2DOF Build log - Status: Construction Begun

Hi all, I am starting my first motion sim build so this will be the log for it. I have collected pretty much all of the parts i need to begin construction and I am now trying to design a compact setup.

Basics:
Full motion 2DOF sim (not a seat shaker)
BIG Wheelchair motors
Modded/Hacked wheelchair controller - controlled with Modded version of SMC3 Code
CV joint for pivot

I am trying to design something that is fairly simple to construct but should be durable.

This is the design I have come up with - Is it just pointless and should I stick with a generic square format? Is there any disadvantage to this design?



It provides 20 degrees of movement on all axes with only about 35 degrees of swing either way on the motor axes (100mm CTC)

Welcoming feedback!

Edit:
Rather than reworking my existing pedals and re-using the G27 I had, I have also decided to go down the route of OSW and some new custom pedals. While this extends the build, it should make something of much better quality.

I'm also considering a wireless wheel. Utilizing some 18650 cells an arduino and an NRF24 module to transmit back to a Pro-micro configured as a joystick. It means you'd have to recharge the wheel but would keep things much simpler with lager lock-to-lock rotations.

The calc tool does not really help much as the specs of the motors are relatively unknown. They are however 24v motors and are about 30% bigger than the other motors often seen here. I can probably determine the gear ratio but I believe it is quite high. The motors are almost 400mm long in total. I can turn the motors around to get the ams closer to the pivot point but perhaps I am just over complicating things with the 'Y' Design though.

So i done some checks:
Ratio - 24:1 lower than i thought but...
voltage is 24v, resistance is only 0.5ohms (48 amps!!)

That gives me ~1.1kw per motor...

I dont know how that compares with yours but thats some big power.

Position sensor feedback on motors

Some of you may have seen my other post on the Wheelchair controller. I was testing using some cheap pots from a local supplier but then I decided a better way to go is with Hall sensors.

I used UGN3503UA Hall Effect Sensors and some 4x4x2mm rare earth magnets salvaged from a dead RC brushless motor I had lying around (I fly RC racing quads...)

This is what I ended up with...


Unfortunately the legs on the hall sensor are a bit too short so I added a recess into the back of the holder to allow for soldering wires but I have not printed that version yet.

This is what the pot version looks like mounted to the motor - I intend on drilling and tapping 3mm holes to secure the mechanism. The holes are elongated to allow for center fine tuning.

It has a hex output that I added:

So I started on the frame base today

Welded up the Y frame and mounted the CV Joint.

Tomorrow:
Weld the plate to the CV joint and make up a seat base that will bolt to the plate.

I finished welding on the seat mount and I have done a concept design for the upper frame.

I'm torn on the simplicity of a single mount for the wheel and pedals (meaning you end up with the wheel mount between your legs) or this type of design.

dimensions subject to change...

I'd appreciate comments for/against either ideas.

At the moment, a g27 wheel so 'soft and laggy' but this is likely to change to OSW in the near future (I need to start a new thread on that - got some ideas that no one seems to have thought of yet). I just worry that the center bar will get in the way but the construction is so much simpler.

Small update - So after waiting 4+ weeks for my Arduino Nanos to arrive, I discovered that I accidentally ordered the Atmega168 version and not the Atmega328 version - Doh! Luckily, the SMC3 code still fits on the 168 (It has less RAM and Progmem but runs at the same speed).

I soldered it into the wires from the wheelchair controller and tested it ok. I also printed a small holder for the Nano so I can safely secure it inside the unit.


I have made up new gearbox covers for the motors as the original ones were angled and I wanted a flat mount. 10mm Aluminium with a pocket machined out as the main output gear comes past the machined surface of the gearbox.

One thing that I did not realise until after going down this path was that the main motor mount bolts are actually on an angle too. It's not that harsh of an angle but annoying. Too hard to change - If i have to I will make up some angled washers for when it is bolted to the frame.

Some more work
I have decided to do a full rework now and have scrapped my G27 in favour of the MMOs DD wheel ('Cousin of OSW' thread here). I have also decided to make some new pedals too. Something lighter and more compact than my previous sim pedals.

I acquired a 1000W 48v DC motor from Motion dynamics, ordered a few IBT_2 H bridges and an STM32F4 board.

I took the idea of the work of .dumbass (Member here) who designed a mount for the rear of the motor to mount all the wheel bits. His design was not quite what I wanted so I re-drew his idea into something that works for me.

I published it on thingiverse https://www.thingiverse.com/thing:3208235

Done some more work on the wheel today. Still waiting on my stm32 board so using a pro micro for now. Annoying that it landed in australia back on the 6th and i still dont have it!

I removed the green screw connectors from the ibt2 boards, scraped away some of the coating and soldered the wires direct to the boards. Many of the green connectors had stripped screws anyway.

Reflowed the chips on the boards to fill in all the vias to help with heat transfer. I also added a thin heatsink pad between the heatsink and the board. I am not confident that the heatsink is properly isolated in the stock form and we are dealing with big current potentials.

All smoke tested ok other than needing a pullup on the zref encoder i forgot.

Next - make a wheel hub to go on the motor shaft.

Looking good mate
My only concern would be the stability of the base.
I know you want it compact, but I got a big shock when I first got my rig to move and it threw me around like a rag doll.
Naturally I had all the settings farked up, but it definitely has some power and you are more prone to trip a tri than a quad base if the legs are too short.
Just my own opinion and you know what your building
I look forward to your OSW build that you spoke of too.

Yeah, I have considered the stability factor. If it is an issue I can always expand the base.

Wheel coming along more, I made up a hub adapter to connect the shaft to a wheel.
The standard 11t steel sprocket is secured into the adapter with 3 4mm bolts. Captive nuts on the back. There is about 1mm clearance to the motor body. I installed the bolts backwards for now because I could not be bothered shortening them...

Printed in PETG @ 25% infill and 4 perimeters for good strength, plus the wheel bolts go all the way through so I don't see that there will be a problem with de-lamination of the layers.



Also some pics of the H Bridges


Interestingly, the encoder that I have is supposed to have 2k pullups built in but I could not get it to work without them externally (the pullups I installed were 10k)

I also picked up a race seat for free - It was basically useless but I removed the trim, repadded it all and retrimmed it in a faux suade...

Looking good mate.
You did a good job on the seat re-trim, my seat was just a Bumtree special but works the same as a $1K new one would.

Minor update
Got the other motor gearbox cover machined up and mounted the motors to the frame


5X20mm mild steel has been welded to the sides of the 50mm square tube to mount them. They use 5/16 UNC thread bolts (bloody Americans and their refusal to go metric!) - There is not quite enough clearance for a standard bolt, it needs to be a socket head bolt, will need to order some...