Wood and Metal 2DOF Rig

Hi,

I am a retired computer engineer and hobby carpenter. I have been sim racing more than 15 years going back to Project Gotham 3 on Xbox 360 with an Xbox wheel. Over the years my static wood rig has evolved and now that I am retired I am looking to make the step to a motion rig.

I built a prototype with motion in mind and am using that as a static rig until I get motors and controls squared away.

Prototype

This is the completed prototype. I will rebuild it pretty much to this spec but take into account what I learned.


The monitor stand is the separate black metal at the front. The monitor doesn’t need to move since I race in VR most of the time.

The wheel support is the standard Trak Racer DD1 wheel support ($99). The 580mm inside width of the uprights is so that this will fit. While I was waiting for the DD1 supports I made wood ones which work fine but are not particularly photogenic. They are a lot cheaper.

The uprights will be extrusions on the next version. NOTE: When I bought them from AliExpress in March 2022 they were free delivery. Now delivery is $40.




This is the full extent of the Universal Joint. This is 22 degrees which is way more than I can actually use. I can move it this far in three directions. The pedals will hit the floor tilting forward but the height of the Universal joint could be easily increased. The black hunk at the back is a ButtKicker Advance.




This detail shows the connection. This is has no perceptible flex.. I plan on keeping the bottom parts wood. 2x6 are actually 140mm so I will rip (saw) them to 120mm. The premium planed 2x6s I will use are 135mm.

The cross piece is under the backbone so that it does not interfere with my feet.

Lessons Learned From the Prototype

Originally I planned to attach the motors to the wheel support cross member. This is a bit problematic due to access and interference. The DOF Reality system does just that but for me it seemed better to use the more common actuation from behind the seat. So the back bar in the design is specifically for attaching the actuators. The prototype doesn’t have the back bar just a seat support.

The backbone should be the best (flat straight not twisted) 2x6 available. I thought I had a pretty good piece (leftover from another project). It became obvious that even a slight “twist” is a big defect. For the next revision I will start with “furniture” grade clear pine 2x6 . These are a little bit smaller (135mm width) than standard 2x6 because they are planed. That’s OK the standard 2x6 is more than large enough. As of May 2022 these cost about $30.

Make sure that all of the pieces attached to the backbone are square. A carpenter’s square (framing square) is generally not square enough (well mine isn’t). If they are square the ends of the back board and wheel support are exactly the same distance apart on both ends. Next time I will precisely set those distances (<1mm). I will provide an illustration in a later post when I am doing the actual work.

The holes for attaching the seat need to be “precision” drilled. Precision means within 1mm distance between the front of the support and the hole. Misalignment of the seat, pedals and wheel will be noticeable.

Design

The design is based on measurements of the prototype. There are issues such as the distance between the seat, wheels and pedals that needed to be worked out experimentally. Those dimensions are pretty universal if you are using a car seat. The dimension of the U-joint from the back of the rig will depend on driver weight and equipment weight. The wheel, power supply, pedals, shifter and ButtKicker will impact the balance.




The stationary base is just the box frame from my original static rig. If all goes well I will replace this with a 3rd axis in the future.

All of the wood structural connections (black circles) are 3/8 inch bolts. 8mm will also work but here in the states 8mm tends to be more expensive.

For the connection between the backbone and the wheel four 3 ½ inch carriage bolts are used. This provides a smoother top since that is where your feet will be. The two bolts at the very back are also 3 ½ inch carriage bolts.

For the main seat support there is “interference” from the U-joint and Butt Kicker attachment on the bottom. This means that through bolts are not a good option. Instead I use these:


LQ Industrial 30pcs M8x25mm Furniture Screw-in Nut Zinc Alloy Bolt Fastener Connector Hex Socket Drive Threaded Insert Nuts For Wood Furniture Assortment - - Amazon.com


I use the 3/8 inch version on the wood frame. I use M8 for interfacing wood pieces to the wheel support extrusions. These provide stiff connections and easy disassembly.

These are also used for the shifter/handbrake at the back. At the front the connection will be to an extrusion so gussets are used. 8mm versions of these are used in the bottom board for compatibility with the gussets and corner supports.



