used smaller Banebots transmissions last year for Mabuchi and Fisher Price motors and found the quality and service excellent.

Agreed, we used a 36mm, 256:1 gearbox to reduce the speed of a FP motor for ball harvesting last year. Worked well... Then again it was used for ball harvesting (which put very little load on the motor) not lifting an arm.

If a solid solution cannot be found in the next couple of days I would recommend that teams planning on using the 56mm gearboxes to drive their robot should seriously consider the single speed AndyMark (http://andymark.biz) gearboxes. They're the closest thing to last years kit transmissions, in fact I believe they might even be lighter than the old kit transmissions. At $98 a piece these teams should ask themselves if a headache is worth $200 +shipping.

[Tim Arnold]

Originally Posted by jjack

If a solid solution cannot be found in the next couple of days I would recommend that teams planning on using the 56mm gearboxes to drive their robot should seriously consider the single speed AndyMark (http://andymark.biz) gearboxes. They're the closest thing to last years kit transmissions, in fact I believe they might even be lighter than the old kit transmissions. At $98 a piece these teams should ask themselves if a headache is worth $200 +shipping.

I don't have anything to throw in about the actual process here, but for us, and likley many other teams, it is far too late in build to just "drop in" different transmissions. Believe it or not, it would take far more time, effort, and money to modify our existing chassis to use the AMs than to fix the already-mounted BBs. We will be looking into hardening once a final plan is developed, as a mentor fortunately works at a brazing company.

[eugenebrooks]

We got a chance to test our 56 mm banebots transmission solution
today, using 12:1 gearing and two cim motors on each transmission.
As noted in the post earlier in this thread, we made new plates and
axle shafts out of 4130 steel, using a square hole, and hardened in
an oil quench at 1600F followed by a temper at 800F. The 4130 is
what we have on hand for heat treated welded parts, we would have
used 4340 for this application (or maraging steel considering the risk
to the function of the robot) if we had bought the material specifically
for it.

Our test was a half hour of ramming the robot into a old tire positioned
against a wall at full power from a distance of a foot or two, back and
forth. The test was stopped when everyone got bored of it. The
transmissions are doing fine with no apparent extra lash, but we have
not yet pulled them apart for inspection. We will do that when time
permits, there isn't all that much time available in the last two weeks
of the build period as we have our share of other problems to solve.

Have fun,
Eugene

[sanddrag]

Originally Posted by eugenebrooks

We got a chance to test our 56 mm banebots transmission solution
today, using 12:1 gearing and two cim motors on each transmission.
As noted in the post earlier in this thread, we made new plates and
axle shafts out of 4130 steel, using a square hole, and hardened in
an oil quench at 1600F followed by a temper at 800F. The 4130 is
what we have on hand for heat treated welded parts, we would have
used 4340 for this application (or maraging steel considering the risk
to the function of the robot) if we had bought the material specifically
for it.

Our test was a half hour of ramming the robot into a old tire positioned
against a wall at full power from a distance of a foot or two, back and
forth. The test was stopped when everyone got bored of it. The
transmissions are doing fine with no apparent extra lash, but we have
not yet pulled them apart for inspection. We will do that when time
permits, there isn't all that much time available in the last two weeks
of the build period as we have our share of other problems to solve.

Have fun,
Eugene

What wheel size, tread material, and floor/ground?

[eugenebrooks]

Originally Posted by sanddrag

What wheel size, tread material, and floor/ground?

Wedgetop conveyor belting. Wheel diameter was about
5 1/2 inches. The robot was operated on carpet for the test.
Chain drive from the gear box to the wheels was 1:1.

Eugene

[Arkorobotics]

Our team is also using these 56mm gearboxes. Now I have been reading through the forums and I am scared now! I am afraid these gearboxes will fail. We have signs of the gearbox binding or jamming. What is the best and simplest way to prevent this?

I am afraid that it may break like the images people have posted. What can I do to prevent this? From what I soaked up from previous posts is add lubricant, support the CIM, and break them in.

Also I found gear dust, very small amounts of metal inside the gearbox. What should I do?

Now I heard harding this plate may prevent the problem of the double D breaking. The problem is we don't know how to harden the plates, and I doubt we have the time. Also if these gearboxes fail, we don't have the money to replace them. So what should we do?

[DonRotolo]

Originally Posted by Arkorobotics

I am scared now!

Also I found gear dust, very small amounts of metal inside the gearbox.

Now I heard harding this plate may prevent the problem of the double D breaking.

Don't be scared. Those with experience in these kinds of issues are working on a solution. They realize they have two weeks.

Small amounts of metal dust are not unusual, nor cause for panic. If you want, clean out the internals of the gearbox and re-lubricate.

