$450 Mige Direct Drive Wheel using MMOS and AASD

Mines 3d printed in PETG, 100% infill. No problem at all. Even heavy drifting and letting it snap back and forwards continuously it just took it like a champ.
Alternatively you could look at ascher racing as they do a clamp set if you want metal.

You can use Ascher racing boss kit and 70mm adapter, along a cheap quick release (from ebay). It all easy to mount:

Great write-up @Peacemaker105! I got this setup and working well following your instructions on it.

To address the humming I used a pair of low pass RC filters with Pn188 = 1 and they completely eliminate it. It's a simple and inexpensive solution and most people may already have the parts on hand. From some filter simulations it should add about 1 ms of latency for large sudden changes in torque but in practice it will be less because the torque commands typically are smooth transitions over a few milliseconds rather than large steps.

Here is the circuit for this filter. The filter frequency is about 1500 Hz. The values are not critical and close values can be substituted.

Another problem I had was the loss of calibration when the AASD driver was enabled. It was significant and after being powered on for a few seconds the calibration had already drifted to full right steer. This was the result of EMI affecting the encoder signals to the STM32. To address it I used the a differential pair receiver and used both wires for each differential encoder pair. I used some SN75176B chips for this that I had on hand but they are 5v parts and not ideal and there are other 3.3v parts would be a better fit. The encoder inputs on the STM32 are 5v tolerant so it worked by connecting the output of these differential receiver to a ~5K resistor and then to the encoder pin on the STM32. I'm not sure why I was having so much of a problem with this when it seems to work fine for everyone else.

I didnt have issues with encoder calibration but my other usb devices would trip out and my loadcell brake pedal was jittery. I just put ferrite chokes on pretty much all my computer cables and ran a ground wire from the motor's flange to the AASD's exposed ground. solved the problem for me.

I'm using one RC filter with 2.2k resistor and 100nf cap and seems to be adequate.

Did you try using shielded db25 cable?? DIY cabling can be trick to make work with all the EMI the motors emitting...

Also, I used DAC that has latency only 8us...

Latency is not bad.
Modern ffb is 60hz-400hz.So there are 16ms-2.5ms　interval steps.
The rise time rounds those edges.
simcube2 and fanatecDD also have the same purpose filter.
I think that Pn188 has the same effect.
As a result, the smoothing(overall filter of mmos) value can be reduced.

did you find out what the usb problem was caused by ?

what I did:
1. put ferrite chokes on all usb and serial cables going to PC, stm board and servo drive
2. connected additional ground wire between servo motor's mounting holes unpainted side and the servo drive's ground.
3. moved the usb cables of stm board to usb 2.0 ports on the back panel, they were previously on the front panel

i'm not sure if 1. or 3. had the overall effect of stopping the crashing of usb devices. 2. was definitely helpful as i also had random jitters in my loadcell pedal input before the servo got the external ground wire.

so i think it was either ground issue, or emi issue, or both

to be fair, the ferrite chokes and the extra ground seems like a general good practice for OSW based on some online search e.g.: https://issuu.com/simracingcoach/docs/tutorial_eliminate_emi_issues

thanks for the feedback

i have tried ferrite chokes on all cables

moving the usb connection to a laptop on battery power has worked

i will keep trying !

Awesome man! thanks for the schematic, will help a fair few people out that's for sure

Did you not need a DAC to clean up the signal to remove the humming ??

Yeah that's what they've been reporting..

a simple rc filter can do the job, but you will get less latency (under 1ms correct me if I'm wrong) with that specific DAC on the TDD board. it's also 1 to 1 conversion between the 12bit pwm of the mmos and the 12bit output of the dac chip.

Hi all, and thank you for the effort you are putting into this. I have started shopping around for an AASD servo drive+matching motor. I noticed many sets are sold in one version without a brake and a more expensive version with a brake.
Should I buy the version with the brake, or can I go for the cheaper one without the brake?

Hi, I am shopping for a servo motor for this project, and found a webshop that sells a 130ST-M10015 (1500 rpm, 1.5 kW, 10Nm) and a 130ST-M10010 (1000 rpm, 1.0 kW, 10Nm). Looks like they produce the same torque, but as far as I can see from you post, the one that has the lowest RPM is the best. Is that correct?
Also, a more general qustion: When selecting a motor, is the rated torque the most important factor, or are there other specs that also are important?

you do not want the braked version

Hi @Peacemaker105 just wanted to give ya update that i got my Wheel working at last , @Thanos gave me a pointer that i missed out, because i bought my Core407 board independantly ,theres and solder bridge that needs to be made on the underside of the Core407 board , once done , the system seems to be working just as you both intended. so gonna go buy another wig cos i ruined the last one wondering what i missed LOL cheers as always mate for great work and support . its really highly appreciated ..gonna go try and tame this beast now Thanks again for bringing this option forward for us tight gezzas

Its the 5v jumper, should be set to 5v OUT !!!



Thanks
Thanos

Did a bit of research, I think I found out why low rpm servos are better.
This table is from the link in the original post (https://www.aliexpress.com/item/32868662195.html?spm=a2g0s.9042311.0.0.67274c4d5qpWWV)



The servo that is rated at 1000 rpm has a torque coefficient of 2.2 N.m/A whereas the one that is rated at 1500 rpm has a torque coefficient of 1.67 N.m/A. So the servo that is rated at 1000 rpm can deliver more torque with a lower current (Ampere), it will give you more bang for the buck, so to speak. The servo that has a higher rated rpm can deliver the same torque, but you will need a more powerful servo driver that can deliver a higher current. At least that's how I understand it, let me know if I'm wrong

you should get the 130st-m10010. this the motor they call "small mige" in the opensimwheel community i think.
better torque coefficient, lower current rating = less heat. rpm does not matter as much as you will not even reach 500rpm in a driving simulator.