DYI G-Seat

[Michal Burisin]

So I finally got to the point, where I can start working on the G-Seat.

Here is my latest progress with the rig in general.

What's important for G-Seat part of the project is tactile installation on the seat, which consists of 3 BK Concert + 1 TST 239, so it's pretty powerful tactile, so I must make sure nothing rattles, this will be challenge as there will be many moving parts.

I'm using Kirkey Aluminium Seat, link is here.

I have 17'' version and it fits me like a glove I'm 168 cm, around 66 kg, quite thin with broader shoulders and it's so so, especially with the G-Seat paddles. I'm striving to take around 2,5 cm of space from the internal volume with paddles.

Here is the parts overview:

A) Bottom paddles - with big space between them - I don't want to simulate heave, maybe just a little, but it's intended mainly as pressure from the sides.

B) Main back paddles - narrower gap between to allow for some pressure during acceleration.

C) Additional back paddles for sides - just for thin guys, I made them removable to allow someone with bigger butt to fit in there

I spent day and half measuring etc., here are some images of progress:

Now when I printed it on A3:

And with additional paddle from the side:

I'm really happy it worked so well! I just have to correct 2 lines where the plate would be bent and that's it! I don't need to tweak the shape any more, it fits perfectly!

I will be probably using 4x ASME-04A servos for 4 paddles and 2 for seat belt tensioner.

The servos for bottom paddles will be located on red profiles like pictured below, on these profiles will be also servos for seatbelt fastener.

Servos for back paddles will be mounted to the seat.

The servos axles are goint to be extended by this connection - 8 mm to 12 mm, for 12 mm shafts I use aluminium hollow shaft with 2 mm thick wall.

The shaft will be like 0,5 m long, so at the end, it will be fixed in ball bearing with housing connected to the seat to prevent shaft bending.

Along the way there are going to be 2 points on each shaft with fixed clamp.

 

[Michal Burisin]

Brilliant work! Honestly, no on the GS-5 I can't detect any backlash...like you said the weight is always on the paddles, which kind of eliminates any inherent backlash in the system.

Really love the wrap-around style of your paddles, and as you say, the fact they're much closer to the seat shell than the GS-5.

I always wondered why small servo motors (as in DD motor style) aren't used with encoders. I would think even a small-fish one could drive such paddles and loads directly without the need for any gearing. Yes it would add cost, but on other hand would be practically silent in operation, which would be a huge win over steppers...it's my biggest issue with the gs-5.

That's great to hear about the backlash.

For me the steppers are easier to work with, as it's just 4 cables from the motor. With servo motor, you have to deal with encoder etc. and the drive has like 20 cables or something and I couldn't wrap my head around it That's why I chose the steppers, but other solutions like servo motors are much better choice for someone who can make them work.

I don't find the motor much loud, but I don't have any comparison. But it might be much louder when running 6 of them at once

 

[BenKay]

That's great to hear about the backlash.

For me the steppers are easier to work with, as it's just 4 cables from the motor. With servo motor, you have to deal with encoder etc. and the drive has like 20 cables or something and I couldn't wrap my head around it That's why I chose the steppers, but other solutions like servo motors are much better choice for someone who can make them work.

I don't find the motor much loud, but I don't have any comparison. But it might be much louder when running 6 of them at once

Understood! Yes it will be louder with 6, goes without saying

Noise of course highly depends on the profile used too. GS-5 on "out of box" profile is chattery and noisy as hell.

Mine sits on a motion rig though, so I just want lateral and longitudinal G type effects, that can have lots and lots of smoothing....that helped the noise dramatically.

 

[Michal Burisin]

Just tried motor with seatbelt tensioner and with lower leverage, this motor is quite strong I got to its limits, but it was quite strong pressure already. It would be better to have stronger gearbox to have some headroom though.

It's just temporary attachment.

Now I need to figure out how to hook it to load cell data, so it will be pressure based and not position based, I have something in the works, hopefully it will work out

 

[asteroulis]

I'd say it would be better to have the seat belt read the g's during breaking and acceleration. Load cell data is wrong because it can be applied even with a stationary car.
Move to see your build moving again

 

[Flag Ghost]

Michal would you mind sharing the STL’s for your upper seat brackets? They are beautiful.