The plates themselves are of a material that cannot be hardened, so again, folks are working on this and hope to have a solution. NOT having a solution might be a serious problem, they understand, so just the existence of this thread should be very reassuring.
===

For those really in a panic , go to a local machine shop and get a quote on making a shaft, 3:1 plate and 4:1 plate all with the square holes, from material that will be hardened to Rockwell C 23 - one set for each transmission you have. Install those and, the current thinking is, you should be good (or at least a LOT better) to go. If the price is reasonable, have spares made for Just In Case.

Don

[ZZII 527]

Originally Posted by Arkorobotics

Our team is also using these 56mm gearboxes. Now I have been reading through the forums and I am scared now! I am afraid these gearboxes will fail. We have signs of the gearbox binding or jamming. What is the best and simplest way to prevent this?

I am afraid that it may break like the images people have posted. What can I do to prevent this? From what I soaked up from previous posts is add lubricant, support the CIM, and break them in.

Also I found gear dust, very small amounts of metal inside the gearbox. What should I do?

Now I heard harding this plate may prevent the problem of the double D breaking. The problem is we don't know how to harden the plates, and I doubt we have the time. Also if these gearboxes fail, we don't have the money to replace them. So what should we do?

All of the things you mentioned (lubricant, breaking in, supporting the CIM) are helpful. I'm guessing you are using just one CIM in each gearbox. The data and evidence so far suggests that if you treat this setup right, it will quite possibly survive without modification. Joe Johnson and BaneBots are working on a solution for making harder carrier plates that doesn't involve the teams doing heat treating themselves.

I can tell you from my own experiences (we are also running 12:1, 1 CIM per gearbox) that if you notice binding and the gearbox gets hot while breaking it in for a half hour, you may want to take it apart an look for tight spots. This will give you a chance to look at the carrier plate too for any signs of wear and add lubricant. There are instructions available from the BaneBots website for how to relieve common problems, such as the motor pinion sticking out too far.

Another fix that seemed to help yesterday was sanding (with fine grit) and/or polishing the round lip on the bearing blocks that fits into the ring gear housing and the flat surface opposite the carrier plate on the bearing block. This seemed to help with alignment issues, and the motor current went down by half an amp or more.

[dtengineering]

Don't panic about the gearboxes... be aware, but don't panic. You are in the same boat as probably close to 1,000 other teams and there is absolutely no way this boat is going to sink.

We have had some experiences with the 56mm gearboxes and have identified two problems, both with assistance from this forum.

1) The "Double-D" bowtie problem, which this thread is primarily concerned about, and which... as previous posts indicate, can apparently be solved by remanufacturing the final stage carrier plate. Given the sheer number of these gearboxes in circulation, you can expect that FIRST and Banebots will be involved in this solution. As for now, you can drive and test your robot. You may (will?) experience increasing backlash as the bowtie forms, but that is a ductile failure that should cause no permanent harm to the transmission. When (if?) a new final stage carrier plate becomes available it will not take long to insert. These transmissions can be rebuilt very quickly and easily with one 2.5mm hex key.

2) The "mounting plate" (aka spacer) problem. There is some variation in the thickness of the mounting plates for the trannies. This is more likely the cause of the metal and grinding that you have experienced. We found that three (of our four) BB 56mm gearboxes experienced this problem. The cause seems to be that the tranny is "too tight" when all the screws are done up, and the spacer that fits on the CIM shaft rubs against the gearbox end of the motor... since the motor appears to be aluminum in this place, and the spacer steel... well, that is probably most of the metal dust. Check to see if the end of the CIM surrounding the motor shaft appears to be wearing away and this will confirm this problem. We found that putting the spacer (that goes between the CIM and the first gear) on the lathe and turning about .010 off one end, then slightly grinding the motor shaft down to match has cleared up this problem and resulted in four nice smooth running gearboxes. In retrospect, we should probably have done this before "breaking in" the CIM... as the breaking in process seems to consist mostly of wearing down the end of the CIM. If you have access to a lathe (or even a vise, a file, and a steady hand) you should be able to solve this problem fairly quickly.

I would like to point out, that these are not solutions or observations original to our team, but rather are based on helpful posts by many other CD posters.

We have also noticed one other thing to look out for if you are using the Banebots encoders... some of the transmissions seem to have a bit of end play in the driveshaft. We have observed this on one of our four trannies in particular. The end play was sufficent that the encoder disk would move in and out enough that it would at times rub against the sensors on the encoder. A serrated steel disc rubbing against plastic optical sensors means it didn't take long before we were working with one less encoder.... we are considering a few solutions for this problem, but have not tried one yet.

Any suggestions CD community?

Thanks to everyone who has been posting their experiences and the teams who are leading the way in testing re-manufactured carrier plates. It has been about a week since the bowtie problem was first noted here, I believe, and it sounds like a solution is well in the works. That is an amazingly fast response, and just one more reason why FIRST is first.

