Showroom DIY cheap 6DOF motion sim

Hello, I would like to show you my cheap 6 DoF platform and provide some useful tips if you want to build one yourself. From to start I want to say that this is not a project for someone who doesn’t have access to a basic machine shop. What I saw in other projects is that they can be built with hand tools and a 3D printer. Unfortunately this project requires a lathe and a welder, these two tools contributed to it being so cheap. I won’t go into detail about how each and every piece was built, I just want to highlight some features which I think are worth mentioning .

Some backstory. This simulator was built for my engineering project. I’ve always wanted a 6 DoF sim rig and during covid I had a lot of time in my garage. It was mostly built from parts I had laying around or I could get locally. It was meant to be a prototype and proof of concept, but somehow (a long story) I set it up in a classroom in uni and it stayed there. From that point its being used (abused ) quite regularly, mostly by kids from other schools. Due to that I can tell you its strengths and weaknesses.

Since designing a 6 DoF frame has been in my opinion thoroughly described in a lot of other posts in this forum, I would just like to mention that the lower frame has a slightly larger radius than the upper one for a greater range of motion on the X-Y plane. Another thing I would recommend is don’t do it as I did . Recessing the cockpit below the upper mounting points of the actuators has increased stability, however it limits rotations of the Z axis and is quite difficult to get into.

What has worked surprisingly well is using a car steering shaft joint as U joints for both the upper and lower mounting points. One part of the U joint is mounted to a piece of steel flat bar with and tightened with a screw. The other part had a hexagonal hole for the original steering shaft. I ground down the welds and pressed the tube out. I grinded down a M14 hex head bolt so its head would have a slight taper and press into the hexagonal hole. A slight problem is that the actuators like to slightly rotate around their own axis which in turn causes the M14 nuts to get loose. Due to the platforms weight and its design it could probably operate without these nuts .


The actuators themself are essentially made from two pieces of square tubing. A smaller tube acting as a piston and a larger one acting as a shell. The way it operates is similar to how heavy machinery has adjustable stabilizer support legs. 4 slide bushings fit into 4 holes on the outer tube and another set fits into the smaller tube. Together they provide a mostly smooth and low friction sliding actuator. Overall this solution is quite simple and cost effective in comparison to actuators made from aluminum extrusions.



The one downside is that I used cheap T16x4 trapezoidal screws and nuts. As a proof of concept it was alright. In practice the nuts wore down very quickly. Additionally they gave all sorts of squeaks and noises during operation. I would highly recommend avoiding using these screws and just use 1605 ball screws like everyone else. I tried using bronze nuts, yet still there is a lot a of noise from the screws.


Moving on to another note worthy part of this build is the feedback loop. At first I tried using 10 [kOhm] multi turn potentiometers. They proved to be delicate and due to them not being contactless, they eventually wore down the most used part of their range. To overcome this I decided to design a somewhat contactless system. A GT2 belt is attached to the moving portion of the actuators and the DIY multi turn hall effect potentiometers are attached to the stationary outer tube. They work by converting rotary motion of the GT2 gear to linear motion by the means of a screw. When the screw turns the two nuts with magnets move from side to side. The nuts don’t turn because of the 3D printed slide which fits the profile of the hex nuts. Its quite simple and proved to be reliable. I only had one issue when due to several guest drivers thinking it’s fun to drive head on to barriers at break neck speeds, one actuator stopped working due to a magnet getting loose and falling off. From that point I used two part epoxy to secure these magnets.

On a last note another thing that works surprisingly well is the DIY 24v PSU. Its made by connecting 4 12V server rack PSU’s in series and parallel. To ensure that no PSU is getting any reverse voltage or back EMF, 6 diodes were also placed. One for each PSU going from the + and – outputs and two more going from their parallel + output connection to the series + output.

Feel free to ask any questions .

Very creative and cost effective design!

Why do you think the acme screws are so noisy? Is it backlash related and they make click noises during direction reversal? Or are they the loudest when runing at high speed?

I thought acme screws should theoretically be more silent than ball screws because the nut is gliding on a lubricant and there are no balls rolling on a ground surface. But with the low pitch of only 4mm the screw has to spin quite fast. Maybe there are double or even triple start threads but they may be hard to find. I think my 3D printer has 8x8mm acme screws with double-start thread.

Is the multi-turn potentiometer or hall sensor really necessary? An encoder on the motor shaft would be much easier. Of course, an incremental encoder doesn't provide an absolute position and would require a homing procedure at startup.

Thank you

I'm not entirely sure. There are several noises all of which come from the leadscrews and nuts. Backlash isn't really an issue since i guess that the weight of the upper platform keeps constant pressure on the whole assembly. The noised are the loudest during high speed movements and it basicaly sounds like metal on metal grinding or it somewhat resembles a sound that a seized wheel bearing in car would sound. It goes away for a while if I grease the leadscew, however its too much of a hassle to do on the regular.

Well i thought so too . In practice acme or trapezoidal screws don't work well with fast movements. As you say perhaps a leadscrew with higher pitch would require less RPM and in turn would make less noise. However as you also mentioned, it is hard to find leadscrews with custom pitches. Another thing worth mentioning is that a higher pitch leadscrew would probably lose its self braking ability, which comes in handy with racing sims.

Well, there are many ways to incorporate a feedback loop. Back when I made this simulator encoders were quite pricey or I had to order them from china. I had to do this project in a short time frame and didn't want to spend a lot of money so I had to use what i had. As you also mention the main advantage of my feedback loop is that it knows each actuatorors position regardless of whether its on or not. Nowadays I would definitely look in to encoders.

Yes, I fully agree.

Backlash is no issue in the linear part as it's constantly biased by the weight but the inner square tube could still rotate a little bit and make clicky noises when torque changes direction as you can't make the slide bearings too tight.

Trapezoidal/acme screws usually are self locking if the pitch is less than the diameter. It depends a bit on the material and lubrication. So 16x8 screws should still be self-locking but 16x16 probably wouldn't.

Using plastic nuts could probably help reducing the noise. On the other hand, plastic nuts are sensible to expansion/seizing when they heat up too much.

Yes, thats an interesting point. After analysis I did notice a slight "clunk" when changing directions. This sound varies from actuator to actuator due to, as you pointed tolarances between the slide bearings.

On a side note these noises aren't as annoying as the high pitched metal on metal sound.

Thank you for clarifying this. I coudn't find any high pitch trapezoidal screws, but there are quite a few high-helix leadscrews. However their price leaves much to be desired .

Yes, even with bronze leadscrew nuts there is quite a bit of generated heat withing the actuators. Maybe a material with low friction and some heat resistance like drylin from igus would be ok. Then once again, when cost is a factor a set of ball screws is cheaper than a trapezoidal screw with a drylin nut.

Yes, it's amazing how cheap ball screws have become. I recently payed only $11 for a 20x20 ball screw and $14 for the nut. The more widely spread 16x5 are even cheaper. The quality is... well "marginal", I'd never use them for precision applications like CNC machines. But they work and are always better than acme screws.