Custom Two Speed Gearbox

[Adam.garcia]

Hello All,

I am currently working on a custom two-speed transmission for an offseason 6 Wheel WCD.

The design intent for this drive train is to keep the gearboxes as light as possible, while obtaining a very high speed (18-22) and a moderately low speed (4-6).

These are the gears I'm planning on using.
http://wcproducts.net/gears-20-dp/


Available CIM gears: 14
Available Spur Gears: 15,28,30,45,48,56,60
Available Dog Gears: 35,45,48,60

I have been having trouble producing the desired gear ratios. Can anybody who is experienced with designing gearboxes of this type help me figure out a set of gears that will produce our desired result?

Thank you in advance!

[Nemo]

The gear ratio depends on the wheel size you are using.

I suggest finding JesseK's calculator in the CD-Media white papers section and calculating your desired gear ratios first. If you do that I can suggest gear combinations.

[Adam.garcia]

Sorry for not including this in the description. We are going to be using 4" AM Performance wheels:

http://www.andymark.com/product-p/am-0393.htm

[Thad House]

Here (http://www.chiefdelphi.com/media/papers/2681) is a gearbox calculator that i wrote. It takes wheel size and the gears and calculates the ratios and the speed. Is the 18fps the theoretical or actual speed that you want. The actual is usually about 80% of the theoretical, and the actual is what is calculated by my app. 18fps actual is VERY fast and very hard to control. I would also try using a 12 tooth CIM gear, which can help give some better ratios. Those can be found at Andymark.

Also dont count out some of the smaller gears from andymark. From andymark you can also get 14,16,19, and 24 tooth spur gears, in addition to a 28 tooth dog gear. You can also get a 50 tooth aluminum spur gear from andymark as well which will give you a weight advantage.

[Adam.garcia]

The 18 fps is actual.

I understand that this 18fps is a very fast speed, especially for FIRST robots. Essentially our team would like to have a proof of concept, with "swappable" gearboxes for our practice bots.

This "swappable" gearboxes will allow this drive train to be used as a very robust prototyping tool, come next build season.

Thank you, I will check out this gearbox calculator.

I'll keep you posted, as I inevitably will need more help from the CD community.

[Adam.garcia]

Do you have any sample gear ratios that your team has used in the past?

[Thad House]

We have never done a 2 speed in the past, but here is a picture (http://team254.com/blog2011/wp-content/uploads/2011/01/IMG952011010995221542.jpg) of the ratio's team 254 did last year. They used 3.5 inch wheels, and they only weighed 100 lbs which allowed them to gear that high.

[Adam.garcia]

Oh wow, this is great!

Do you know of any other gear ratios that teams have released? (besides www.frc-designs.com where teams have posted their CAD models)

Thanks

[R.C.]

Sorry for not including this in the description. We are going to be using 4" AM Performance wheels:

http://www.andymark.com/product-p/am-0393.htm

Adam the two common spur gear ratios are:

15:60 - 30:45

&

15:48 - 28:35

The First set of ratios is really nice for around a 4" Wheel. It allows you to get a really nice choice of reductions. We do a 14:45 with a 15:60 - 30:45 shifter set to get around 16/6 FPS. Which is really darn nice. Wanna gear faster or slower? Change the input ratio.

But the ratio above allows a really nice, compact gearbox. If you went to a smaller wheel ~3" Or so you could go to the smaller set and use the Latter ratio.

All the information above is for a 2 stage 2 speed, you can use either one if you go to the three stage. Its just more weight/money/less efficient.

If you have any questions, please post them up!

-RC

[artdutra04]

Adam the two common spur gear ratios are:

15:60 - 30:45

&

15:48 - 28:35

The First set of ratios is really nice for around a 4" Wheel. It allows you to get a really nice choice of reductions. We do a 14:45 with a 15:60 - 30:45 shifter set to get around 16/6 FPS. Which is really darn nice. Wanna gear faster or slower? Change the input ratio.

But the ratio above allows a really nice, compact gearbox. If you went to a smaller wheel ~3" Or so you could go to the smaller set and use the Latter ratio.

All the information above is for a 2 stage 2 speed, you can use either one if you go to the three stage. Its just more weight/money/less efficient.

