# Gear Ratios

For next year, I’m looking to use 4" IFI wheels. I’m planning on using a crab drive and I’m designing it right now. What would be the best ratio for it. It would need to be pretty well rounded in speed and pushing power.

Well this is not as easy as just saying “ok you need a 5:1 ratio” There are alot of things that need to be considered when designing a transmission. On these forums there are quite a few resources that will help you but here are some initial things to think about.

What motors are you using to drive?
what obstacles are you going to be facing?
What you want your top speed to be?
What kind of gear set up are you trying to use (efficiency depends on this)?

I would take a look through the white papers section on the site and get a feel for how and what needs to be looked at to design a transmission. You are on the right track though by asking, there just isn’t a straight forward answer to your question.

You could do a chainzilla, and mount some 2 speed transmissions between the 4 crab modules, and have bevel gears drive the crab modules via chain from the transmissions. it would work great!

Well, I plan on using four small cims. I plan on a live shaft, two stage reduction. I’m not sure how fast i need it to go, but no slower then 7.5 fps and no faster then 12 fps. I’ve made about 4 designs so far for a crab drive and when i finally did the math, it would go like 4.5 fps. So now I’m just trying to find the best ratio.

IFI

wheels run on roller bearings and dead axles.

We used 6" IFI’s this year and modified them to run on a live axle. It works better than having 8 feet of chain on your robot.

Are you talking about 7.5 - 12 fps theoretical speed using the motor’s free speed or are you talking 7.5 - 12 fps actual measured robot speed?

I have found that the actual robot speed tends to be about 75% of the theoretical speed using the free speed of the motor. We design our single speed robots for no more than 7 fps using 75% of the free speed. (many people will tell you to go faster than that)

The other thing to keep in mind about robot speed is control. If you do not have a very good feedback based control system it will be a challenge for you to make good use of a robot traveling at 7 or 8 fps. Often you can perform the task faster if you slow the robot down.

I personally would not use a live axle in a crab module. You can flush mount the last gear of your reduction to the side of the IFI wheel, (make sure it is concentric to the axle center) and taking the wheel out is as simple as pulling the dead axle out. Ground OD shoulder bolts work really good for axles.

I hope that helps.

Matt B.

First, I would reccomend being cautious when designing a system to be used in a game we havn’t seen yet. There may be obstacles on the field that a 4" wheel cannot pass (like the steps in '04). Additionally, you may find that, depending on the field size/shape, you may desire a different speed/torque output than you were previously planning. You may want to use motors that would be allocated to the drive motors, or the steering motors elsewhere on the robot, like the manipulator. Also, I’d be wary about using a high traction wheel in a swerve drive system. The very nature of the system recquires you to counter-act the traction of your wheels in order to be able to turn, so you may wear down the tread relatively quickly, and will need more power to turn the wheels than when using a lower traction design.

I’m saying actual speed. The motor specs say that 5310 rpm (+/- 10%) is the free speed. The normal load is at 4320 rpm.

This year, our robot went roughly 10fps and I had no problems controlling it at full speed. We had no sensor or feedback either.

Live axles are actually easier to do with crab drive in my opinion. I’ll look into dead axles, but hey what ever works best I’ll do.

Well, the whole point of this is a summer project. If next year’s game has obstacles, I’ll change up the configuration to use 6" or 8" wheels, or not use crab. I just want to get a prototype so we can give the programmers something to do, holonomic drive practice (i know its not the same as Mecanum or omni but its the same general concept), and to look into if its worth implementing on a robot.

I’ve already thought about if we need a small CIM for another apparatus. We either go to 3 crab modules or use two Fischer prices. Also, we have too many steering motors to choose from as it is. So we can easily integrate a globe or a window or a vandor in really quickly with minimal headaches.

I plan on using 1" w, 4" d IFI wheels. There shouldn’t me much of a problem turning. If we go with 3 modules, it’ll be even easier to turn because there will be an independently controller wheel.

Thanks for your concerns and I hope that with your help, I can put together a monster crab drive.

Cut the ratio in half?

Using free seed, if you put a 12 tooth gear on a CIM then to a 36 tooth gear for a 4:1 reduction, you’d get 1327.5 rpms. Put a 15 tooth gear on that shaft which drive the wheels which are using another 36 tooth gear you get to 553.125 rpms. That translates to roughly 9.65 ft/s. Use 95% efficency for each gearing, that goes to 8.71 ft/s realistically. A respectible speed that is, IMHO, a good middle ground speed. Hope tht helped.

BTW: Those are commonly available gear sizes, so its a realistic gearing

Alright guys. I got my stuff figured out and I made it in Alibre Designs. I used an 8.75-1 ratio which is about 8.61 fps. Thank you all for your help and I’ll post some pictures. Thanks again. O yeh the reason the gears look wierd is to simulate the meshing, I’m to lazy to make real gears. I’ve also done the research and found places to buy every single piece i need. Aluminum, shafts from Mcmaster; gears from W.M. Berg; and bearings from Boston Gear.

Any particular reason? What are the specs on the gears you plan on using? I’ve found Berg to be a bit pricey at times. Also, what are the specs on the bearings you’re using (ID, OD, thickness, and flange/no flange)? I’m wondering why they have to come from Boston Gear.

Well I found all the gears at Berg. I’m using 60-20 and 70-24, 20 pitch. The reason i used these is for spacing purposes. The Bearings are .375 and .5 id. The widest they can be is 3/8 from the inside of the flange to the end. They are flanged. If you know of a cheaper/better place off hand, then please send me the link.

For bearings, unless you can get them somewhere cheaper (<\$4 each), I’d recommend Small Parts p/n brf-08 (1/2" ID) and brf-06 (3/8" ID) You won’t find them in their catalog but if you do a p/n search on their website you will.

For gears, at first I was shocked at how large they were but I took a look at your design and did a few numbers on the clearances and center to center distances and whatnot and it seems like you’ve got it pretty well figured out. You’ll be able to take a lot of meat out of the gears and save a lot of space. You might take a look at Martin gears part numbers S20XX (3/8" face width, 14.5 degree PA) and TS20XX (1/2" face width, 20 degree PA) where the XX is the tooth count. You can order them from Motion Industries (who has a store local to you (afaik) 1310 MILLER PARK WAY, MILWAUKEE , WI, 53214 -3603, Phone: (414) 649-2400

Now, I haven’t looked at the Berg gears but the Martin gears have thick heavy hubs that you’ll definitely want to machine away most of.

You can also take a look at McMaster to see what gears they have. I don’t think they have the 70 tooth.

Also, make sure you get a bore small enough on the CIM motor gear, but you probably already knew that.

Ahh yes. I overlooked small parts inventory. I’ll give motion industries a call also. Well thanks for your help. When I finally get one built, I’ll post alot of pictures. Thanks again for everyones help.

Well I think Mr.Needel and Sandrag answered your questions about the gear ratios, I would like to add an idea about your 8 ft of chain comment. I was wondering if you have ever seen how team 71 turns there crab modules it’s really cool and is a much lighter and fool proof concept then chain and 71 uses it so you know it works. They use a type of window motor belt strapping it’s a really strong plastic that is used in your automatic car windows. I would definitely pm someone on 71 about it or at least see if it would work for your application.