Showroom SRT80 3DOF with dual traction loss

I previously built a 2dof rig with gear motors and was very unhappy with the backlash. I decided to start somewhat fresh using actuators. I settled on SRT80's. Unfortunately, when I do something, I can get a little bit out of hand and ended up with 7 of them. The rig consists of 4 vertical actuators for heave, pitch, roll etc. and 3 horizontals. 2 for front and rear traction loss and 1 for surge.

The rig is fully built and functional. I've moved into a stage of trying to tweak the settings to make it work the best and would be happy to hear any thoughts on axis settings that might help with the experience.

Here's the current finished product:

I designed and 3d printed many parts for bushings, spacers and POC prior to obtaining real parts. The traction loss design is an adaptation from lebois. I bought similar parts but designed and 3d printed my own pieces to mate the parts all together.

Here is one of the TL apparatuses.


I also designed and printed my own e-stop control for the Thanos controller.


The POS22 rod ends were too long and the ball screw's would run into the ends. I have jam nuts but I wanted to keep the thing as short as possible as it's already about 6' tall and my head almost hits the ceiling. I ended up drilling relief holes in the ends of the joints to allow the actuator fully retract.


Dump of some additional photos from the build process.


Here are my current axis assignments. Very open to any suggestions you seasoned pro's have:

I like your Rig, the only thing I don't like so much is the very high build. If you mounted the slides on the side, you would save a lot of height and the whole rig would be more stable.

I appreciate your thoughts. I also do not prefer how tall it is but spent a fair amount of time trying to find ways to reduce the overall height without spending a ton of extra money.

Which frame are you suggesting rotating out of curiosity? I figure most of the forces are worst in the vertical axis. The frame is strongest vertically as it is. The actuators are plenty strong enough to bend it. Don't ask how I know

Your Rig is similar to mine, you can see it in my project threat.

Your project has definitely given me new motivation to start my Tractionloss Project again. The entire structure of the Tractionloss will be a maximum of 15cm with a very simple but very similar structure. However, I will not be working with linear rails, but with solid shafts and linear bearings. This simplifies the construction considerably as different directions and angles can be compensated. I started on the new design yesterday, maybe I'll have something presentable ready by Sunday.

Can the feet of your vertical actuators compensate for the lateral travel when tilted? That would be important to avoid transferring unnecessary load to the actuators. The force is then also transferred to the frame, which can bend.

Will watch your build! (Actually I already was. I remember asking you about the linear slides which got me going. Your build helped me get moving. Thanks!) I looked at linear rod but for whatever reason preferred liner rails. Maybe someday I'll change up.

Are you referring to the rod ends? They were carefully shimmed so that they can tilt side to side more than the maximum needed if the rig is hard rolled one direction or the other.


The only "stability" issues I have had is that it will jump the whole dang thing off the floor if you aren't careful. I have the Thanos dialed in to pretty much eliminate hopping due to large fast motions. The whole thing weighs about 250lbs, or about 400 with me sitting on it. It's POWERFUL.

My mishap was with the Thanos. Either a bug or user error, it kept reverting to the wrong actuator configuration. Then at boot, the four corners went opposite directions and twisted a bunch of crap and broke some of the 3d printed traction loss parts. It was a sad day. Only one aluminum beam was bent and I was able to straighten it. Most of the rest of the mess was just all the brackets racked in the channels so it was all held twisted. I had to basically tear it down and rebuild it in place. Needless to say, I power the Thanos first then the actuators and am ready to power off the actuators if it's torquing. Hasn't happened again.

Had a similar issue with the Thanos controller. This is why i don't prefer a solid connection.

Simhub has the best solutions about filters and limits. SRS was pure crap. Had the same issue and i started to lower the seatposition as low as possible. Look on your rig from the back and think about the centerpoint of weight when your vertical actuators tilt from one side to another. Also it will benefit your immersion.

