The HAMMR is the High-mobility All-terrain Multipurpose Modular Robot platform. The idea is to have a basic model for a differential drive four wheel drive robot that can traverse tough terrain. Additionally, I want this thing to go really fast and be all-around bad-ass.
This project was hugely inspired by the Hormes All-Terrain Robot. There are a lot of similarities, but this is going to be my own take on it.
The frame is made from 30mm 80/20 extruded aluminum. It was available and didn’t require me to be married to any specific dimensions since this is a prototype. I was able to quickly move and re size things that needed to be adjusted. I’m glad I did, since there were a lot of major problems initially.
One awesome thing I discovered was that instead of using 80/20 Inc.’s stupid proprietary (read: overpriced) connecting hardware, I just used 1/4″-20 square nuts along with their bolts. The square nuts fit perfectly in the channel and there was plenty of ‘catch’ when tightening the bolt. The only problem I ran into was when fastening thin metal to the 80/20, I needed some washers since the shortest bolts I could find were about 1/2″.
The axle plates are 1/4″ aluminum that I cut using an angle grinder. I found some youtube video one day that suggested using a fence and just going at it with a cutoff wheel. At the time, I hadn’t any experience with a milling machine so it was my only option. It worked pretty damn well and I’m satisfied with the results.
The 10″ wheels are from Harbor Freight and they have 40 tooth #35 chain go-kart sprockets attached to them using long bolts and 2″ PVC spacers. It was impossible to get the spacers perfectly square so the wheels weren’t wobbly (as you can see in the above image), so I eventually turned them on the lathe. Axle is a 5/8″ steel rod from Tractor Supply, and there’s a bunch of little collar spacer things I found on Amazon. Two CIM motors are used to drive it, and each are connected to VEX 12 tooth double sprockets. I really like the idea of using only two motors instead of four like most other robots. This trades off a lot of power, but saves space, weight, and cost.
For the electronics bay, I used some of the metal that I scrapped from an old dryer. I had been holding on to this rusty sheet for a couple years and finally found a use for it. Angle grinder, flap disk, and my 18″ sheet metal brake helped me turn this scrap into a nice area to hold all the electronics. I neglected to get a good picture of the finished product.
Now you’re thinking:
3.3:1 gear ratio? What kind of stupid are you?
Yeah I know! Well that’s why its called a learning experience. My goal was to gear it so it would be really fast, but I overlooked the significant torque loss that would disable the robot’s ability to roll over sticks and bumps in the road. Spoiler alert: it went really fast. It took a good few seconds to get to top speed, and it went about 15-20mph – but it wasn’t able to turn or roll over small bumps.
I’ve got two SPARK motor controllers by REV Robotics, a cheap buck converter from china, an RX module and two 12V 20AH lead-acid batteries in parallel. Fuses not shown here.
With all the gear and hardware together, it weighed about 70lbs.
Testing was really exciting. It was so fun to see something this big and heavy just fly down the road. In fact, I was even more excited when it was barreling toward a parked car and lacked the ability to decelerate quickly. Narrowly missing the car, I continued the test drive.
Those controllers only handle 60A continuous and 100A peak. What are you doing?!
Yeah well, money. Those are super cheap controllers and will be much more suited for the task at hand when I adjust the gear ratio. They got pretty hot, and I’m 100% sure they were operating over their rated current for longer than rated periods of time. Eventually the robot died, all electronics were off and I just assumed it was the end.
Weeks of design and building for 5 minutes of fun. I thought I fried my controllers, and I was all out of money. Huge bummer. But it turns out after actually testing everything a few weeks later, I only fried the main fuse. I swore that I was getting 12V in places that I shouldn’t be if I had a blown fuse, but whatever. So nothing is fried and I can continue on.
Progress since testing
I went back to solidworks and also did a bunch of math to figure out what ratios and speeds I wanted. I have to fabricate more sheet metal electronics bays to work around the new gearbox going in. I’ll be using an Andy Mark CIMple gearbox which will give me about a 15.5:1 ratio and a top speed of about 8mph.
The new electronics bay was able to be a bit deeper. I also painted it since the old one had started to rust rather quickly. This was fabricated using an angle grinder, drill press, and small sheet metal brake.
This summer I’ve actually setup an independent study on campus that will entail designing and building a 2-speed gearbox for this robot. I’m pretty excited since I’ll have a chance to design and fabricate something with moving parts and high precision. I’ll be writing it up on this site as it happens.