• aditya mehrotra.

robot arm funsies: hacking a Dobot M1 [robot arm project]

So ever since I saw Iron Man, I've always wanted a robot arm that's like the JARVIS Dum-E thing where you can talk to it and it can do stuff. I think I finally found the platform where I can make that happen.



So this is a Dobot M1 the ORRCA team got a while ago here at MIT. It is supposed to be able to 3D print, laser cut, CNC, etc... But it never worked so it's been collecting dust. What I kinda want to do is make it not collect dust again, get it working over the course of a few months, and then make it do things (like CNC or 3D print in very fine detail). The first thing I'm going to do is take all the useless body panels off. Dobot may call them aesthetic - frankly they could've done a better job.



So here it is... all the wires taken apart, all the faces of the robot taken off. And if we look at it, we may have a problem. The problem is that these motors do not appear to be stepper motors, they appear to be brushless DC motors at least that's what they're labeled. They look like brushless DC motors with some ridiculous encoder strapped on the back. So it's a 48V brushless DC motor.



Here's the arm after some more disassembly. This is definitely telling us that the motor controllers have two parts, first is the three-phase BLDC motor, and the second "hat" is an encoder with 9-Pins coming out of it. I'm going to see if I can find the specs on the motor controller to get an idea of how to control this thing. It seems to be labeled: PB105001AC


That returned zero results on google. Like actually zero. And even after some disassembly the motor controller is also unlabeled. Completely. Have to love low-quality manufacturing.


So what we're going to try to do is replace the WHOLE SYSTEM of electronics besides the motors. Because whatever we come up with will probably be a lot better to work with than the nonsense we have on here right now. We're going to have to figure out the encoders or have to figure out some way of doing position sensing.


IDEA: APRIL TAG ON THE END-EFFECTOR, HIGH-MOUNTED CAMERA, DO POSITION DETECTION OF THE APRIL TAG IN THE FRAME AND PID CONTROL FROM THE APRIL TAG, NO NEED FOR JOINT ENCODERS.


Honestly... not a bad idea.


So we're going to need to find a 48V power supply (got it, it came with the robot arm). And we will need to find some motor controllers.


Or wait... we don't... because we have ODRIVES! I keep forgetting I have like (6x) ODrives sitting back in my dorm room. Maybe we will just use two of those for control. That's a better idea! We're going to need to get them hooked up too a linux computer and working. It might be easiest to start with those from scratch and re-flash them considering the lack of experience we have with them. There's only one problem. I think the ODrives I bought are 24V and not 56V. Which makes meeee, a dummy.


I might just sell my old ODrives and get new 56V versions and just 2 of them at that.


In the meantime. I'm going to lug this thing back to Simmons until I can figure out what to do with it.


#robot_arm #projects #making #modifying #hacking #robotics #omgrobots

2 views0 comments
© copyright 2019 | aditya mehrotra