Showroom 2DOF + traction loss using Brushless DC motors

I needed to try something different. Either new BLDC drivers, or something other than the arduino SMC3. New drivers will cost me $200, and take a while to arrive so I decided to try the other option first.

As mentioned above, FlyPT's (pmvcda) system runs on an ESP32, and uses the motor hall sensor for positioning which has a big advantage for me.

So I got PCBs for the ESP32 created at allPCB.com. Gerber file for this, created by @fcppenta (thanks mate) is in the link below. For a new accounts at allPCB the first order is free (including worldwide shipping with DHL) so they were created and delivered within a week! Wow!

https://www.xsimulator.net/communit...hless-diy-actuators.10799/page-38#post-218195

I've made all other connections, including limit switches, and uploaded the ESP32 code created by the genius @pmvcda (thanks mate) from the link below:
https://www.xsimulator.net/community/threads/flypt-6dof-brushless-diy-actuators.10799/#post-139682

There are still a lot of things that could have been done wrong, a lot of potential for crossed wires despite the PCB doing much of the connections work. My BLDC driver direction pin is active low, so will it run in reverse? Will it run at all? Will it go up in a puff of smoke?

I'm a little afraid to find out right now, and running out of time. I'll recheck all connections, and give it a go tomorrow.

[EDIT] The motor ran as soon as I powered it up. Next I pressed one of the limit switches, and the motor stopped (makes sense it was at it's limit). Unfortunately I have not managed to get it to run again. More effort is still required.

Now that we are starting the holiday period, I'm hoping to make some progress.

Since my last post some parts have arrived. They are:
1 x 1500w 24v PSU
1 x Universal Joint
3 x Wantai BLDC-8015A drivers
and a little more steel

PROGRESS I think... with one of the ESP32 boards connected to a Wantai BLDC driver, and after uploading ESP32 code by @pmvcda, via Arduino IDE, my motor now runs smoothly and completes it's calibration. Very happy about that.



Tomorrow I'll try to go beyond calibration and see if I can get smooth movement via SMC3Util and Sim Tools input, and then it will finally be time to start building the actual frame itself, starting with the traction loss part.

Thanks for checking in @noorbeast. There's certainly been a lot of effort, but not much progress unfortunately. I've burnt out one of the ESP32 boards and now using my one backup.

As mentioned above, motor calibration works well and runs smoothly using the ESP32 with associated code provided by pmvcda, and Wantai driver

The main focus now is to get FlyPT Mover to connect to do something useful with a "loop" as input, and this is where I'm stuck. It either does nothing or it goes a little crazy (randomly). Frustrating!

Following that I want to switch to Simtools as the input.

I'll do more reading (there's soooo many posts) and will also post questions on either of pmvcda's threads soon, but if anyone has a simple FlyPT Mover setup that can run a "single rotating actuator" Rig with output to pmvcda's ESP32 code, then could you please send it to me?

Thanks a lot

After much more effort that I thought it would take, it finally seems to run smoothly in response to in-game feedback. I must say I'm very happy to have reached this point.

The overall aim is to use FlyPT Mover (so that BLDC motors run smoothly) with Sim Tools for game input.

Sorry for the rubbish video but hopefully shows it responding to feedback from LFS:


There's still an issue with a very specific order of events needed to connect the ESP32 and make it run (that took some time finding!), but at least for now I know that sequence.

I'm now moving on to building the frame, and I have redesigned it again for simplification. I needed to keep in mind that there must be room for the 3 motors where they are needed and to reduce overall height by allowing the motor arms can go all the way to the floor at their lowest point. The image shows the two components and then how they will work when mounted together.



EDIT: For my benefit the sequence is:
1./ reboot the ESP32 using the onboard reboot button
2./ while it is rebooting (or even calibrating), click on Connect in the Flypt Mover output to connect the ESP32
The motor/lever should then centre itself and you know it is ready to go.

Very happy to report that after a little more effort than expected (always the way!) I've now completed the physical build for the traction-loss frame. It seems to be working as well as I could have hoped. The remaining 2dof part will build on top of this and should be less of an effort overall.

It will look a lot better after a coat of paint and some more end caps, but painting will have to wait until all hardware (motors etc) are mounted as that will require more welding.



Things I learnt:

• Getting the 2 frames flat and level was difficult. I clamped them together while welding but still needed to adjust the height of one wheel.
• The wheels are difficult to fit due to needing spacers either side, inside the spare profile. Maybe need a better solution that putting the wheel within the metal profile.
• I can't go roller blading any more (it's probably been 20 years now anyway)

It moves so easily that I'm wondering if I use a wiper motor for this part... but then I will have an unused BLDC and driver. I guess I'll stick with plan A, and getting the motors attached, connected and tested is the the next step.

Finally got nearly everything mounted, including all electronics and power supply:
- 3 bldc motors/worm gears
- 3 bldc drivers
- 1800w 24v Power Supply
- ESP32 electronics (in the plastic box at the front)
- Limit switches

Just wiring to go now. My aim was to make is self contained and relatively small in size.