 

[Michal Burisin]

I'd say it would be better to have the seat belt read the g's during breaking and acceleration. Load cell data is wrong because it can be applied even with a stationary car.
Move to see your build moving again

I probably haven't explained myself clearly, so more in detail:

There would be 2 load cells, one for each seatbelt. Load cells would be connected to the seatbelts.

I would know calibration point, when there is no pressure applied to the seatbelt. Then software would read telemetry - braking, acceleration, turning etc. and it would decide how much pressure it should exert on the seatbelts.

For example, let's say I'm going straight, so there is no force applied to the seatbelts. Then I start braking. As I start to brake, the software says the belts should start to gradually apply force. Let's say the software says it should apply 1 kg of pressure, I know from the data from load cell, that it is at 0 kg, so the motor starts to move in direction to tighten the seatbelt and will continue to do so, until 1 kg of pressure is applied and then the motor will stop and hold the load.

This value of 1 kg from example would be constantly changing and the motor would continuously move toward the desired pressure, if it would reach it, only then it would hold the pressure (basically just when the car would stand).

This would have advantage over position based approach as it would give more consistent and repeatable results, which would not be influenced by how much you tightened the seatbelt when you started. Also it would apply always correct force even when there would be paddle pressing into you and therefore increasing felt pressure from the seatbelt.

With this approach when the paddle would be pressing into you, the seatbelt would loosen so that correct amount of force is applied all the time.

It will be quite challenge to implement it, but I would like to try.

This idea came from @saxxon66 from one thread about seatbelt tensioners, if I remember correctly, to give him credit

 

[HoiHman]

@Michal Burisin, you are making great progress.

I think you are over complicating the seatbelt tensioner thing.

I have been testing the GS-5 with active seatbelt tensioning for the last couple of weeks and it's pretty awesome. Actually it doesn't get any better than that

However the main factor is how tight belts are at the start, the G-panels are almost a non factor. You brake in a straight line most of the time any way. Let go of the gas, the belts will loosen before you turn in and feel the G-panels

Seat belt tensioning is about travel, not force.

If you change sweaters without adjusting your seat belts, you will have a different experience.

 

[Michal Burisin]

@Michal Burisin, you are making great progress.

I think you are over complicating the seatbelt tensioner thing.

I have been testing the GS-5 with active seatbelt tensioning for the last couple of weeks and it's pretty awesome. Actually it doesn't get any better than that

However the main factor is how tight belts are at the start, the G-panels are almost a non factor. You brake in a straight line most of the time any way. Let go of the gas, the belts will loosen before you turn in and feel the G-panels

Seat belt tensioning is about travel, not force.

If you change sweaters without adjusting your seat belts, you will have a different experience.

I was thinking mainly about trailbraking when paddle and seatbelt would work together artificially increasing seatbelt pressure than what it should be.

But you're right, not worth the effort right now, I'll try to keep it more simple for now and put this on nice to try list for later

 

[saxxon66]

Michal, thank you for giving credit.

Atm our servo can only do position but they are able to be controlled by torque.
I need to do some experiments with this torque mode, but need to find some time.

The main advantage using torque versus position mode is that the belts will not snap.
This is the case if a belt gets too loose and then tries to pull hard and fast, it can feel like a wipe sometimes.

 

[Michal Burisin]

Michal, thank you for giving credit.

Atm our servo can only do position but they are able to be controlled by torque.
I need to do some experiments with this torque mode, but need to find some time.

The main advantage using torque versus position mode is that the belts will not snap.
This is the case if a belt gets too loose and then tries to pull hard and fast, it can feel like a wipe sometimes.

This torque mode is functionality of AC drivers or you still need load cell to read the pressure?

I was thinking about seatbelt roller to be on an shaft - small circle on the left. It would be mounted to plate (black line) with seatbelt roller and motor (big circle on right).

