RiftFlyer G-Seat project

I'm finally getting to a point where I have something to show in my helicopter G-Seat project. I was inspired by the Bergison G-Seat previously discussed here:

http://forums.eagle.ru/showthread.php?t=122577&highlight=Bergison+gseat

The idea is to simulate motion by applied pressure both via flaps and a tensioning harness. I will also be stripping down the 908 gametrix Jetseat cushion and installing the vibration motors from it into the flaps of the seat.

So far I have constructed the seat and a base. The seat itself is isolated from the base and sensitive controls by four rubber engine mounts:

This was the first time I have ever picked up a stick welder or done any fabrication in metal. It's a steep learning curve.

Here is a view with the controls temporarily in place:

I have also completed the mechanism for the flaps in the backrest. The motors are VW wiper motors. I will be controlling them via Arduino and a Monster Moto Shield

These are old pictures. I will update this post shortly to show recent progress.

EDIT: Update 11th April:

I made some progress finalizing the mechanism in the backrest. The flaps and belt tensioners are now aligned. I need to secure the feedback pots into position and secure the brackets on the rear of the flaps then I will paint and move on to testing.

EDIT: Update April 12th

Mock up of a flap actuator in motion

EDIT April 15th:

I have moved from the garage back to the computer. I need to start work on the electronics and plan the wiring. I cleaned up some dodgy welds and added paint. I now need to source a few more motors for the base.

EDIT Updated April 17th

I have finalized the power system for the actuators. As recommended here I am using separate 12volt supplies for the base and back of the chair. Power is being provided by the trusty HP DPS600pb server power supplies. Each is rated for 47amps.

I made up a switch box to keep the wiring tidy. Each supply has its own switch. Power is passed through a fuse to a relay to the motor monster shields. The relays are connected to a kill switch which is mounted within easy reach.

Internals of the power box

Power Supply switches

Proposed mounting location. Note the easy access to the kill switch between my legs :music_whistling:

EDIT: June 8th

I started work on the base of the seat. I've switched over to MOD 1 gears in place of the belts and pulleys. These are working out much better

Today I managed to make up the brackets for the pulleys and motors for the base of the seat.

I made a jig from an old wiper motor gearbox which allowed for much better alignment. This made life so much easier.

The brackets will be mounted to a removable plate for ease of axis. Here I am marking the holes on the plate for mounting.

Reserved.

Reserved for future update

updated original post.

Looking really good! I made a G-seat too, and it's incredibly immersive. Everyone needs one!

Looks quite a thing. Waiting to see this take shape and start moving

Looking really good! I made a G-seat too, and it's incredibly immersive. Everyone needs one!

Thanks Rico,

I have followed your G-Seat thread. Have you any videos of it in action? Are there any pitfalls or mistakes that you found that you could warn me of in advance? I will be calling on your expertise once I start interfacing X-sim and dcs ;-)

Thanks Rico,

 

I have followed your G-Seat thread. Have you any videos of it in action? Are there any pitfalls or mistakes that you found that you could warn me of in advance? I will be calling on your expertise once I start interfacing X-sim and dcs ;-)

I don't have any videos, but I guess it's pretty much the same as Bergison's.

If it's not too late, try to incorporate some way to adjust the pivot points on the panels, especially the bottom ones. There is a big difference from just a small adjustment in how wide or narrow they are in terms of effectiveness and comfort.

Another thing I want to add is a force strength potentiometer on the side of the chair. You can adjust the strength on the fly with a software slider, but doing it with a knob on the chair would be better, I think.

Oh, and BTW, make sure you use a nice big mushroom style kill switch within easy reach. The first time you loose a rotor blade in the shark you'll be glad you did. :lol:

I had some trouble with burning up the motor controllers, which I think it was caused by ground loop issues. The seat power supply, speed controls, arduino, and computer were all tied together by ground. I added optoisolators to separate the arduino from the motor controllers and haven't had a problem since.

