pic: Omniwheel Minibot



This is a minibot I CADed up a little while ago and then spent some time learning how to render with. I really like the results personally.

The bot itself is made using 4 identical sideplates that are symmetrical in all directions and inspired by gearboxes found on WCDs like 254s. Other than that the only manufacturing is cut the tubes to length and press star nuts into the ones in the middle.

In it’s current (easily modifiable) configuration it’s geared 4.29:1 with 4 in omni wheels for a final free speed of 21ft/s with 1 CIM per side expandable to 2 per side if you widen the center beams.

The plan is to use RC receivers and LiPo batteries to save weight, space and money (and maybe add a little extra juice, the CIMs and Talons can handle it). Before the controller and battery (so everything in the rendering) this should cost just under $500 to make.

That’s kind of cool. What on it makes it cost $500? $150 for the talons, $50 for the CIMs, and $300 for the aluminum/wheels?

Edit: ignore my post

I actually parted out every part and pre-tax and pre-shipping it works out to $465. Those gears, pulleys, belts, bolts and tube nuts really add up. 4.5 in bolts are expensive.

I think with something of that weight could get away with a significantly higher top speed.

Recall that a 150lb robot with 6 cims could top out at ~18fps, and that only because they can only travel 50ft before stopping.

If you were running it down a street or something you could push a full FRC robot to 25+fps over a hundred feet.

An antweight in comparison with just 2 cims should have no problem reaching 30 fps in 15-20 feet or so.

Then again, if you are operating in more confined places like classrooms or such, you would gain a control advantage topping out slower.

Also be careful, CIMs produce a significant amount of noise in the FM band. We converted a kitbot to run on a FM system and had control problems when we got 30+ feet away. We decided to seriously isolate our antenna from the CIMs, by putting the antenna higher and also wrapping the motors in layers of aluminum foil in an attempt to Faraday cage them.

We use a 2.4 GHZ system on our V1 minibot and have had no issues. We’re currently looking at running it on a 14.8V to 22.2V lipo so it should be faster than 21 ft/s that was estimated at 12v.

If you post your BOM, we might be able to help you find less expensive sources for your components.

1/2" OD Round Tube Axle Stock (3 feet) http://www.vexrobotics.com/vexpro/hardware/shaft-stock.html?lulz= $11.99 2 $23.98 VEXpro
VersaHub (1.125" Bearing Pilot, 1/4" Thick, w/ Plate Sprocket Mount) http://www.vexrobotics.com/vexpro/wheels-and-hubs/versahubs.html?lulz= $2.99 6 $17.94
60t Gear with 1.125" Bearing Bore & VersaKeys http://www.vexrobotics.com/vexpro/gears-and-gearboxes/versakey-gears.html?lulz= $21.99 2 $43.98
4" Omni Wheel http://www.vexrobotics.com/vexpro/wheels-and-hubs/omni-wheels/omni-wheels.html?lulz= $17.99 4 $71.96
VersaPulley Kit (HTD 5mm) - 42T, 18mm wide http://www.vexrobotics.com/vexpro/belts-and-pulleys/htdversapulley.html?lulz= $4.99 4 $19.96
14t CIM Gear (Steel) with Mounting Hardware http://www.vexrobotics.com/vexpro/gears-and-gearboxes/cim-motor-gears.html?lulz= $7.99 2 $15.98
Timing Belt (HTD 5mm) - 170T, 9mm Wide http://www.vexrobotics.com/vexpro/belts-and-pulleys/htdbelts9.html?lulz= $10.99 2 $21.98
Flanged Bearing - 0.500" x 1.125" x 0.313" http://www.vexrobotics.com/vexpro/hardware/bearings.html?lulz= $2.99 8 $23.92
tube connector*25 http://www.mcmaster.com/#94290a510/=sdvhn7 $15.13 1 $15.13 McMaster
4.5in bolts-1/4-20 (pack of 5) http://www.mcmaster.com/?lulz=#91251a089/=se2p5q $8.26 3 $24.78

total $279.61

This was copy pasted from my spreadsheet so sorry about the formatting.
the first number at the end of each line is the cost per unit then the quantity then the total. It’s mostly vex stuff because it’s cheap, available and simple other than maybe a different source for belt i doubt it. We’re also from Canada so that cuts down options. And honestly we will be making McMaster orders and Vex orders anyways so it’s very little trouble.

That looks like a very clean design. Well done.

But I can’t understand the reason for using identically-oriented omniwheels in all four corners. Won’t that let the 'bot slide sideways uncontrollably?

Sliding sideways uncontrollably sounds like a lot of fun. It’s designed to be brought around to demos but the vex parts make it easy to switch the omni-wheels out for something less silly.

If you do go to 4 CIMs I would gear them 1 to each wheel. If you keep the control system and batteries in the center I bet you could get it to flip and keep driving “upsidedown”, that would be a pretty cool trick for demos. (added benefit, no belts)

Make sure the lipos you get can handle the potentially huge currents 2 CIMs would cause. Trying to draw too much from them can cause thermal runaway and fire.

In terms of size and weight, I think you’d get a more satisfying performance by switching to 550’s or 775 sized motors as you can shave a lot of size and weight off, without a large power loss.

I would have to agree with Adam; for this size, you wouldn’t need that much power from two CIMs to move the weight of the bot.

Essentially 33 ran this drive this year.

Quite a few teams have run “butterfly” drives that have this + drop down tractions as a second mode.

We found that the robot tracks straight as on carpet the omnis still have appreciable lateral rolling resistance.

Our driver quickly figured out how to drift with it, and how to control those motions roughly.

In summary, it’s more far more controllable than people assume.

I’ve made several FTC robots in my day that incorporate this drive. It’s completely controllable and it’s pretty fun to drive. I’m almost certain FTC 3864 has used it 5 years in a row now.

The reason I chose to go with CIMs is because it’s actually cheaper and simpler than using those more sensible motors. The Cim motor isn’t much more expensive than a 775 for instance but because of the lower speed and higher torque we don’t need any additional stages and we don’t need to buy either non-vex gears or versaplanetaries. Also, because it’s so light I doubt we’ll see the massive stall currents that CIMs are capable of.

I’m actually planning a WCD styled minibot that uses 500 series brushless motors (made out of 30mmx15mmx1.5mm alu extrusion, I need to practice my metric) so it will also fit the RS 550.

Additionally you have to look at its intended use. We regularly do demos for 6 hours without cool downs. The additional thermal mass will prevent over heating the motors.

I see you’ve got the rendering pretty well figured out. Could you share a bit about how it was done? It looks photorealistic :slight_smile:

I’m no expert but it was done in Solidworks photoview. Make sure perspective is on. I also added a spotlight to make things look a little more realistic.

Controllability is relative. If you want to slide sideways then it does exactly what it’s supposed to. Have you not seen 33’s drive train (inspired by the JVN Build Blitz team I’m sure)?