3-motor transmissions

[Andrew Lawrence]

Based on a topic brought up in the "Why not shift" thread (http://bit.ly/UgKdrg)

With the new Vex 3-motor, single-speed gearbox, something I've been looking into is why certain teams have or have not used a 3 motor transmission, and the logistics of such.

Here are some of my main concerns:

1) Have you used a 3 motor transmission in the past before?
If no, why not?
If yes, why did you want one? Do you think it was worth the extra work? Did you have any problems? Was it a shifting transmission or a single speed?

2) What kind of motor would you use in the 3rd motor slot? I've heard a lot about teams ditching their 3rd motors due to burning them out too often. How is this caused, and what precautions can be taken to prevent it?

3) What kind of wheels did you use with the 3-motor transmission? Did they work into a strategy used with the 3-motor transmission? Had you had a 2 motor transmission would you have used different wheels? When in the design process did you decide to use a 3-motor transmission?

4) Last but not least, what other advice would you have for a person looking into using a 3-motor transmission?


Thanks!

[ttldomination]

1) Have you used a 3 motor transmission in the past before?
If no, why not?

I'll add this one because, in short, this is the only experience I've had.

The short answer to this simply the fact that there wasn't a resource available that provided us with this capabilities.

One of my mentors always believed in using everything that gives us power in the kit. If we didn't have a PDV from AM, I'd pick up one of the VEX gearboxes and cross my fingers that we'll get a chance to slap six motors onto the DT.

- Sunny G.

[sanddrag]

3 motor shifting in 2005 with 696. Didn't really need nor using. Not a smart decision.

3 motor shifting in 2008 Overdrive with 968. Worked fantastic, but we ended up changing the robot design, and borrowing two motors off the drive. Robot accelerated noticeably slower without them. Geared for around 21 fps top speed that year.

3 motor single speed in 2012 with 696. Worked great. Geared for pretty good speed, gave extra torque to get over the bump, push robots up bridge, etc. Never a problem.

[dcarr]

3 motor single speed in 2012 with 696. Worked great. Geared for pretty good speed, gave extra torque to get over the bump, push robots up bridge, etc. Never a problem.

Your 2012 bot certainly looked good moving around on the field. What third motor did you use?

Edit: nevermind :) http://www.team696.org/history/index.php?qhr

[Billfred]

2815 (and 1398) used a six-motor shifter in 2010; our design had the motors available, so why not?

We used two CIMs and an F-P with the AM Planetary on each side. No issues with either. AM Plactions were the wheels, with roughtop.

In hindsight, it was overkill for our one-zone strategy...but it also didn't harm things either.

[MichaelBick]

The three teams that I know of that used 3 motor gearboxes were 33 2012, 973 2011, and 254 2012(I'm sure there are a lot more, these are just the ones I know of off the top of my head).

1) I know 973 ran into problems with their 3rd motor because it was a 775(case short problem). I've heard that if they were to do it again they would use a 550 instead. Not quite sure if 254 or 33 ran into problems. I know 254 and 973 ran theirs off their WCD shifting gearbox. IIRC 33 ran theirs off an AM shifter. One of the things that I noticed was that 254 was able to gear higher this year because they used a 3rd motor. According to their website, they geared over 20 FPS with a full weight robot, something that would have been unable to do otherwise(very slow acceleration).

2)Never heard of burning out motors. I'm pretty sure 254 and 33 ran a 550. I know 973 used a 775 later and regretted it after due to the case short problem.

3) 254 ran their normal wheels. Not sure about 33. 973 ran wider (2" wheels), though they ran these same wheels once they took the extra motor off. I know it allowed them to push most people, even with only 4 motors(they pushed 469). I don't know if they used 2" wide wheels specifically because they were running 3 motors per gearbox.

That's a brief overview of what I know about 3 motor gearboxes. I've never actually made/run a 3 motor gearbox, so take this all with a grain of salt.

[Ravage457]

Does anyone have an Autocad drawing or pictures of this type of setup? Im curious to see it

[Gregor]

Does anyone have an Autocad drawing or pictures of this type of setup? Im curious to see it

Check out team 254's website.

[ttldomination]

Check out team 254's website.

I think 254 rebuilt their website. Their image gallery has been unavailable for a while.

- Sunny G.

[Ravage457]

I think 254 rebuilt their website. Their image gallery has been unavailable for a while.

- Sunny G.

Actually it back up, just saw it right now, thanks gregor

[MichaelBick]

254 has rebuilt their website, so their gallery is empty. However they do have renders of their 2012 gearbox posted. In addition you can look through Patrick Fairbank's 254 Build album.

[Nuttyman54]

The reason why not is that it tends to be overkill in a lot of cases, and the motors that add a lot as a 3rd motor (BB775's, 550's and FP's) tend to be your best motors to use elsewhere. The DT motors tend to be abused a lot too, and that can have a detrimental effect on air cooled motors like the banebots and Fischer-Prices if they're not designed in right.

