Tutorial G-Seat where to start

Where to start. That was my question.
The legendary GS-4 seat from Simexperience is no longer made and still in demand. My project is roughly based on their design. But where do you start?
So this thread is my start.
The chairs design you use can vary wildly.
I would advise a wider chair.
Be prepared to pretty much destroy your new chair.
It is next to impossible to really know what you will find underneath that beautiful upholstery .
I am starting on the chairs back. This should give surge and sway.
The foam is removed.

I made the back paddles from 1/8" aluminum plate.
The bevels edges are bent at about 30deg and 20deg
Left is drivers left shoulder.

The paddles are attached to the chairs wooden back with 1"x1" x 14.5" angle aluminum 1/8" thick. The hinge material is called continuous hinge and is sold in lengths of up to 6' long.
I simply used the pre drilled holes in the hinge to attach to the aluminum angle and the paddles..

The servo pushrods are 4mm turnbuckles made of stainless steel .
I had originally used some very nice looking all metal aluminum 1/10 scale steering arms from an R/c car. However they fell apart very quickly and ended up being unuseable.
Pushrod link..
https://www.ebay.com/itm/4-Heavy-Duty-Marine-Grade-316-Stainless-Steel-Jaw-Jaw-Closed-Body-Turnbuckle-M4/253504518830?ssPageName=STRK:MEBIDX:IT&_trksid=p2060353.m2749.l2649


57186 57187

Here are a couple of pics of the real seat.

The GS4 seat used four 20kg Hitec HS805 servos per paddle.
I am going to use a single 260kg Chinese servos from WingXine.
These are a fast servo rated at .12 seconds rotation speed per 60deg. The faster the better....
https://www.ebay.com/itm/ASME-04A-H...h=item468a50fbb5:g:3GwAAOSwUlFao0cK:rk:1:pf:0

Just up front like to say that it's great to see another member having a go at building a G-Seat, as they really are a great addition to any simulator or even stand alone. Believe me, once you have tried one your sim feels 'dead' without it. G-seat are great for simulation and there is nothing wrong with basing it on the GS4/5 product which do a good job. Although I would ask you to take a spirited drive in your car on a racetrack or mountain road and really concentrate on what you are feeling around your body from the seat and forces in general. How well are these paddles going to replicate all those pressures and forces felt in a real car? Is this the best solution? Food for thought.

I know they are not and I have read your excellent articles posted here on the forum. The complexity of your system is awesome and I hope to replicate it some day.
The paddle based g seat is somewhat simple and can be fabricated with some basic hand tools. That is its advantage.
I have seen your work and I must ask if you own or work in a prototyping engineering production facility. I know someone who does and he produces personal project items to die for.
His company produces pretty sophisticated one of''s
Including really cool composites.
Cheers
Tom

Not really knocking the paddle system as I have used it and it worked OK, but stay tuned for my new G-Seat - you may be surprised by it's simplicity, on seat lightness and capabilities. I'm retired, but have a Defence Engineering background and a reasonable workshop.

I thought you might like to see into the 260kg .12sec /60deg @24vdc servos I am using.
The circuit board on these provide 5vdc for your logic or pwm device. Pot "B" controls the degrees of rotation. Pot "A" is a bit confusing. It might be speed but what ever it does just turn it fully clockwise.
I did run into a problem with these servos and thought I had received a defective one. The problem turned out to be the colored jumpers making intermittent contact with the pins.
It shows up as a total fail or worst an intermittent problem .
It might be best to just solder the pins.
The F/R jumper does control direction....

Just plain bearings and degreased for the pictures

I also picked up one of the 380kg/cm @.5sec/60deg servos.
The control board is not as nice and has no 5vdc output on it.
There is an extra gear in the train.
And it was $20 cheaper than the faster WingXine servos....

The Surge / Sway servos are mounted and I have adjusted the paddles for a total of 2" of movement measured at the outside of each paddle. I moved the pushrod attachment point closer in toward the hinge line by about 2 inches. This kept the throw the same while allowing the servos to only rotate 90deg making for a faster response time. I cannot stall out the servos so there is plenty of torque left.
The pushrods I used are not going to be strong enough and I have some new ones coming this week..

The surge/sway paddles are installed and adjusted. Overall I like the motion they present. I have only been running it with the NoLimits 2 roller coaster sim which makes it easy since I do not have to provide any inputs like you would if you are driving or flying. The roller coaster sim comes with enough different tracks to vary between extreme motion and a much more gentle effect.
The most impressive coaster for trying out the new system was where the coaster lays you on your back. It felt real although I was still sitting up..