100% agree on the solid connection but boy, that thing jumps and I just can't see it not launching itself out of those little cups and having a much worse experience. First iteration was just the rubber feet without the traction loss. It walked across the floor multiple times.

Yup, it's heavy up high but that hasn't been an issue in tipping/listing. I'm not sure how lowering height would change immersion. The pivot is still as low as it can go for the primary actuators being all the way down on their bodies. I want the cockpit to be more like a regular car, not f1 so that sets the seat to pedal height.

Perhaps I'm not following your point.

I could have imbedded the surge platform inside of the mid frame but it would have required me to buy a bunch of different sizes of extrusion I didn't have. If you hadn't followed my journey, I started with gear motors and a u joint but was unhappy. I tried to build this iteration without throwing out too much of the metal I had already bought.

Good to know on SRS. I had wondered if I would give it a try some day. I'm still using xbox so things like haptics are waiting for me to build a PC.

Your rig is really well built. It just has a small but possibly dangerous design flaw I've seen many times: A 3DOF rig with 4 actuators in the corners is over-constraint. This doesn't have any bad consequences if the actuators just sit on rubber pads on the floor. If anything goes wrong and the actuators get misaligned one of them can just jump off the floor and won't damage anything. But if they are bolted to a fixed frame with ball joints they will bend it.

I wonder why we don't see 4DOF rigs with the configuration like the D-Mover in the ad shown in the top right corner of every forum page more often.

IMHO this is quite clever. With both rear actuators moving in the same direction the act together like one in the heave direction. If they move in opposite directions you get rear sway aka traction loss. And you don't need an extra actuator for that.

But you already have traction loss implemented with two horizontal actuators. So I'd recommend getting rid of one of the 4 vertical actuators. If they are strong enough (I guess they are) just place one of them in the middle behind the seat. Alternatively, you can use this configuration which has a better load sharing.

But be careful. Even with only 3 actuators there is still some bending. The ball joints can "absorb" the angular deflection of the pushrods our of straight vertical position. But when rolling or pitching the actuators themself have to change either distance or angle slightly. If you force them to stay parallel by bolting them to the upper frame the result will be excessive bending forces on those bolts.

Look closely to the picture of the D-mover. The rear actuators don't have ball joints but hinges with an axis in longitudial direction. This provides a stiff connection in the surge direction while allowing roll and sway. It's not clearly visible but I guess the front actuators have hinges with laterally oriented axes. This allows for a small movement in the pitch angle while being stiff in the roll direction. So everything is held in position but not over-constraint.

Look at your Center of weight and where your Actuators touching the floor. You get an very high Triangle and on the picture you are almost over tipping point. Your whole weight is on the right actuator. When you tilt it to the other side, on the left actuators will be no weight and you get serious dangerous weight movement. Like you would have a really high car, it will not like any load change from left to right and in the worst case you will flip over. You have to slower your max speed of the actuators to compensate this. I had almost the same issues and yes, also my rig took off completely a few times and my rig is significantly lighter, which makes the whole thing more dangerous and more susceptible to load changes.

I like the D-Mover a lot, maybe i would build a similar one when i would start a new one from scratch. However, with the rear actuators you will very quickly have a problem with the working radius. Theoretically, they need twice the working distance of the front actuators, but the basic length then becomes too long. You would therefore have to design the actuators in a fundamentally different way to minimize the length and increase the working length, which quickly makes the actuators twice as expensive.

You're not wrong. However, this was purposely pushed into this position with a bogus axis assignment to show the rod ends will not bind. The axis percentages wouldn't actually allow it to ever be like that. Max it can roll is 50% which is 20% roll and 30% sway. The rest would be heave/pitch/surge which wouldn't [directly] contribute to roll. As an aside, when it sway's, the dual traction loss moves the opposite direction which brings the center of gravity more over the center of the bottom frame as an added bonus.

I'm not sure there is much that can be done without basically starting over. The srt80's have a lot of travel so they are much of the height. Having a surge frame and a traction loss frame adds 6+" of metal. I could turn the bottom frame on its side and let the floor be the stiffness to get ~2" but not have much to mount the TL actuators to. I could widen everything and that would reduce the roll degree.

