The Adventures of DIY Sim

Hi all,

Thought it would be neat to catalogue the journey of DIY, with an ever present "why didn't I just buy off the shelf", the satisfaction of using something made my your hands is always worth it.

A couple of years ago I built myself a sim rig frame out of 25x25mm (1") SHS with a G27. Several shortcomings have been discovered but between wedding plans, the wedding, baby and a hectic work life, absolutely nothing has been done.

I hope someone may not make one of the numerous mistakes I have and/or will make to save at least some frustration and delay.

Current upgrade;
DD Wheel - inspiration from 'Cousin of OSW'

Future plans on the table;
Remake simrig - plan was to have easy change out accessories to go from drive to fly
Load-cell hydraulic brake
New pedal set as a whole
Seat paddles (gseat?)

Completed;
TBA

I may end up moving the below to another post - will see how it all pans out first

Having read as much as I could on a DIY DD Wheel, I eventually settled on the 'Cousin to OSW' style. Parts where a little easier and cheaper to source in Australia, compared to the '$450 MiGE AASD build' and even the 'DD Wheel with a Hoverboard'. It also suits my limited skills and equipment, though I have seemingly picked up a lot...

Components so far;
1x XYD-6B - Motiondynamics.com.au.

3x IBT2 (BTS7970) - eBay

1x STM32F407 Discovery digikey.com.au. Initial Nucleo board didn't work

1x 40mm fan. Going to swap for a 120mm fan for airflow through the box.

1x AMT103 incremental quad encoder. AEAT-6012 absolute encoder doesn't have the required channels.

Numerous wire etc. Settled on 14awg with ferrules to go between the IBT2s and terminal block, 12awg for the motor wires. Ribbon and 6 core cable as required. Lots of cutting and soldering. This will be neatened up once things are up and working.

1x male and female DB9

1x Switch

1x 12v Server PSU

Required tools;
Soldering iron
Extra hands with magnifying glass (Only a nice to have)
Side cutters
Wire strippers
Good set of crimpers
Pin punch kit
Drill press (highly recommended)
Dividers / Compass / Calipers (For the steering wheel coupling and adapter)

More details to come with pictures.

No pictures yet ... They will come shortly.

I finally got the motor pulled apart, tapped a thread into the back of the shaft. Turns out the drill press table had an angle so there goes the M4. Thankfully I has able to drill it out and rethread to an M5 (After resetting the table).

Welded some supports for the sim rig. Just the one cross brace in the middle, a place to put the big box of wires / STM32 and IBT2s and extended the mount for the wheel. The G27 were both too wide and not close enough.
After testing, the uprights for the wheel definitely need more bracing.

Also made up a wheel adapter - 10mm ID for the shaft with a 3mm securing hole. This is brazed to an adapter with 60mm PCD to suit the G27 during testing. Somewhere along the line it went slightly out of square and true. Revision 2 should fix this all going well.

Tedious process of connecting all the pins - they have to come off again soon to be able to feed them on the correct side of the interface box but tonight was just for testing.

Little bit of hotglue to hold the encoder onto the back of the motor and help with the shaft *note below*.

I had tested the motor before taking it apart through 'WheelCheck' without an encoder hooked up to the motor. Forces were quite good, as was the speed.
Today however, the feedback and force when turning the wheel left (against you) and when the wheel moves left by itself, it clicks and clacks, with a momentary null in any force. To the right works perfectly.

Hopefully something simple in terms of how the brushes and magnets are aligned and nothing that I have broken... If anyone has any ideas to narrow down otherwise I'll be opening it all up tomorrow afternoon.

*Sidenote - the AMT103 has a compression fitting. Suits an M4 size perfectly.

Secondary sidenote - Don't swap L / R pins when trouble shooting. Nothing good comes from it.

Well, one IBT2 board is dead. Definite magic smoke. My mistake there.

Further updates, opened the motor up and religned everything. Didn't seem out of sorts.

Now when I power it all up, motor only spins in the one direction. Swapping between PE9 and PE11 will change the direction. Speed changes every 10 seconds. Motor runs as expected when hooked up directly to the PSU.

Perhaps I have burnt out more than just the IBT.