There are two main reasons you would want to do it:

1) You want to be traction limited but have a high top speed (>13 ft/s) with decent acceleration and not require a shifting gearbox

2) You have a shifting gearbox and want a really high top speed without an acceleration penalty (>17 ft/s).

Both of those speeds are rough rules of thumb based on a 150lb robot (inc. batteries and bumpers) running wedgetop tread. That's just how the math tends to work out. The main advantage you get from the extra motor is the acceleration and torque increase at higher speeds. It really has no effect on the low speed performance, provided you're still traction limited*.

Personally, I don't see many games where it's necessary, again going back to the motors that are used. They need to be fairly high power to have any noticeable improvement (remember, 2 CIMS is a pretty beastly drivetrain already). I would argue that for the vast majority of teams, it's more worthwhile to go 6 motor driveline to get rid of shifting than to get that extra jump from 17 ft/s to 20 ft/s, but each team has to make that call. If the majority of your robot mechanisms are going to be pneumatic, then it's less of a sacrifice.

Don't forget to think about your robot from an entire system standpoint. While it might be great to zip around the field at 20 ft/s, if your arm take 3 seconds to score an object because all you had left were window motors, you might actually be taking longer to score than if you'd left your robot at say 15 ft/s, but take only 0.5 second to score.

254 has used 3-motor transmissions several times, although very few have been successful. I recall their 2005 (and maybe 2006?) drivetrain had FP's as the 3rd motor, and they burned several of them out. In 2012 they intended to have three motors (BB775's IIRC), but ended up reallocating them to other robot functions, and changed their gearing to a lower top speed to account for this.

*I would never, EVER recommend designing a drivetrain that isn't traction limited in it's lowest (or only) gear. That's a great way to burn out motors and blow breakers when you run into robots and walls. 3 motors will NOT make you stronger in a pushing match.

[roystur44]

A 3 motor 3 position shifting transmission could be used for a number of things. The third motor could be shifted to a winch or roller pulley for example(power take off) and still have your main drive working. But if you needed the extra turbo power the transmission could shift the third motor to the drive train gear. Couple that with a electrically actuated disc brake and a over current protection sensor and we have something pretty darn cool.

I hope this years game is a racing game with full size robots climbing poles at end game aka minibot style

[CalTran]

The third motor could be shifted to a winch or roller pulley for example(power take off) and still have your main drive working.

Is there anywhere one can find some in depth information about how an FRC power take off would work? I know there are teams who have used them, IE 1114 in 2010, but I've never been able to find a paper or an explanation or a CAD of how they work.

[sanddrag]

For a PTO, look at 254/968 in 2010. It was basically another shifting setup. Another piston and another dog and dog gear.

The first PTO I saw in FIRST was on Team 60 in 2004. Amazing mechanism they had.

[Nuttyman54]

Is there anywhere one can find some in depth information about how an FRC power take off would work? I know there are teams who have used them, IE 1114 in 2010, but I've never been able to find a paper or an explanation or a CAD of how they work.

If 254's 2010 robot had it, you can find good pictures here:
http://www.chiefdelphi.com/media/photos/35993
http://www.chiefdelphi.com/media/photos/35997

If their build log is public somewhere from 2010, that might have some good info on their designs, but they might not be accessible after they rearranged their website.

[CalTran]

For a PTO, look at 254/968 in 2010. It was basically another shifting setup. Another piston and another dog and dog gear.

The first PTO I saw in FIRST was on Team 60 in 2004. Amazing mechanism they had.

Funny, I'm wearing a Cheesy Poof sweatshirt right now...Would a PTO take more customization of a gearbox, as I imagine you would need to shift to a different axle instead rather than the axle the wheels are on?

[MichaelBick]

Is there anywhere one can find some in depth information about how an FRC power take off would work? I know there are teams who have used them, IE 1114 in 2010, but I've never been able to find a paper or an explanation or a CAD of how they work.

254/968 and 973 also had them off the top of my mind. 973 has their CAD posted so check that out if you want to see more in depth of how it works. Basically, there is an extra shifting stage on a different axle. It it's simplest form, this stage can either be in neutral or driving the PTO. Both 254/968 and 973 also incorporated anti-backdriving and PTO lock into their PTOs too(kidn of hard to explain, best to look at the 973 CAD). Both 254/968 and 973 had shifting gearboxes, so they also had a 3 position actuator on the drive shift so that they could place the shifting dog in the middle of the two shifting gears so that their wheels wouldn't turn when they were lifting.

254 build blogs are back in place: http://team254.com/resources/. It's under FRC build.

[apalrd]

33

We designed our 2011 and 2012 gearboxes to accept more motors.