I'm thinking about building a 3rd version now, but instead of panels, I'll use a set of parallel rollers and nylon fabric (like a camp chair) suspended between them. The rollers would tighten or loosen the fabric seat so you'd sink or lift in it. I'm going to be building a mock up one pretty soon.

As far as x-sim, also look at http://www.xsimulator.net/, apparently it's some kind of derivative of x-sim. I didn't see it until the other day and it looks much simpler and cleaner than x-sim. The main problem with x-sim is that it's got a billion features, but they've just been tacked on and now the software is super convoluted. I feel like I got it working by accident.

I'm going to check xsimulator out for v3, but I need to learn how it does the serial output so I can rewrite my arduino code.

I don't have any videos, but I guess it's pretty much the same as Bergison's.

 

If it's not too late, try to incorporate some way to adjust the pivot points on the panels, especially the bottom ones. There is a big difference from just a small adjustment in how wide or narrow they are in terms of effectiveness and comfort.

 

Another thing I want to add is a force strength potentiometer on the side of the chair. You can adjust the strength on the fly with a software slider, but doing it with a knob on the chair would be better, I think.

 

Oh, and BTW, make sure you use a nice big mushroom style kill switch within easy reach. The first time you loose a rotor blade in the shark you'll be glad you did. :lol:

 

I had some trouble with burning up the motor controllers, which I think it was caused by ground loop issues. The seat power supply, speed controls, arduino, and computer were all tied together by ground. I added optoisolators to separate the arduino from the motor controllers and haven't had a problem since.

 

I'm thinking about building a 3rd version now, but instead of panels, I'll use a set of parallel rollers and nylon fabric (like a camp chair) suspended between them. The rollers would tighten or loosen the fabric seat so you'd sink or lift in it. I'm going to be building a mock up one pretty soon.

 

As far as x-sim, also look at http://www.xsimulator.net/, apparently it's some kind of derivative of x-sim. I didn't see it until the other day and it looks much simpler and cleaner than x-sim. The main problem with x-sim is that it's got a billion features, but they've just been tacked on and now the software is super convoluted. I feel like I got it working by accident.

 

I'm going to check xsimulator out for v3, but I need to learn how it does the serial output so I can rewrite my arduino code.

Thanks for the helpful tips.

I have a large mushroom kill switch for the front of the seat just within easy reach. I was going to rig it to a relay for the power supplies. I'm using two server supplies which I think I can safely connect in parallel.

I have not yet started work on the base mechanism. The struts supporting the flap pivot points are not secure yet so I will bolt them in place rather than weld/rivet to allow for adjustment.

I like your force potentiometer idea. Please share details in your thread if you implement it. I mainly fly helicopters but had considered having two X-sim profiles to tailor the forces for the faster jets. I was afraid if I set max deflection for say 9g/-2g that I'd have too little movement when flying helicopters.

I have an account on Xsimulator also. I actually found it before x-sim. I decided to go with X-sim as its what Bergison used. I do find the software very convoluted. Regarding your ground loop issue are you sure your motors were isolated? The wiper motors I got required modification as the motor housing could become live when rotating in one direction. I'm not familiar with optoisolators. Could you please share a link and how you have them connected? The last thing I want is damaging my motherboard. Any suggestions on how to ensure that won't happen would be greatly appreciated.

Thanks again.

I have not yet started work on the base mechanism. The struts supporting the flap pivot points are not secure yet so I will bolt them in place rather than weld/rivet to allow for adjustment.

:thumbup: I think that will prove to be useful.

I mainly fly helicopters but had considered having two X-sim profiles to tailor the forces for the faster jets. I was afraid if I set max deflection for say 9g/-2g that I'd have too little movement when flying helicopters.

I have -2/+6 to get full scale deflection with the strength slider at mid point. With the strength slider at the most sensitive setting it still requires >3g to get full scale deflection, which is just barely enough for the helicopters. Two profiles should work pretty well though.

Regarding your ground loop issue are you sure your motors were isolated? The wiper motors I got required modification as the motor housing could become live when rotating in one direction. I'm not familiar with optoisolators. Could you please share a link and how you have them connected?