In other news, sprang for EMC Pro. Runnings just the two IBTs now, however the perform much better with active cooling. Those little chips certainly get hot quickly otherwise. From MMoS to EMC leads to some different pins, coincidentally this fixed the movement issue entirely. Possibly messed up the PWM pins (PE9 - PE11?)

Motor gets terribly hot. Over the weekend I'll be opening it up, yet again, to drill some holes.

I did do another wheelcheck with EMC and it still isn't doing the left - right rotations but in-game, the forces are working as expected.

The plan for the weekend;

1/ Active cool the motor. Sheet metal for a tube, hose clamps to secure it and another fan

2/ Redo the wheel adapter. Bought some longer M10 sleeve nuts (need to change to M8 and drill out properly)

3/ Secure everything down, make the component box overall neater and mount the connector pins for neatness.

4/ Clean up the dining table - The wife has been understanding but luck may run out soon!

Finally got some photos.

This was done over the weekend, Monday through to today has been a mad house at work.

Wheel adapter was redone. It is definitely closer to centre and true (always room for improvement). The hole in the motor shaft seems to be a little bigger than an M3. Callipers show 3.1mm. I think I'll leave the M3 in there and file a flat into the shaft and then secure with a set screw. This will also, hopefully, have the added effect of removing any play.

Weekend coming I have a work trip but I'd like to get around to properly mounting this wheel as clamps only go so far... I found some old plastic containers that may be perfect to mount some active cooling. Should be a slim fit.

As I have a second server PSU, I'm going to try a 24v setup and see if it gives any difference in results (Purely for science). I believe the amps are the driving force for the torque but the higher voltage may allow more to be drawn.

I set the stop force to 100% also measured the amps coming off the PSU as well as going into the motor.
PSU - From 12.25 to 12.00V and 52A.
Motor - About the same amount of amps. Didn't measure the voltage.
The PSU is a 62A model so possible nothing is gained. It's a fairly straight forward conversion so will try none the less.

*Forgot to submit the post. I did end up putting that plastic box with a 50mm fan.
1/ Fan is too small.
2/ Handled 3 laps of Nordschleifer then moved on the motor, allowing room air in.
2.5/ Somehow have to secure this better.

I have decided against the 24v idea, especially for the time being.

Overall the wheel feels great - probably a little stronger than needed at 100% but it is just too fun! Come up with another temporary solution to the cooling as seen below.



Yes, that is a plastic flower pot that has been sliced and diced. I thought I had sheet metal but it was more like thin plate which would have taken more effort to come up with a mock design than I was willing to spend.

Took the motor apart again to drill / file some more holes. The back has 12 x 7.5mm (Groups of three linked together) and the front 6 x 10mm. I may still need more but the back is definitely running out of room around the brushes to keep some integrity. RC motors have big holes and small supports though so this may be an option.

A Corsair Sickleflow 120mm doesn't have enough pressure. It moves air but there is a lot of turbulent flow. Motor lasts a little longer before getting hot and cools back down within 5 minutes or so. I have ordered a Noctua NF-F12. I have tried both push and pull. Push ends up heating the steering wheel, that wasn't the best idea...

The wheel is still unfortunately only held with clamps. I'd very much like to sort the cooling issue first, as it lasts 3 - 4 laps of Nordschliefer or 10 of Bathurst. Once this is done, I can start experimenting with the 'Dampers' and 'Frictions' of the EMC software a little more.

New fan is helping a lot but it still felt like a $59 AUD bandaid.
Did some more reading into EMC and the 'Constant' Tweak had been bumped up to 120%. I don't know when I did that. Changed it back to 100% and the wheel gets warm but far from hot. I'm more than happy with that temperature and the FFB is still nice. Not as strong as before but if I'm being honest, it was probably more than necessary

This fan is loud though. I'll try some resistors to help lower the voltage of it which should quieten it down. I can see it will be a game of fan sound vs vent holes vs FFB. I'll find the balance sooner or later.

In other news, the steering wheel Quick Disconnect finally rocked up, that will be added somewhere on the To Do list. The flower pot needs to be replaced before any of that though.

On a side note, I am far from a drifter but the speed of this wheel far exceeds that of the G27. Very nice indeed!