Universal Joint Connection

The U-joint is connected to short 2x6s top and bottom. The holes must be drilled with a drill press and a template or some other means to get the holes precisely square with the U-joints.

These Yokes are made for an old Land Rover but were the least expensive parts I could find. I am not big on junk yard roaming. These provide 22 degrees of rotation as measured by iPhone.

Flanged Yoke - 262469P - Aftermarket | Rimmer Bros £10.75 (x2)

Universal Joint - RTC3690P1 - OEM | Rimmer Bros £9.20

These arrived in the states from Britain in 3 days! These are press fit and I don’t have a press. When I put the second one in I dremelled the holes out a bit and lubricated to get them together. The bolt pattern is 2 ½ x 2 inches. I used 3/8 bolts and screw in nuts to connect the yolks to the wood.

The wood provides additional height and clearances for the extreme movement cases so doesn’t add to the height above the floor. Once attached to the 2x6 they are easy to attach to the backbone 2x6 and adjust.

Controls

I was originally attracted to this site based on the smc3 control system. I have some experience with the Arduino and it was a completely understandable approach. I picked up some H bridges and a pretty hefty 24V power supply. I also had a 20W motor purchased for a TBD upgrade on another Arduino project.

With these parts I put together a motor test stand to allow me to understand how things work before I invest in real motors:



I have not been able to get reliable performance with this very simple setup. When I run the sin wave motion generator the motor will track for 10-15 seconds and then stop tracking. When the tracking stops I can see that the motor is OFF according to SMC3Utils. I cannot turn it back on. If I exit and restart SMC3Utils the motors will come back on momentarily and then stop. Seems like the motor moves to the target position and shuts off.

Here is the setup and example of the failure plot with settings:



Can anyone suggest what I might be doing wrong? I get the same result with two totally different pots.

This is the stock smc3 version 1.00 set to MODE2. I did make a change to the version just to verify I was loading what I thought I was loading.

Is the h bridge getting hot? Maybe a cooling issue?

Thanks for the suggestion! The h bridge is cool as a cucumber. There is a heat sink on the back and it only runs for few seconds.

The motor is OFF and the yellow line goes to zero. It does not seem to even be trying to send power to the motor. Trying to turn the motor back on does not work.

I started in thread in the Q&A section. Please post replies there. I will post the solution here once I have one.

https://www.xsimulator.net/community/threads/smc3-looses-tracking.17118/

My conclusion is that I have to be more robust in grounding and powering the external devices. Aka running pin wires to the H bridge isn't good enough.

I think with USB 3 you might be OK but with USB 2 it is marginal in terms of current. It's also possible that the H bridge I originally used was causing more significant power issues. These led to the motors "turning off" and also the serial communications breaking down.

Based on what I have read the Arduino is designed to use an external power supply even when USB is connected. So I will do my build with a separate power supply for the Arduino and the signal side of the electronics. I will use a voltage regulator off the 24v motor supply. Then everything will have a common ground reference on that board. The USB is just a serial port connection.

Thanks for all of the inputs.

Yes, it’s been awhile and I have successfully “completed” my 3DOF rig. The air quotes are because it’s never done it’s a process.

Let’s start with the obligatory look at my working rig video. I would like to preface that with a few notes:

• I have gotten tripped up by the fact that the camera is pointed toward the back of the rig. So the motions of the levers are "backwards".
  
• This is a 1962 Ferrari 350 which is extremely touchy to drive and tends not to go where it's pointed due to the 1962 suspension. It does make for a good demo.
• This track is Maple Valley which was part of Forza 4. It's not a real track because the hills would be too steep. Again good for illustration.
• There is a "traction loss" axis. You can see the rig move that way if you look for it but it's not tuned up yet. I have had a lot of issues here and only 2 working motors at the moment.

Looks pretty good but soon after I made this video two of the gearboxes had a problem. There is too much play between the shaft and the main gear due to wear. The wear is due to abuse while I was figuring out various issues. It was really slamming around and often breaking other stuff until I got to this point.