007 G-Seat: Ultra Compact/Servo Based

Really thinking I need more power (Oh, and Oculus Dash is fantastic for SimTools tweaking)

I'm seriously thinking of buying my 4th pair of servos. The bill for this thing just keeps going up. I've just been flying in FS2 and it was bothering me a bit that the heave flaps seemed to be moving inconsistently. I measured them with no load and they were exactly symmetrical, but then I paused the game in VR, went into Dash and opened up Output Testing in VR and experimented. Basically the right heave flap is stronger than the left, or I'm putting more weight on my left leg. Either way I can be on about 70% and the right flap feels quite a bit higher than the left and only feels level when I increase heave to 100%.

So, there is one more Wingxine servo I can upgrade to, the 380kgcm. Same dimensions as the others but it has extra gearing so has more torque but only moves at 0.5s/60 deg. I didn't want to go with the slower model as I wanted quick reactions for a nice thunk on landing, but I've now given up on that and will use the 3DOF to supply that.

Here's the 380kgcm I'm looking at. https://www.ebay.com/itm/302541117345

Just one thing before I do - upgrade the cable from 0.75mm to 1.5mm to get an extra 1.5v or so to the servos.

Heave still needs improvement
I'm really pleased with the surge component of this G seat, but in comparison there's something lacking with heave. It is more difficult of course due to the extra power needed, but i'd like to invite comments from those who have built G seats so far on two aspects of My G seat.

1. Thigh lift v. Whole body. I'm seriously considering doubling up on servos so i can apply pressure right across the seat, thighs and backside - British fellas don't really say 'butt' ;-) it takes more power, but would it be more convincing?
2. Seat width. One consequence i now realise of converting a normal seat is that the seat cushion i'm lifting is relatively narrow. Once the flaps start lifting and closing in this is especially so. I have to place my legs carefully in fact to ensure even pressure. I may well get a better sense of heave if the flaps and cushion were wider and so gave more side pressure

Any thoughts from those who have experimented with gseat configurations? @MarkusB @early_m ... who else is in this niche club?

I've only used my setup for sim racing so can't comment on the flight sim side of things. But I use flaps under the thighs only and feel this is substantial enough to emulate the braking g's. IMO when you break in a car you role forward and your arse comes off the seat and the backs of your legs go into the seat, so I think this setup mimics that.

A further idea I've had to increase this effect is to also make the wheel move towards you when you brake, as in real life you push on the wheel during braking as your body is thrown forwards. A kind of g-seat for your wheel if you like.

I was also not convinced by my seat flaps. At the moment I am rebuilding my rig from scratch, and with this second approach the seat flaps will only apply pressure and not lift me. I am focusing on flight simulations where I need a longer lasting heave force rather than inducing bumps to my body. (For bumps and vibration I will probably add a bass shaker.
Unfortunately I cannot report from any positive experiences yet, and I will probably still need some weeks or even months due to other work in progress.
By the way: I will give the SUPER500 servos a try. I already got them from China, but did not have the time to test them.

Working on plans for V2

Great feedback guys. The partial flap/full flap thing is very interesting and I hadn't thought about the different requirements for cars v planes.

I've decided to rebuild my G-Seat with these changes:

• Somehow squeeze a second servo in on each heave flap - I'll pair the 260kgcm with 380kgcms so I have 640kgcm per side
• Full flaps, not just thighs
• Move the hinge point in about 30% on each flap so there's more squeeze than lift
• Try to widen the seat cushion/flap area

Having already been down this road, It's interesting watching you guys going through your G-seat development process, I wonder if you will come to a similar final conclusion as myself? Time will tell .

@SeatTime: Yeah, inspired by your posts I am indeed thinking of a bowden cable solution, at least for parts of my new g-seat. My current thought is to still have the backrest paddles for positive acceleration and sway, and to add a bowden cable that pulls my upper body forward against the seat belt for negative acceleration (breaking). This would be a replacement for the belt tensioner solution. I already browsed the web for a suitable harness for mounting the cable to my body, and I came from motorcycle armor protection vests via tactical vests and climbing belts to medical back support braces. Still don't know what will work. That was also the reason why I was asking you for a hint regarding your harness solution some time ago. Well, it seems that we have to reinvent your wheel.

I was in Australia visiting relatives earlier this year. I should have pestered you to let me come see your rig close up.