The 260kg/cm servos provide plenty of torque. Even with me sitting in the chair and forcing myself back into the paddles they still move with great authority .
Presently I have them set to about 130degrees of rotation. According to the specs this should give a full travel response time of 250ms . I plan on measuring this in the near future with an oscilloscope and a photodetector.
I think it will be interesting to see how fast they really need to be....

I bought one of the vibrating car cushions off E-bay.. 10 motors no less .
It has a zippered back so pulling out all the foam and vibrators was simple.
You are left with what appears to be a very durable cover with a thin foam backing. The thin backing contains some carbon fiber tow used as a heating element. Its totally ineffective but the foam coating they used to seal it is near perfect for just enough padding.
I also bought a low cost automotive car seat cover.
No padding but still works great. Cheap and simple.
I am ready to start on the heave modification next.
It will be very interesting and I am sure more difficult than the surge.

You are really making great progress. It's good to hear that your surge/sway paddles work. Seems that these servos are the better choice compared to the Super500 I have (or 'had', because I am currently changing to bowden cables for the heave paddles after burning one of the servos).

About the heave paddles: See here for some considerations about the distance between the paddles. When you put them closely together, they have to lift your body, which needs quite some power. If you leave some space between them, they rather squeeze your body and need less power.

That's right and a good idea. I did this on my first rig (see the first link on the page referenced above). At the moment I am using bass shakers with GameVibe, but I was already thinking about adding the rumble motors again. I am still using fans for getting an air stream into my face, and I just need to connect the rumble motors in parallel to the fans. It gives you a humming vibration that you can connect to any Game Dash output, like speed or RPM.

Thanks for the tip. I ordered some stuff to get the heave paddles mounted.
I am going to use 2 of the 260kg servos per paddle. Yes I cannot imagine the servos lifting my fat ass so we will do the squeeze.

I looked over the game dash thread .
It appears to be dedicated to running a fan for the most part.
I ordered some monster motor shields from Ebay so it will be a few weeks until I can try out the vibrating motors.

I did hook up some transducers . The main one is a Clark Synthesis TST239.
However there are no game vibe plugins supporting DCS .
And when I play PCARS2 the transducer just did not do much.
Possibly I just have it tuned wrong.

I ordered a motor setup like @Grigory
Just one to experiment with.

I hope to make some sort of video of the G-seat operation very soon.
and will post it here.

Many Thanks
Tom

I set up the scope to measure the speed of the 260kg .12sec /60deg @24vdc servos. The setup is a Rigol DS1104 scope, a photocell (glued on to the servo wheel) and a 5v 1mw red laser diode for the light source.
The servo was not under any load.
The servo was adjusted for 130deg rotation with 900 to 1900 millisecond pulse width.
The servo traveled in both directions (total of 260 deg) to make one measurement.
The result was very close to the specifications posted by the seller.
300ms to rotate 130 deg and about 140ms to do 60 deg.
The servo does want to see the full 900 to 1900 pulse width no matter where you set the degree of rotation. Otherwise you will lose a lot of speed.
The sketch I got from Spit40 does output the full PWM range.
I did set the Pulse width parameters to max.
If you need to lower the rotation use the "B" pot on the servo
If you need to adjust centering that will have to be done with the mechanical linkage. (Centering is really a not the correct term since with the G Seat you never use the the center point. Only one extreme to the other.)
The sketch does have 4 servo outputs R L S T
In the game engine you will have to manually enter the configuration.
Interface type Serial
Com Port Check your device manager
Baud 9600
Data Bits 8
Parity None
Stop Bits 1
Output bit range 8
Output type decimal

Interface Output R<Axis1a>~L<Axis2a>~S<Axis3a>~T<Axis4a>~
NOTE: the lower case "a" assumes you are using interface 1
If using interface2 change the lower case "a"s to "b" s

Output rate 10ms
Arduino Uno used
Pin outputs R is pin 4
L is pin5
S is pin6
T is pin7
If you are using the ASME-4 servo there is no need to supply +5

You will need the servo library to install..

I am preparing the heave installation. How I wish I had a real drawing of the GS-4 Seat to work off of. The best I can eek out of available videos is the seat is 17"wide. This pretty much puts the paddle width at about 6.5" total width. Inside bevel 1" and bent down 20-30 deg. Outside bevel .5" and again 20-30 deg bent down. I assume this allows the paddle to slide over the cover easier and maybe more comfort to the pilot.
So this will be my starting point.