Milt's budget build - DIY 2dof compact seat mover

Howdy folks. I'm on the verge of starting my 2dof seat mover build. I've always been intrigued by motion platforms and computer / physical feedback interfaces. I've got lots of DIY experience including building a custom CNC router in the last year or two. I have been browsing these forums for a while now and finally ready to pull the trigger on my own build.

A few details I have planned out in my head:

Arduino + 2x dual monster moto control
2x wheelchair motors (160W - 250W)
12V server power supply + some beefy capacitors to handle surge.
Motorcycle final-drive universal joint
Frame build primarily of metal square tube with plywood sheeting

I know the wheelchair motors are designed to run on 24v, but I'm trying to do a budget build using the cheap MM drivers. I already have a 24v 20A unregulated power supply that might be up to the task if I change to 24v drivers (Pololu dual G2?).

Arduino Mega and MM drivers are on their way from China... I have the U-joint and motors in my ebay cart.

Quick question about wheelchair motors. I understand I'll need to remove the braking mechanisms and I imagine the motors will freewheel when not under power. Will this pose any problems in a seat mover build?

Looking forward to documenting this build!

Just did the servo RC model with simtools this weekend. Fun little project and neat to see a miniature seat mover in action! Lots of design considerations going forward and the little model will certainly help.

Video:
https://imgur.com/a/dAqch9x

My wheelchair motors have 25:1. They definitively move down when power is cut off.

BUT they are really fast. Kick in the back

Good to know. I can't find gear ratios for the motors I'm looking at but guess I could figure it out once I get them in hand. I'm thinking about leaving the electromechanical brake on the motors so I can lock them out when needed. From my research it is usually 12-24v applied to the brake will release it...?

Mine had manual brakes, but I know from CNC applications that 24 V is used to release the brake.

Thanks. Off topic, but out of curiosity, what did you use wheelchair motors for on a CNC?

Ah, no I don't use wheelchair motors for CNC purposes. But the brakes of axe servos almost the same. And those use 24 V to release.

Slowly but surely gathering parts.

Motors purchased off ebay. $50

Found a cool vintage seat on the local Craigslist. $20


Moto Monsters arrived from Aliexpress. After ordering I began reading about all the problems people have with getting bad drivers. Soldered headers on to one of the three MMs I ordered and checked the output voltage. Sure enough, one side runs fine, the other has reduced voltage in one direction. I still need to test the other two before requesting a refund from the aliexpress seller.

Any suggestions on the easiest way to test MMs without soldering headers on to the MM or soldering directly to the board? I have a couple mini-grabber cables, but not enough for the full job. Can I just run 5v to the motor enable and PWM input pins to test?

If these MMs keep giving me problems I'll bite the bullet and get a 2x32A Sabertooth and run at 24v

Onward!

I've been looking at motor drivers and settled on the Pololu G2 24V18 for my own project so maybe you should consider that too? The 18A is continuous, plus it has current chopping ability so you can set the controller to keep the current at a set point. The price point is pretty good (~$80CAD) and it has good features.

As for the testing without soldering, maybe you should grab a couple of alligator clips for easy testing because changes are you'll need this capability again in the future.

what kind of motors are those? 24v from a wheel chair? super nice price
fer

I have looked at the Pololu G2 and thought it looked like a good package. I couldn't find many cases where people were using them here on xsimulator and would like to use something I can get community support for if I run in to problems. I also found a few reviews suggesting the peak current capability of the G2 was not enough to deal with the rapid reversing of motor direction under load resulting in failure. The price point of the Pololu is hard to argue with, but after reading more about the Sabertooth and rave reviews it gets I am willing to spend the extra cash.

They are 24v wheelchair motors. I believe they are from a Pride Jazzy 614 HD wheelchair, according to my google search. The motors are marked CM808-110B. Output after the gear box is ~120rpm at 24v. They seem really nice to me, although the inline gearbox can be backdriven easier than a 90 degree gearbox. I can turn the gearbox shaft very slowly by hand if I really really crank on it... not sure if that will pose problems down the line but I doubt it. I can pass along the ebay seller's name if you are interested. It looks like they have a ready supply of them available.