For now, it feels plenty stable short of the front jumping if you don't limit rate. I think I'll just enjoy it as is. Maybe I'll build something new in the future.

After about 3 months of use, I've been underwhelmed by the sensation provided by the surge axis on my rig. Basically it acts as a seatbelt tensioner and doesn't do a great job at that either.

Would I have better results if only my seat moved slightly instead of the entire platform with wheel, pedals and seat?

Or should I ditch that whole platform and convert my actuator into a seat belt tensioner?

How big is the stroke of your Surge?

Also it depends on where you have attached your seatbelts? If it is like on your last picture, it can not work properly.

Can you please post a picture of your current configuration? The last I saw were these:

... in post #1 and #5. But the surge actuator is missing. Are the frame-bending problems solved?

So now you have 6 actuators but not 6 DOF, right? A complete re-build converting it to a hexapod would be way too much work. But I have an idea of how to mount a 3DOF rig (pitch, roll, heave) onto a flat slide on ball rollers. That would add onother 3DOF (surge, sway, yaw) so you'd had true 6DOF instead of just 3DOF + traction loss.

Look at the very first picture.

All 7 of my actuators are SRT80 which are 220mm stroke.

I've seen multiple opinions on which proper way surge is supposed to move. I settled on the platform moving backwards with breaking to release pressure on your back. The cord attaches to the belt, passes through a pivot and then attaches tot he mid frame. When the platform goes backwards, it pulls on the belts. As such, it technically "works", it just doesn't really add much. Tugs on the tops of my shoulders a bit. Doesn't feel anything like hitting a true belt.

I'm suspicious that a dedicated tensioner would feel similar as it's just pulling down on my shoulders. Difference being it could be better controlled.

In so much as the Thanos hasn't gone wonky again, sure, it's solved Redesigning the entire rig to have no over constraint isn't something I was ready to do. If I find a better design, I can somewhat easily change to that doesn't have over constraint, I'm all for it.

I have 7 actuators. The way I see it (could be wrong), I have pitch, roll, heave, sway, surge, yaw (although I don't use it), TL.

The 2 bottom actuators can move the entire setup left or right a total of 220mm. I use it for sway a bit and traction loss.
The 4 mid actuators give the pitch, roll, heave (and surge representing the compression of suspension).
The Surge actuator tries to simulate surge.

I don't think I need yaw? Yaw is the direction component of the vector right? Unless my rig can spin unconstrained, that doesn't seem doable. Maybe I'm misinformed.

I would be plenty happy enough to change the mid frame down to 3 actuators to solve the over constraint but with the current surge platform, there is no easy way to do so as the sliding surge frame (and my feet) would hit an actuator directly off the front or back of the platform.

Traction loss is some sort of coupled yaw + sway. For example, rear traction loss during a left turn sways to the right with a bit of yaw to the left at the same time.

So with a different configuration you could do the same 6DOF but with only 6 instead of 7 actuators. But because you already have them there is no advantage in optimizing this. It would only take many hours of work and wouldn't give you any money back... So if it works I'd just leave it the way it is.

The actuators are quite long for the stroke, I estimate something around 700mm. Building a hexapod out of theese would be quite difficult, I think.

I think this is basically what I have created with my lower frame.

Front TL: TL 30% (inverted direction), Sway 30%
Rear TL: TL 70%, Sway 30%

I measure 66cm fully collapsed including the ball joint so you're pretty close. The other issue is there is no easy way to have a pivot at the opposite end of the actuator as that's the servo.

It sounds like your Softwaresettings are the issue. If you don't mix the directions of movement, you will get a better feel. For example, for Surge, let the Surge actuator do this compeletely.

And adjust your sim's maximum stroke and reduce the filter. Shifting become relatively extreme, but these can be filtered out separately depending on the software.