Hello everyone. I am getting back into 3kg sumo. I have a score to settle and have been inspired to make a new killer sumo. Many people build these bots with a wheel directly connected to the gearmotor. This approach limits the motor size, specifically the length. Some have created custom wheels and gearboxes to get around this restriction. My approach will use an MXL timing belts to allow one largish motor per side. Because the motor lies on top of the drive wheels it can be much larger / longer. The motor used in this mock up will be replaced by a much larger motor and gearbox. I’ve just started really, but parts are starting to arrive. I have some progress to report. The parts will eventually be aluminum. This is a very early stage prototype. 8)
I’m just wondering though what your plans are in regards to connecting the wheels to the gearmotor? In the setup you have pictured it looks like you’ll use drive belts that are half the width of the timing pulleys in order to use two belts on each pulley? ie. belt from motor to wheel 1, belt from wheel 1 to wheel 2, belt from wheel 2 to wheel 3?
It’ll be a real nice setup. I’m curious whether such narrow and fine pitched belts can handle the torque. I’ve never used MXL, but am familiar with XL size.
Good luck, looking forward to seeing the progress!
If the tires are multisegment rings on a hub (or could be), you might be able to make a simple mold and produce your own using silicone sealant (which reportedly has good traction qualitys). DIY tires might also allow some experimentation for traction improvement. Hmmm, now I’m thinking of thick o-rings on a large tubing hub.
The tires are not multi segment, but one piece molded polyurethane cast directly onto the wheels. The marks on the surface are from the mold itself. It was a stack of 1/4 inch steel plates. Their CNC mill couldn’t cut anything thicker. They even degassed the polyurethane after it was mixed to eliminate small air bubbles.
The polyurethane tires (ShoreA 20 as I recall) are about 50% better than silicone sealant and, perhaps, too soft. Good luck trying to come up with better tires for sumo. You will need it.
Sorry I forgot to answer this yesterday. The wheel and motor pulleys are aligned and can accept a 1/4" wide belt. My plan is to use 4 idlers to position the single 1/4" wide belt to have the required pulley contact. This is going to put a pretty good amount of tension on the belt. It may not last very long, but it doesn’t really have to. I may have to replace them after each event. There are some impressive MXL belts with Kevlar that are very strong. I hope we can hit them hard and fast without slipping. We will see.
Here is an update. I got the bearings in for the idlers. Everything is lining up well. I need to remake the motor mount panel to add a way to add tension. It works!
Here is an image of one of the motors I am considering. If it’s not got the umph I need it has two larger brothers in the catalog. hehe!
Considering the length of the motor housing, are you going to make the other wheel assembly identical as the first so that when you flip it around, one motor is toward the front of the sumo and the other motor is toward the back? This configuration will allow you make it more compact width wise anyway. Only thing though is one motor will need to be reverse biased or the sumo will run in circles.
Great job. I love the stout compact look of this thing.
Yes that’s the idea. The timing on the motors is adjustable so we can either set them both for neutral or adjust them both for optimum performance in one direction.
It looks like a Promax “400” or similar. I’d go with the “05” big brother from the start. Remember too much power is never enough.
As you correctly pointed out, adjustable timing is mandatory.
The belt drive appears to be about 1:1 ratio, so the gear box should be about 25:1 or 30:1, but not outside of that range. Plan on using more voltage than the motor rattings. Add batteries until you get the speed that you want. Also figure on a speed control that can safely handle 40 to 50 amps per side. Again, overkill is necessary.
I’ve never done anything with belt drive, but you may be loosing a lot of power in the idlers. Perhaps removing the idlers and eliminating the center wheel will be an improvement. Experimenting is required.
Wait. Why am I telling you all this? Forget this post.
If he removes the center idler, wont he be loosing power to the center wheel? If all the wheels are to be belt driven, the I would guess there would be a need for all the idlers.
I would not be surprised if I am missing something.