I still need to add the movement arms to the front motors (these will swing up/down close to the front wheels), and one last 40mm square bar which those movement arms with push up/down and which my existing rig will be connected to.

Also, I now need to redesign/rebuild the entire pedal end of my existing rig quite as well due to a new set of pedals arriving tomorrow. That's the trouble with building things to fit equipment... doesn't work when new equipment arrives :-(

If you want to control a brushless motor
Vesc and ODrive seem to be more powerful brushless drives It has been successfully applied to DDW (torque mode)
I think he can also be used very well in motion simulator (position mode)
Might need some modifications to these powerful drives

Thanks for the informati0n @fzxj520 I haven't heard of DDW torque mode, and was not aware of Vesc. I'd looked into ODrive a previously and it does look good alright.

I'll definitely do some reading on these options for my own learning. I would buy one or other to experiment with but ummm, have you seen the prices? Unfortunately that's a bit of a problem

Really appreciate your effort in sending me this information. Thanks again

sorry for my english from translation
I'm talking about the force feedback steering wheel project
DDW = Direct Drive Wheels

Based on modifying the drive's firmware
Two communication modes with brushless driver
OpenFFBoard + VESC
https://www.racingfr.net/wbb/index....réation-d-un-firmware-ffb-pour-vesc/&pageNo=1
MMOS + VESC
https://www.dropbox.com/sh/bbs92yfinfr63la/AAAcAwNRpS1ms2fB8cmdIACia?dl=0&fbclid=IwAR1kz5Kx5LzQWrP-QIRik_DzL43LLnHtFd5yomlXoyyzF1u1zhu28J2fw7E

In addition, if the motion simulation uses springs reasonably, the load of the motor can be reduced and the response speed can be improved.

Both Vesc and ODrive have rc pwm input
There is also the host computer code output by rc pwn in the forum
Hope someone can integrate it

Thanks @fzxj520 That's another thing that I hadn't heard of, but RC PWM could simplify things as far as wiring goes if I have understood it correctly... I need to do some more reading on it, and see if I can find the code you refer to. These motors and drivers are such cool technology.

I read all your post and I loved your chassis designer how do you intend to make the upper part? Will it be just a rectangle to put the cocpit on top? about controlling BLDC motors i would try something like odrive but i don't know how to do it yet, i'm looking forward to the rest of the project, you have my respect for the fight so far.

Hi @workfayer. I'm very flattered that you like the design, thanks for the feedback. A lot of thought went into the criss-cross design so that the motor arms can swing low towards the floor without hitting the lower frame while swinging sideways for traction loss, as I'm trying to minimize the overall height.

The Wantai BLDC-8015A drivers I'm using (the black heat sinks you can see in the centre of my frames) seem quite good, and are approx USD$20 per motor. From memory an oDrive is more than $200 for 2 motors, and I would need 2 of them.

Yes my racing rig, which currently sits on the floor will be mounted on top of this unit as you know. I have recently completed the mechanical build (just wiring to go but other things like house painting keep getting in the way) so I can show you how it will work in the photos below. The blue bar in the photos represents the central bar of my racing rig (photos of that are in my first post). The blue bar is not the correct length, and I may still have an opportunity to lower the universal joint it will sit on. A rectangular frame should be able to be mounted in a similar way I expect.

Hi all, I'm happy to report that I've finally got things working to a certain extent... currently with one of the three motors so that I can test running traction loss. In the video I'm sending it a sine wave and it's responding nicely.



I'm now very motivated to get the other 2 motors wired up, so will do that very soon. I just hoping they can provide enough force for where they are located.

Remaining to do:
- add limit switches to the remaining 2 motors
- connect 2 remaining motors/drivers to the esp32 processor.
- connect the emergency stop button
- replace the 2 small fans in the power supply with a single larger and quieter fan
- consider lowering the overall height by shortening and rewelding the universal joint
- tidy and clip wires

Very good, the platform seems very compact under the cockpit my only fear of a compact platform is the instability in more sudden movements, however I am waiting for the tests with pilot in motion to see how it behaves, how much in mm you got the displacement of the loss of traction?

Hi workfayer. Thanks for the feedback. Yes I am waiting as well! It turns out that there is still one more hurdle... the ESP32 (I have 2 of them and they act the same) continually reboots when plugged in to my gaming PC, but they have no problem on other PCs. All tests so far have been performed connected to a different laptop (actually it worked once on the gaming PC a couple of nights back so I didn't realise there would be any problem). Arrrgh!

EDIT: By a lucky accident, I have found that it connects perfectly when Simhub is not running. I have Simhub connecting to an Arduino for my wind simulator. So is there a clash of comms ports or a clash of devices IDs or similar. At least I can connect the ESP32, and I know where to focus my attention to resolve whatever is going on

Regarding the traction loss movement, it's about 75mm on either side. Not a great deal but I hope it will give the required sensation. EDIT: It feels like enough movement when in game, but I think I'll now have to investigate motion compensation as I use VR.

Question: Does anyone know what the motor levers/cranks (see photo) are called, and where they can be purchased? Mine are DIY and I never want to do that again

in my country there is no store or market for this type of part I think you need to order it from a company that does custom machining by sending the measurements to them