Update on my progress thus far. Realized my sanity was more important than dealing with the shoddy monster moto boards so I splurged for a 2x32 Sabertooth. Definitely money well spent.

Took my first foray into milling aluminum on my DIY CNC router to cut some 17mm holes and keyways for the lever arm. Turned out great in my opinion!




Now on to finding a universal joint and ordering some of those fancy rotary hall effect sensors!

Finally got to test out the sabertooth and SMC3 using a quick test rig and a 50K potentiometer I had on hand. I ended up getting a 24v 800W switching DC power supply off ebay for a more compact design. I wired the pot to an Arduino Mega, connected one motor to the Sabertooth and installed the SMC3-SPS library. Following all the instructions in the SMC3-SPS thread I successfully got movement and I am blown away at how powerful and quiet these motors are!

Quick question about SMC3-utils. As I tune the PID loop and PWM parameters in SMC3-utils, these variables must be stored on the Arduino somewhere, correct? So after I get the motors tuned how I want them, I can close SMC-utils and open SimTools and the PID/PWM tuning will still be applied, right? I don't see any mention of this in the documentation so I want to be sure I'm not missing something.

Waiting on a U-joint then I'll start fabricating the frame.

Nearly two years later and I've decided to pick up this seat mover project again. The lack of free time, wifely concerns about finding space for the rig, and inelegant solutions for positional feedback led me to put the project down and step away. Free time and spousal approval are still lacking, but I finally settled on a positional feedback strategy that I like.

I purchased a Kangaroo X2 motion controller from Dimension Engineering to pair with my Sabertooth 2x32. Since I don't have an easy way to mount a potentiometer to the output shaft of my gearmotors, I opted to put a quadrature encoder on the motor shaft itself. The motors have the short 8mm shaft exposed on the opposite side of the gear box so it was relatively easy to 3d print an adapter plate and mount encoders. I'm using ATM102 encoders and will incorporate limit switches for homing.



Last night I got my test setup working with SimTools. Took some focused reading of the Kangaroo X2 manual and relevant postings on this board to get serial communications working correctly, but in the end everything was moving as it should. Loving the resolution of these encoders! Provides wonderfully smooth operation with ~5000 discreet positions within the 90 degrees of travel on the gear shaft.

Next step is ordering some yoke flanges and universal joint to build the pivot mechanism!

Alright! Making progress with my Sabertooth/Kangaroo x2 configurations. I have successfully configured my motors to use a single limit switch for homing then restricted the lever movement to ~90 degrees of travel using the soft limits set in the DEScribe software.

My question: Is there any disadvantage to having Simtools output 14 or 16 bit output resolution (16,384 or 65,536 discrete positions)? It seems like 12 bit (4096 positions) would be plenty, but with my encoder resolution I could conceivably up the bit resolution for finer positional control... curious if there is a reason not to.

Slowly but surely moving along. Universal joint is complete. I ended up ordering two Spicer yoke flanges (part 2-2-939) and a Moog 369 universal joint (bearing?) to make what I needed. Took a bit of work to press the caps on with a c-clamp and a hammer, but finally got it together.



I started mocking everything up on a wood stand to get the geometry right before I figure out how I want to fabricate a more sturdy base. My goal is to have this a compact build and I think it's looking good so far.



Also added a single limit switch to each motor so I can home the motor orientation on SimTools boot up (still need to get these wired). Only $10 for both these switches off eBay...

Well, two years after I started the project I'm finally nearing the finish line. All it took was a global pandemic and stay-home order to find the time to finish putting it all together. I did a little testing in LFS to make sure my geometry was right then tightened everything up. Had to scrounge around the garage to find all the nuts and bolts I needed, but made it work in the end.

Unfortunately, my simulator doubles as my desk and is currently disassembled so I can work from home. I plan to pull out my wheel this weekend and give it a proper few laps!

I really appreciate all the help documentation and forum posts on here... It really made building this possible.

I'll post a video of movement this weekend after I get all set up.