If you have any questions, please post them up!

-RCYou can use the standard AndyMark ratio of 48:15/35:28 and direct drive 4" wheels using only two gear reductions if you use 11t press-fit pinions on the CIM motors.

[R.C.]

You can use the standard AndyMark ratio of 48:15/35:28 and direct drive 4" wheels using only two gear reductions if you use 11t press-fit pinions on the CIM motors.

Def missed that guy :p

You guys still using the 11T pinions?

-RC

[Adam.garcia]

Adam the two common spur gear ratios are:

15:60 - 30:45

&

15:48 - 28:35

The First set of ratios is really nice for around a 4" Wheel. It allows you to get a really nice choice of reductions. We do a 14:45 with a 15:60 - 30:45 shifter set to get around 16/6 FPS. Which is really darn nice. Wanna gear faster or slower? Change the input ratio.

But the ratio above allows a really nice, compact gearbox. If you went to a smaller wheel ~3" Or so you could go to the smaller set and use the Latter ratio.

All the information above is for a 2 stage 2 speed, you can use either one if you go to the three stage. Its just more weight/money/less efficient.

If you have any questions, please post them up!

-RC

Wow, That's something that I did not know. I will start working with my design team on designing the gearbox.

We have been confused about this for about 4 weeks or so, and it just hit me that I had a huge and knowledgeable community right here on CD. I'll be sure to keep everybody posted if I have questions.

[Nemo]

A couple more questions:

1) Are you planning to machine any of your own axles?
2) Are you willing to include a steel gear or two?

You'll have to answer yes to at least one of those.

Option 1: If machining your own axles, you can do a 2 stage using the 60:15 / 45:30 gear set. 2 stage will of course be lighter, smaller, and more efficient. You would need to make your own shifting / wheel axle with the dog gear slot and the hole bored for the shifting shaft. That gear set has a high/low ratio of 2.67, while you're looking for something like 4:1 (20 ft/s and 5 ft/s). So you'd have to settle for something like 16 ft/s and 6 ft/s with this gearing, or 20 ft/s and 7.5 ft/s.

At any rate, you only get to select two of the gears in this setup: CIM gear and 1st stage output gear.
CIM Gear: 12 or 14
1st Stg Output: 40, 45, or 50.

That gives you ratios between 4.86:1 and 3.33:1 for high gear. Those are very fast. (AndyMark has a lightened steel 40T gear that only weighs 0.17 lb, so I consider that a good option for a light gearbox).

*The option Art mentioned using 11T gears would be nice and light, but you'd have to come up with 11T gears somehow. They are not available off the shelf, so you'd have to find some 11T spur gear stock or get a company to Wire EDM them for you or something. Incidentally, does anybody know a place that sells 11 tooth stock? That would be a pretty sweet find...

Option 2: If you use this shifting shaft (http://www.andymark.com/product-p/am-0131.htm), you can make a custom 3 stage box with all COTS parts except the gear plates.

Stage 1: 12T or 14T CIM gears; 40T output gear is the only one that fits.

Stage 2: Basic set is 48:15 and 35:28, but you can do the optional set that flips it to 48:15 and 28:35 for the 4:1 speed ratio you want. That makes your high gear faster without changing the low gear. If you do 3 stage with 4" wheels and you want really fast speed, it will be hard to get a low enough gear ratio if you don't choose that flipped 4:1 gear set.

Stage 3 input: Only 17T and 20T input gears are available off the shelf (the shaft is 1/2" keyed). You could also buy a 3/8" hex gear, then bore it to 1/2" and broach a 1/8" keyway on your own. Requires arbor press + broach tools. If you did that, it opens up the 28T and 30T gears as input gears, which would provide faster gearing.

Stage 3 output: Can be 40, 45, 48, 50, 56, or 60. BUT, the stage 3 center distance has to be big enough to avoid interference between the wheel shaft and the 48T stage 2 output gear, which eliminates, for example, 40:20 as an option (unless you shrink part of the hex shaft...). If you bore and broach a 30T gear, you could do an all-aluminum 45:30 or 50:30.

...

The 2 stage designs seem more appealing for small wheels and fast speeds.