It probably still made sense for me to build this way though. I build like i write software - incrementally, fixing mistakes as i go and evolving my thinking.

What I can tell you is that paddles can only go so far pressure wise, they then start moving your body in the wrong direction. Eg. Your going into a Right turn - with paddles you would apply pressure to the left side of your body, unfortunately if you apply a realistic force it will start to push your upper body also to the right, which is not what you want.. (it may also push your body into the other seat bolster, or paddle). If your after realistic big Gs like me, then they just don't work. Also a well fitting seat works best - doesn't matter how you do this. FI, My body harness is hand made by myself for the job - I could not find anything available commercially that was suitable.

Are you saying there is another approach that gives g-forces? What is it?

I moved away from G - paddles quite some time ago. It is mentioned scattered through my build thread.

Here is a brief summary of what I still have in my mind from reading SeatTime’s posts:
SeatTime uses bowden cables that are pulled by strong DC motors. The motors are not part of the moving platform, so that they don‘t add any weight to it.
The other ends of the bowden cables are connected to a self-made harness.
At first, he had trouble because the surge cable (the one pulling your upper body forwards and backwards) was interfering with the sway cable (the one pulling your upper body sidewards).
But somehow SeatTime has solved this interference problem.
Besides there is (or was?) a cable connected to a belt that goes around your legs/knees, also pulling them sidewards and causing a sway force. You can find a video of this belt (without legs inside) in his thread.

Thought I'd share the two arduino scripts I wrote for

1. Testing and setting the limits of movement for my servos
2. Running all 4 Wingxine servos talking to SimTools on pins 4,5,6,7

Hi @Spit40,

At the moment I am in the progress of rebuilding my g-seat from scratch. It will again become part of a 2DOF simulator.
2 remarks about the list of your planned changes:

Some time ago I had a PM discussion with another forum member about pairing servos, and I also read some related posts published in an RC modelling forum.
What you should consider: If the paired servos have a rigid connection to the paddle and if they are not perfectly aligned, they may work against each other (at least a little bit). For example, one servo arm may try to move some millimeters farther than the other, which will then cause a continuous load on both motors. The solution could be a connection that is not completely rigid, but a little bit flexible.

This is also my new approach. For getting the correct distance between the hinges, I measured the distance 'd' between the two lower parts of my pelvic bone (see the image below).

The idea is as follows:
If the distance between the hinges is larger than 'd', the paddles tend to squeeze your body, and if the distance is smaller than 'd', the paddles tend to lift your body.
I got the basic idea from @RiftFlyer, see here.
The difference is that I will still mount the hinges at the edge of the paddles instead of 30% away from the edge. I will only increase the distance between the paddles as explained above.

That is a very interesting idea, and it makes a huge difference to the load on the servos of course. I may wait to see how you get on though, as at present I do get some lift off the paddles, which I think is a good thing, but doesn't feel the best being under my thighs only. Surge is so good that it draws attention to the imperfect heave.

As for servo alignment, yes - fine tuning that would be critical. There would probably end up being some tiny flex in the steel plate.

I can see that little bit of play between servos as being an issue. If one reaches its full limit, and the other pushed beyond, it seems like it would cause undue stress on the servo. The one is trying to hold it's position, while the other if forcing it further. In the long run it maybe bad on the motors or even pop a tooth off the gear.

Can't wait to get my servo arms, then I can contribute some more to this.

Andrew

I was thinking about that with my build. I will be using 4 weaker servos with 4 arms/rods per back panel. The rods from all 4 will be attached to a single plate though mine may not have associated flex to it. I was planning to limit the Servo travel slightly in both directions as a safeguard. For calibration, assuming all servos have the same degrees of rotation, set all servos to one extreme of the range before attaching the Servo arms and then bolt it all together. Was that your plan as well?

I did lots of testing and calibrating of servos before attaching the flap/connecting arm. With up to 2 servos per heave flap i even have the option of sending different data to each servo on each side and fine tuning the arduino code to allow for misalignments. In the arduino script there's an important line where the simtools values of 0-255 (actually 128-255 if you ignore -ve) get mapped to the digital servo max/min values and you can do this for each servo.

Why don't you use wiper motors or even electric power window motors instead?
It would be a lot stronger and faster.

I guess it wouldn't be quite so 007 Ultra Compact then. I'm optimistic about what I can do with the current servos now actually and eagerly await the result of @MarkusB's tests with adjusted flap placement.