Our 2011 robot was CAD'd to accept a FP/550 through 2-stage Banebot planetary (driving the 28-tooth high gear on the intermediate cluster shaft via a larger spur gear), we later found (no surprise) we didn't have the weight. We later designed a lighter system with 775 motors (before the case short issue was well-known) into CIM-u-lators, where the extra motor would be directly across the gearbox from an existing CIM. We would have welded two CIM pinions together into a pinion/coupler for the two motors, but we didn't feel like we needed the extra power by CMP, we didn't have the weight before CMP. We had very little choice in gear ratios since we were running super shifters direct drive to 6" wheels, any gear changing would require us to machine gears, something we cannot do well.

Our 2012 robot was designed to hold a CIM-u-lator on an intermediate plate in the gearbox, we CAD'ed it with clearance for two 550's (making 8 drive motors in CAD). We assembled the gearboxes with a single 550 (plus the two CIMs), and removed the extra 550's due to weight concerns. We were able to re-gear the robot appropriately since we used a chain final drive, although we still didn't gear low enough.

We do not have definitive data showing anything about the third motor. The math says it should be good, we've been going purely on that.



As for a power take-off, your goal is essentially to have two or more clutches: One engages the motor to each load. A single dog from Andymark acts as two clutches - In one position one gear is engaged and the other is not, in the other the reverse happens (it exists in neutral with neither engaged for a short time during a shift, if you play with it you can get it there manually).

In the case of 254's 2010 robot, they had two dogs which could engage two separate outputs in three different ways each, plus a one-way ratchet. The lower dog engaged the drive into either motors via low gear, motors via high gear, or neutral. The upper dog engaged the arm into either the motors, neutral, or locked (a fixed plate with dog-face). Something like that would likely be the simplest way to design a PTO in FIRST. I would have to look at their design again to tell you anything else about it.

[AllenGregoryIV]

I was just looking at 973's 2010 robot on FRC-designs.com last night and they have a power take off system.

http://frc-designs.com/CAD2010.html#frc973

[CalTran]

I'll check the CAD tomorrow and open a new thread if need be to stop hijacking this one

[roystur44]

For a PTO, look at 254/968 in 2010. It was basically another shifting setup. Another piston and another dog and dog gear.



Not 100% sure but I believe they used a 3 position piston for the main drive shaft. They had 2 speeds and neutral. They shifted to neutral and activated the second piston to engage a gear on to the main drive shaft. All the power was routed to the second drive shaft via the engaged gear.

Boy was that cool to see.

[sanddrag]

Boy was that cool to see.Even more fun to help design ;) But most of the credit on that lies with others.

[AllenGregoryIV]

Anyone know what 3 position piston they used in these setups?

[R.C.]

Anyone know what 3 position piston they used in these setups?

Allen,

I'm unsure what the exact p/n 254 used. But similar to what we'd use:

FOP-040.25/0.25-L

Its a Flat Multiple Position Cylinder from Bimba.

http://www.bimba.com/Products-and-Cad/Actuators/Inch/Compact/Low-Profile/Flat-One-Multiple-Position/

You can order them here:

https://secure-102.portline.com/e-bimba/item.cfm

-RC

[Tristan Lall]

188 had this capability in:

2002: Old Bosch Drill, Johnson F-P & Old CIM (single-speed)
2003: Bosch/Scintilla Drill, Mabuchi F-P & CIM (single-speed; never did get the radial-engagement shifter to stay engaged)
2004: Bosch/Scintilla Drill, Mabuchi F-P & CIM (two-speed pneumatic dog clutch)
2006: gearbox was designed to be adaptable for combinations of up to 5 motors (F-P with planetary, big CIM, 2 small CIM, and even a Globe if you really wanted to), but only built with 2 small CIM (three-speed servo-shifted DeWalt XRP1; in retrospect, the shifter was pointless for that game, and the servo too slow to actuate the mechanism)

(I have to admit it's been a few years since I dealt with these robots, so where this differs from previous information, assume this list is inaccurate.)

1 That was a crazy design that combined all these motors into a single XRP gearbox—which actually worked surprisingly well, apart from the manufacturing errors, and the tendency to break shafts. (The interface shaft between the XRP gearbox and the motor gearbox was too difficult to make on our lathes as a single piece of steel; it had a joint that wasn't initially robust enough. Also, the output shaft was AISI 12L14, replacing the hardened steel output shaft that comes with the DeWalt gearbox; we found out that after about an hour of running, they would fail in a ductile manner, in torsion. We should have picked a better material, preferably something easily hardenable like AISI O1.)

[Tristan Lall]

Incidentally, I also built some bolt-on adapters that add F-Ps to a pair of SuperShifters, with only reversible modifications to retain COTS status. That's another way to do more motors quite cheaply and easily.

(Andy, you should sell those....)

[MichaelBick]

Incidentally, I also built some bolt-on adapters that add F-Ps to a pair of SuperShifters, with only reversible modifications to retain COTS status. That's another way to do more motors quite cheaply and easily.

(Andy, you should sell those....)

He does sell those, they are called AM planetaries.

[Gregor]

He does sell those, they are called AM planetaries.

I think he means adding the FP as the third motor in the super shifter, not replacing a CIM.