Seat Harness Tensioner Questions

@ferslash - thanks! This is a box for a smallish and bare power supply with the JRK up top and attached to a heat sink with the lid acting as a clamp. I did another design for the JRKs that has stronger cooling here: https://www.xsimulator.net/community/threads/jrk-12v12-overheating.10327/#post-133655

I'll post a build thread when I get it all done and tested and hopefully it will make sense. I keep finding applications for 3D printed stuff. It is way cheaper and faster than getting things machined and has been a lot of fun to boot. I've never been able to go from design to part so quickly or easily. I've always had to draw things up and either send them off or wait for our own machinists to be able to get to whatever it was.

Finalized the rope routing to get straight pulls (as straight as possible) on the belts as well as get symmetric pull down on the universal joint seat post. It's a meeting scribble...



The pull from the motor goes through a hobby horse spring and a turning pulley to then split into two to turn and go vertical at the front and back of the seat post. The rear rope just turns at the top of the seat post to the back of the seat to pull directly on the shoulder straps. The front rope turns at the top of the seat post to go forward to turn back towards the seat post. A line from each lap belt crosses each other to hit their respective turning pulleys to go back forward of the seat post so each can be pulled symmetrically by the rope that was forward of the post and turned back at the front of the seat.

I know this sounds and looks complex but it divides up the motor/spring pull pretty much equally among all the belts, gets pulls in line with the way the belts lay around the seat frame, and keeps all but a few ounces of weight off the seat frame. The motor, feedback, and power supply all ride on the lower, non-moving part of the frame.

With an approximate 50 lbs maximum pull from the motor and arm over a two inch throw, and a hobby horse spring rated at 48 lbs max, there should be 12 pounds of pull on each belt neglecting friction. I got a pack of 10 fidget spinner bearings (using 8) to 3D print ball bearing pulleys and made seat belt guides that clamp around the belts and hopefully reduce friction as the belts turn around my seat frame.

If this doesn't make sense, it should get more clear once I get the last of the pulleys printed and installed. It looks tits.

You have to love all the different ideas that different members develop on this site. @Zed - You could also use a 'Bowden Cable' setup like mine .

Maybe a stupid question, but i have to ask it. It's some years ago, that I've driven a car with 4+ point harness, and back than i didn't focus on the forces, i can feel while driving.
So isn't it more realistic to pull each belt separatly? So if you're braking in a right turn, you feel more the right shoulder belt and vice versa?

@SeatTime - Dude! That was staring me in the face and I missed it! Excellent idea and you are 100% correct about different ideas. My 3D printer uses a Bowden tube to feed filament and the old VW's I worked on back in high school used them for the throttles. Excellent solution!

I was wondering how you would do something similar without pulling against your actuators and hadn't solved it, but you had. Props!

Not a stupid question at all. You probably are correct but all of this is approximation anyway. It's not really practical (though would be possible) for my setup to pull each belt separately but @SeatTime certainly could do that with his Bowden tube setup. That would also require an actuator for each belt, though (stating the obvious) and would get expensive quick.

I haven't even tried motion yet to know how convincing it is with a seat belt tensioner so no idea how much individual pulls would add to the effect or how much more realistic it would feel. Bet you are right that it would be more realistic.

Thanks, great to hear, that you think the same
I think, that with a G-Seat pushing one side of you against the belt in turns you can feel the difference on the left/right side?!? @SeatTime should know this...

Another thing, i don't quite understand is how you can put the motor etc. on the nonmoving buttom and connect it to the ropes on the seat, WITHOUT moving the ropes through the Motion of the sim? Is the Bowden Cable the solution?

This was exactly my thoughts when both @SeatTime and @Zed were talking about seat tensioners and positioning the actuator away from the platform. How can the actuator stay stationary without inadvertently tensioning the ropes as the platform moves from one extreme to the other? But now I realise that if using a Bowden cable this becomes possible, excellent idea!!

The Bowen cable idea came to me when I was trying to think of a way to pull my body around in the seat without bolting big motors to it, (I just wanted to see what you came up with - without polluting the waters with my idea ) as I went to allot of trouble to make the whole rig light. Although it may sound simple, there are allot of things to get right for it to work properly - even down to things like ease of entry and exit. I've had to make allot of bespoke parts . FI - My harness tensioner already uses independently controlled servos for each belt.

The Bowden cable approach does eliminate the motion platform motion from influencing the seat belt pulls, but the way I route the ropes and where they split does too.

The one rope splits to two before one rises up the center of the front of the seat post and the other comes up at the center of the rear. There is definitely a downward pull on the seat post but it's evenly split front to back so should balance.

