Excellent ! Congrats Sir. If I may, do you plan to create a tutorial on how to control your board thanks to an Arduino / ESP32 board ? It will be very handy for a newbie like me. Have a great day !
It is on the TODO list, although it both needs additional software work and isn't usually an optimal solution. Usually if you are using something as capable as moteus, you're better off with a more capable host platform. Raspberry Pi's are on the lower end, but more capable SBCs with dedicated PCIe CAN-FD adapters are also likely candidate host platforms. Basically, the cost difference between a rpi and an Arduino/ESP32 is not significant in a moteus based system. I will admit that the development experience with an Arduino or ESP32 can be better in some cases though.
I don't have a tutorial yet, but I do now have a draft library to control it from Arduino boards with a MCP2517FD or MCP2518FD shield or integrated controller. If you're interested in testing, check it out: github.com/mjbots/moteus-arduino or hop into the mjbots Discord to discuss: discord.gg/W4hUpBb
Thanks for the interest! Right now, there is not a documented build procedure for the A1. Keep your eyes peeled for the eventual A2, which is designed to be easier to construct!
So that you don't have a sparking event. moteus has large capacitors on the input. The power supply has large capacitors on its output. If you turn on the power supply, then connect it, you will cause a large current transient (possibly hundreds of amps) as two capacitors balance out. Large transients like that can cause damage through a number of mechanisms.
The specs for the mj5208 motor say that it has a peak power of 600W but the moteus r4.11 controller is rated for a max of 500W and the power supply provided with the dev kit can deliver a max power of 120W. Under what conditions does the mj5208 draw more than 120W and more than 500W? Is the 120W supply adequate for testing?
You are correct that you'll be limited to the lowest power of the motor, controller, and supply. The power the motor draws can be roughly modeled as the speed multiplied by the torque. With the 120W supply you won't be able to reach the peak torque when stationary, or as you noted, the maximum power while in motion. That said, it is still pretty "torquey" and can go at the maximum speed for 24V no problem. AC supplies just get really big fast when you start increasing the maximum power.
@@awesjosh Thanks for your reply. I'm new to building large robot actuators so I'm trying to gauge my expectations. It's difficult to find a benchtop power supply larger than 300W, there seems to be a step function on price and availability after that. So I was curious if providing for the maximum was really necessary.
I'd recommend starting with anything, even the devkit supply and see where it gets you. Often switching to a battery can be a more cost effective high power solution when you need to go there, although that comes with its own challenges in terms of system power on. DIY solutions also sometimes use re-purposed server power supplies, which can be found in the 500W-1000W range more readily. If you do go with a battery, you have to be able to connect the battery that is never off to the giant bank of capacitors that are multiple motor controllers without sparking, thus pre-chargers like the power_dist.
@@awesjosh Thanks Josh, I just got my dev kit and it's working great. I think it will be nice to have a working baseline as I try to experiment with different parts.
The quad A1 is specifically designed around the qdd100. More generally, quadrupeds are only barely feasible with direct drive motors, and that is you count every gram and accept a bunch of other compromises.
Hard to say. Unless you have access to free motors and PCB manufacturing / assembly, the development kit from mjbots.com is actually a pretty good deal: mjbots.com/products/moteus-r4-11-developer-kit