OK I’m coming clean. Here is all the data so far. I’ve been meaning to do this anyway…
The pulleys are all 32 tooth and can accept belts up to 0.289". I’m using a standard 0.250" wide MXL belt. The idlers are 0.125" ID, 0.375" OD and 0.156" wide, I’m using two of them per position. The tires are 2.125" diameter and 1.5" wide.
I completely expected comments about the timing belts. This is uncharted territory for many. If I were building for Japan rules, vacuum and or magnets, then it would be out of the question. But with just 3kg pushing down on the drive train the bot is likely to bounce before anything breaks. Off the forum I have received comments ranging from “Those timing belts will not be a problem at all” to “The thing will fly apart!”
Here’s the information on the cobalt motor.
Maxx Products PMX4812.
Diameter 1.14"
Length 2.13"
Weight 4.1 oz.
Shaft 1/8"
Winding 12/#22.5
Typical voltage 8.4-9.6vdc
Test voltage 8.0
Static RPM 15,600
Static amps 19.2
Static Power 154W
I believe static means not flying?.. These motors are for airplanes. The current demand is what made me want to try these before going to the 05 model.
The gear box is a planetary box with 27:1 reduction and dual planetary gears, just double the height of the current planet gears for more strength. Dual ball bearings on the final gear of course. 15,600 rpm with 27:1 is 577 rpm at the wheels. Which is 64.2IPS! We can slow it down from there if necessary.
I agree to some extent that 4WD has inherent advantages over 6WD, especially in my situation, but I really want 6WD…
I hope to have it moving under it’s own power in a few days. 8)
I think this will work. The only thing I can think that might go wrong is when the motor goes from 0 rpm to 15,000 rpm in an instant, there is a slight chance that the belt going around the drive pully could jump out of the groves from the stress, and go Zzzzzzip!
If the power is ramped up then I think it will be ok. I say try it full on power and see what happens first. Just a thought I wanted to throw out on the table.
EDIT
Second thought, if both motors work together then there might not be as much stress since all six wheels and two motors are working together.
I’m still working on the bounce problem. After five years I still don’t have a good solution. The bounce will put shock loads on the timing belts.
Static in this case means with a 6X4 prop and not moving. In an airplane this would be a worse case scenario, i.e. maximum current draw and heat build up. In a robot that has its maximum load at 0 rpm, the current draw will be much, much higher. I’d double or triple the current to 40 - 60 amps for the speed control design.
Your motors are 12 turn, 22.5 ga. The big brother 05s are 6 turn 19 ga. That will be a huge jump in current. I’d plan on 90 to 100 amps per side. That is why I stuck with the stock R/C car armatures with 27 turns.
The gear box is ideal for the application. Too bad you don’t have more of those PGHM-14s like I use. At under $30 they were a steal. No, you can’t buy mine back at any price. 64.2 IPS is good. Don’t even think of slowing it down. I just noticed that I haven’t measured Executioner’s speed in its present configuration. Early on, with 6 cells, it did 85 IPS. Now it has eight cells, so I guess it is over 100 IPS.
No problem with that. 6WD was always much more intimidation than 4WD. And, once your opponent knows that he is going to loose, the rest is easy.
Well I have found the original source. I have been wanting to get them back on the website. 8)
Hmmm, I must have missed that little detail when I was spying, um I mean researching your robots. Huh, 85 IPS, really…
I’m not going to get too caught up in specific numbers. The goal will be to move as fast as what is controllable. If we can control it well at 65 IPS we can add more cells till it becomes less stable. Then slow down in software. The most headroom possible, the most performance possible. Weeeee!
That is great news. I have wondered what I would do if one of the gears broke. There is nothing else on the market that comes close.
Bot Brawl 2006 when Executioner spun off its rear tires and Bot Brawl 2007 when the speed control smoked the speed was 85 IPS. When I rebuilt the smoked control, I found much better N and P channel MOSFETs. They were so much better that I added two more batteries. It is scary fast and I am intimidated every time I plug it in. I must have forgotten to update my web site with that information.
That is the way to do it.
I spent last weekend in Marion, Ohio, at the National Robotics Challenge. There sure was a lot of your stuff there.
I have a lot of catching up to do! 