It would lie on solid frame (red), to which load cell (blue) would be connected and mounted above the plate with seatbelt roller.

When the seatbelt would tighten, the black platform would try to rotate along the path of green arrow, thus putting more force on the load cell.

But it would be quite complicated and the load cell might get lot of bumps from the rig moving on top of SFX100, so it would probably not work very well.

If the motor drivers can alter current / voltage to exert more / less pressure, that would be best of course!

 

[saxxon66]

The ac driver will work w/o any load cell.

We discussed the load cell set up earlier in another thread, I would mount them in between each belts.
I think it should work but I agree, it will be much more complicated.
You need to run more wires and do have some amplifiers for the load cells.

Using closed loop Servo w/ tourque control will be much more straightforward and simple.

So long the position based approach is working well.

 

[NeoGrotesk]

This looks amazing! I have an I have an NLR v3 and think this would really add to the immersion.

Is there an idiot's guide on what I need to buy and do to build a belt tensioner myself? (Keyword being 'idiot'... My technical knowledge is pretty limited here )

 

[Michal Burisin]

The ac driver will work w/o any load cell.

We discussed the load cell set up earlier in another thread, I would mount them in between each belts.
I think it should work but I agree, it will be much more complicated.
You need to run more wires and do have some amplifiers for the load cells.

Using closed loop Servo w/ tourque control will be much more straightforward and simple.

So long the position based approach is working well.

I'll put this idea on ice, position control is going to be great alone, thanks for input.

 

[Michal Burisin]

This looks amazing! I have an I have an NLR v3 and think this would really add to the immersion.

Is there an idiot's guide on what I need to buy and do to build a belt tensioner myself? (Keyword being 'idiot'... My technical knowledge is pretty limited here )

I'm still working on it. There is no guide how to do it yet, at least that I know of, I'm figuring this out as I go.

I'm currently testing with stepper Nema 23 so far with 4,25: 1 gearbox ratio, it's almost strong enough, but I would go for stronger one. Then driver is needed and some bearings etc. to put it together, also probably some 3D printed parts, but that depends on the seat and how would you attach it.

I wouldn't recommend jumping on this project yet though

I drive it through SimTools, but there is no official code for steppers for SimTools project. I'm trying to put something together, it works for 1 motor so far but it's not ready to be published, if it will ever be

 

[NeoGrotesk]

I'm still working on it. There is no guide how to do it yet, at least that I know of, I'm figuring this out as I go.

I'm currently testing with stepper Nema 23 so far with 4,25: 1 gearbox ratio, it's almost strong enough, but I would go for stronger one. Then driver is needed and some bearings etc. to put it together, also probably some 3D printed parts, but that depends on the seat and how would you attach it.

I wouldn't recommend jumping on this project yet though

I drive it through SimTools, but there is no official code for steppers for SimTools project. I'm trying to put something together, it works for 1 motor so far but it's not ready to be published, if it will ever be

Thanks for the reply, I'll keep following with interest! Good luck with it.

 

[Mr Latte]

Well Michael, man you've got balls to take this challenge on and stick with it.
Certainly not an easy task but the quality of work in the build to be seen and your determination are commendable.

I'm sure the process of it all has to be frustrating at times too but it could be great when its all up and done, you've come this far. I hope you find help and encouragement yet also a great sense of satisfaction from this as you progress closer to its completion.

No doubt still some more challenges and headaches ahead but wishing you well with the build.

 

[Michal Burisin]

So I'm finisihing bottom:

With BK installed:

And TST:

From the side:

Now it's almost done regarding flaps / hinging mechanism etc., yesterday I drilled last hole into the Kirkey seat , hopefully for a long time.

Now I will be preparing shafts and other minor details while waiting for the new motors and gearbox, hopefully last .

 

[Michal Burisin]

I got some progress on intended final layout .

 

[Michal Burisin]

First testing, just with keyboard and off the rig, but it feels really good already .

Sorry for audio, something wrong with my camera.

 

[Michal Burisin]

Back paddles done, I had to rework mechanism yet again so no belt & pulleys