I don't have any experience with robots that go this fast. I must be missing something, because the numbers look bad to me on paper - it seems like one would have to be driving longer distances to make 20 ft/s theoretical gearing better than 16 ft/s theoretical gearing since the 20 ft/s bot will accelerate more slowly. Unless your robot is very light or the gearboxes have a third motor.

[Jeff Waegelin]

Def missed that guy :p

You guys still using the 11T pinions?

-RC

Yep. We used them last year on our drive modules, and again this year in our transmissions - we made custom plates and shafts for the AM 2-speed transmission and used the 11T press-on pinions. I love them because you don't have to deal with keys or clip rings or anything.

*The option Art mentioned using 11T gears would be nice and light, but you'd have to come up with 11T gears somehow. They are not available off the shelf, so you'd have to find some 11T spur gear stock or get a company to Wire EDM them for you or something. Incidentally, does anybody know a place that sells 11 tooth stock? That would be a pretty sweet find...

We get ours made via EDM - I've not seen any 11T gear stock, although that's not to say it doesn't exist - I just haven't looked for it that hard.

[Tristan Lall]

One thing to think about: if you want a bit more versatility in keeping your high ratio high, add more motors. That way, you can still manoeuvre hard in high gear, without worrying about current drawn due to acceleration. They gave us a lot of motors in the KOP, so you might as well use them.

I tend to design 2-speed gearboxes with the ability to spin the wheels when pushing against an immovable object in high gear, to account for the fact that shifters don't always work right. Failing to do this means that if you can't shift, you'll stall the motors pushing on things, and possibly damage the motors. (Depending on your confidence in the mechanism, this may be an unnecessary tradeoff.)

[Adam.garcia]

I tend to design 2-speed gearboxes with the ability to spin the wheels when pushing against an immovable object in high gear, to account for the fact that shifters don't always work right. Failing to do this means that if you can't shift, you'll stall the motors pushing on things, and possibly damage the motors. (Depending on your confidence in the mechanism, this may be an unnecessary tradeoff.)

Can you please elaborate on how I should go about ensuring that I design the high gear to slip at stall?

Are there formulas/a calculator/white paper that somebody has written describing this phenomenon?

I'm using Blue Nitrile Tread from McMaster. However, I do not know the coefficient of friction.

Thanks

[Tristan Lall]

Can you please elaborate on how I should go about ensuring that I design the high gear to slip at stall?
If you're designing the wheel to slip instead of stall, the force the gearbox can exert on the ground through the wheel should exceed the maximum frictional force that the ground exerts on the wheel. The maximum gearbox force is a function of the motors' torque at stall. The maximum frictional force is a function of normal force (itself a function of weight) and frictional coefficient.

If you're designing the gearbox to always shift into low in case of a failure, you should characterize the possible failures and try to deal with them. This isn't easy for shifting mechanisms that don't return to a home position (e.g. servos) during a malfunction, and is also difficult if you can't walk up to the robot to override a malfunctioning valve (which you can't, at least not during a match). Since redundant actuators and control channels aren't really feasible under most circumstances, taking this design approach will invariably lead you to compromise—but that might be good enough.

Are there formulas/a calculator/white paper that somebody has written describing this phenomenon?
John V-Neun wrote one (http://www.chiefdelphi.com/media/papers/1469), and has updated it over the years (http://www.chiefdelphi.com/media/papers/2059). It's very useful. Others have written similar tools (like this (http://www.chiefdelphi.com/media/papers/2274) and this (http://www.chiefdelphi.com/media/papers/2681)), but I haven't tried them.

I'm using Blue Nitrile Tread from McMaster. However, I do not know the coefficient of friction.
Probably about 1.0 on carpet (an educated guess). See here (http://www.chiefdelphi.com/forums/showthread.php?t=51400) for a bit more information.

[sdcantrell56]

Probably about 1.0 on carpet (an educated guess). See here (http://www.chiefdelphi.com/forums/showthread.php?t=51400) for a bit more information.

Any type of roughtop seems to be more like 1.2 to 1.3. I would try to ere on the side of caution rather than pushing things so definitely use the higher coefficient of friction.