Up in that diagram where the one rope splits to two and it's annotated "Pull/2" on each cord, if that connection is a pulley then the platform tilting front to back will just roll cord through the pulley with the tension being equal on both cords.

If this still isn't clear, I'll have more info this weekend when I get it all laced up. Weekdays are hard to make much progress.

@SeatTime - that's very cool! So with different servos are they truly controlled separately where forces are different in turns and such too or ganged together on the same motion signal? Did you use a servo per belt to get adequate pull or for that differential effect? About how much pull do you get per belt now? And lastly, since you said they are independent, if I am interpreting that correctly, how different is that effect and feeling from just yanking on all belts equally?

I guess if it's a big difference, I can just use deceleration to activate my tensioner and leave off side forces.

Damn, dude, it sounds like your platform is a true Ferrari of platforms! Envy...

So, the bowden cable is a solution for lots of Problems! @SeatTime you should rebuild your g-seat with bowden cables, you could save some moved weight
Seriously... It's really great idea, be to move the seatbelt and the paddles of the g-seat from the bottom platform, so you dont have to look at the weight of the servos and buy cheaper stuff

For what I can now purchase big servos for, there is little difference in cost once you add up a Motor/Gearbox/Controller/driver/powersupply. What I do get out of the big DC motors is speed and torque beyond what the servos can produce, which is required for parts of my G-Harness system. The servos are although more straight forward to setup and quite adequate torque wise for G-Seat Paddles and Seat Harness Tensioners. FI - I feed Surge and Heave data to my seat harness tensioners - only the positive excursions (controlled through some Arduino code I wrote) for braking and negative vertical Gs, although this will change once my new G-braking system is on-line.

Made progress today. My seat tensioner is all done except for the motor output arm and works great with manual pulls. It's still in two parts and I'm up to a full compliment of 10 pulleys but the tension is even and friction is low.

Here's a pic of the bottom of the seat with the two tensioner ropes coming towards the camera that would travel down front and rear of the seat post if the top frame was mated to the bottom. The top of the image is the rear of the seat and the tensioner rope just ties to the shoulder straps and turning pulleys aim it down along the rear of the seat post.

The lap belts have turning pulleys to pull them straight and there is a partially hidden pulley that pulls on that loop in front of the seat post, with another turning pulley to route the tensioner rope down along the front of the seat post. The front and rear tensioner ropes just form a loop where another pulley rolls along that loop (to pull evenly) with its rope turning to then pull through the hobby horse spring.

Complex but I think for this sim it might work better than Bowden tube pulls. It was a bit of extra work to custom make the various pulleys though but that was half the fun and they fit the application perfectly by doing it that way.

It's been slow but getting closer. Had to make some revisions to the power supply box/mount and the mounts for the Hall sensor and sensor magnet. But it's looking like it should work well now. The Hall sensor mount is printing a revision so I'll add it soon.

And the Hall sensor mount. Just need the power connections to the motor and this one will ready to calibrate.



The little red magnet mount in the post above doesn't hold the output arm down. There is another 6 mm nylon lock nut under the red adapter. There was no thread left above it though. The output shaft threaded part was flush with the top of the lock nut. The red adapter has a hex opening in the bottom of it, just big enough for a snug fit on the lock nut. There is a small lip at the very bottom of the nut cavity to grip the filleted edge at the bottom of the nut even though there is no force or contact between the magnet and Hall sensor.

Nice, looks like it is coming along fine. FI It always looks good on paper, but every time I have used springs in whatever system on the sim, I have ended up removing them - as they just seem to mess everything up in regard to the system tracking the simulation data. As far as tensioners go, you will be surprised how much compression/flex you will get from the rest of the system (including your body) rendering their use academic. All I have done is set a limit so that the tensioner does not crush me 'too' much.

Thanks. It's a pretty stiff spring though. If it adds too much flex, I'll take it out since it's in the main pull rope and just use a slightly longer rope. Not really much flex in this configuration, though. The ropes don't really stretch and the turning pulleys don't give any at all.

@Zed : very creative! congrats


my first thought when I saw it

Hey Matt, I still haven't mated the top platform to the bottom to see how it actually works other than a test fit, but I think it should work ok. The top seems to work ok and the bottom too so it has a good chance. The pulleys are all ball bearing for low friction and it should pull fairly equally on the straps. But like some of the other stuff I'm doing, it's not tested yet. Will report soon. Just about ready to combine top and bottom and fingers are crossed it all goes as planned.