Jason

[Arkorobotics]

Originally Posted by ZZII 527

All of the things you mentioned (lubricant, breaking in, supporting the CIM) are helpful. I'm guessing you are using just one CIM in each gearbox. The data and evidence so far suggests that if you treat this setup right, it will quite possibly survive without modification. Joe Johnson and BaneBots are working on a solution for making harder carrier plates that doesn't involve the teams doing heat treating themselves.

I can tell you from my own experiences (we are also running 12:1, 1 CIM per gearbox) that if you notice binding and the gearbox gets hot while breaking it in for a half hour, you may want to take it apart an look for tight spots. This will give you a chance to look at the carrier plate too for any signs of wear and add lubricant. There are instructions available from the BaneBots website for how to relieve common problems, such as the motor pinion sticking out too far.

Another fix that seemed to help yesterday was sanding (with fine grit) and/or polishing the round lip on the bearing blocks that fits into the ring gear housing and the flat surface opposite the carrier plate on the bearing block. This seemed to help with alignment issues, and the motor current went down by half an amp or more.

About the pinion, sorry, I haven't taken apart the transmission, but what is it? Where is it? Are there any pdf's or files that talk about taking the transmission apart?

[MrForbes]

I know this isn't a full solution, but I think there are some things you can do to make the gearboxes live longer, at least through a regional competition, without spending much money or needing exotic machining abilities.

First, make sure the end play is properly adjusted. There is a spacer that goes on the motor to space the gear into the motor the correct amount, this procedure will get that spacer to be the correct thickness. Support the gearbox housing, with the shaft down, and make sure the gearbox is not resting on the shaft or bearing. Use a depth measuring device such as the pin that sticks out the side of a dial caliper, and measure the depth of the first stage planet carrier plate in the gearbox. The place you need to measure is shown in the photo. Now measure the distance D on the drawing from the motor mounting plate, to the end of the gear. The gearbox measurement should be larger than the motor measurement. Subtract the two measurements, this is the end play on your motor/gearbox. I suggest that a desireable end play would be 0.005 - 0.010 inches (0.13 - 0.25 mm).

The fun part is adjusting the end play. If there is not enough, you could turn the spacer in a lathe or file it down to make it shorter, or you could add a shim between the motor mount plate and the gearbox. Paper would probably be sufficient if it only needs a little bit of shimming, normal paper is about 4 thousandths of an inch thick.

If there is too much end play, then you need to make the spacer thicker. You can't use paper for this task, as things are turning, so you need something that won't wear easily, such as nylon, or steel. You might see if you can find a thin flat steel washer, or make your own using "shim stock", which can be purchased at some old-fashioned automotive parts stores, or places such as MSC or McMaster-Carr. You can also buy "round shims" at McMaster-Carr and probably others. If you have access to a lathe you can make some extra spacers, and play around with machining them to the correct thickness. There are probably other relatively easy solutions I have not thought of, hopefully others will let us all know what they are when the figure them out.

There are some other gotchas on the end play issue, such that you should recheck it after you have run in the motor/gearboxes, and again after they have been in use on the robot for a while. There is some end play on the output end of the gearbox, because of how the bearings are mounted. If your drivetrain design pushes in on the output shaft, then you may have problems with binding or excessive wear, so try to make sure it is not doing that.

The next thing to make the gearboxes live is to understand that the problems with the carrier plate being damaged appear to be due to sudden direction change. Talk to your drivers, and tell them that any sudden acceleration of the robot, especially a sudden direction change, is essentially abuse of the robot. The gearboxes will live MUCH longer if when you need to change direction, you make sure to come to a complete stop, then gently accelerate in the other direction. Jamming the joystick from one end quickly to the other end of it's travel is an invitation to disaster.

One possible method of ensuring that the motors are started in a controlled way, would be to have the programming team put a limit on how fast the PWM outputs to the drive motor speed controllers increase. If they allow only a small change per unit time, then the gearboxes cannot be damaged by sudden direction reversals, even if the driver gets a bit exuberant (and I do realize how exciting the matches are!). Look into this, hopefully some of the programmers here can come up with some code to control motor acceleration rate.

I think if you take care of your Banebots gearboxes, you should be able to get them to survive the season. If some of you can try out these suggestions and then take the gearbox apart and see how much wear they have, it might put a lot of teams' minds at ease....

[ZZII 527]

Originally Posted by Arkorobotics

About the pinion, sorry, I haven't taken apart the transmission, but what is it? Where is it? Are there any pdf's or files that talk about taking the transmission apart?

http://www.banebots.com/docs/GP-56012-Assembly.pdf

That doesn't show the internals of the gearbox, but the post above by squirrel explains some other internal fixes well.