The isn't any kind of chassis ground path between the motors, and the feedback pots weren't grounded to the wiper motors, so I'm not too sure what was going on there. Autopsies on the failed controllers seemed to show that the bootstrap diodes for the high side mosfet drivers were failing, and then that would fry the mosfets, but I couldn't work out exactly what was going on, but figured isolating the different power supplies would be a start.

The optoisolators (aka photocouplers, and some other similar names) are IC packages that contain a LED and a phototransistor. It allows you to send information while keeping the circuits electrically separated.

I had some challenges implementing them though. Early on, I changed the arduino's PWM frequency from the default 500Hz to ~30KHz to move the (very loud) motor hum into the ultrasonic range. I thought I spec'd some optoisolators that had enough bandwidth to work at that frequency, but I couldn't it to work well at all; low pulse widths would barely get through. Fortunately using a buffer it cleaned it up enough that I didn't have to do anything else.

If it's not too late, try to incorporate some way to adjust the pivot points on the panels, especially the bottom ones. There is a big difference from just a small adjustment in how wide or narrow they are in terms of effectiveness and comfort.

Rico , since you already touched on that, after your experiments what is the ideal width between pivot points you arrived to? I remember you had them quite wider then what I have in my WIP seat.

Looks quite a thing. Waiting to see this take shape and start moving

Thanks Anton. I'm watching your build with interest also. Good luck with your project.

:thumbup: I think that will prove to be useful.

 

 

 

I have -2/+6 to get full scale deflection with the strength slider at mid point. With the strength slider at the most sensitive setting it still requires >3g to get full scale deflection, which is just barely enough for the helicopters. Two profiles should work pretty well though.

 

 

 

The isn't any kind of chassis ground path between the motors, and the feedback pots weren't grounded to the wiper motors, so I'm not too sure what was going on there. Autopsies on the failed controllers seemed to show that the bootstrap diodes for the high side mosfet drivers were failing, and then that would fry the mosfets, but I couldn't work out exactly what was going on, but figured isolating the different power supplies would be a start.

 

The optoisolators (aka photocouplers, and some other similar names) are IC packages that contain a LED and a phototransistor. It allows you to send information while keeping the circuits electrically separated.

 

I had some challenges implementing them though. Early on, I changed the arduino's PWM frequency from the default 500Hz to ~30KHz to move the (very loud) motor hum into the ultrasonic range. I thought I spec'd some optoisolators that had enough bandwidth to work at that frequency, but I couldn't it to work well at all; low pulse widths would barely get through. Fortunately using a buffer it cleaned it up enough that I didn't have to do anything else.

Good info Rico. I think I read that some optoisolators will invert PWM signal. It is a beyond my electronics knowledge however. I'm going to stick with plenty of fuses and a powered USB hub for now and hope I don't have any mishaps.

what is the ideal width between pivot points you arrived to?

Originally, 19 cm, and the v2 seat is at 25 cm. It's *still* not right, but even on v2 I didn't make it easy to change. :P

I think I read that some optoisolators will invert PWM signal. It is a beyond my electronics knowledge however. I'm going to stick with plenty of fuses and a powered USB hub for now and hope I don't have any mishaps.

It will indeed invert the signal. It's easy enough to take care of in software, but I had the fortune to have inverting buffers on hand, so I didn't need to change anything.

Yeah, I think you'd be fine without them, it's just something that came up after a few months of use and is probably particular to my setup. I don't think you'll have any problems.

When you're setting up the final stages of hardware it's helpful to have a lab type power supply with variable voltage and current limiting, so you can set things up without the rig tearing itself apart if things go haywire.

When you're setting up the final stages of hardware it's helpful to have a lab type power supply with variable voltage and current limiting, so you can set things up without the rig tearing itself apart if things go haywire.

Got the bench supply on hand with an inline low amp fuse ;-)

Very cool video Frusheen. Can't wait to start working on power as well. One note though, be careful connecting PSUs in parallel. It might not be a problem, but per my Electronics engineer friend ( who for the project gifted me three 12v supplies of not matching amp rating) there is some sort of interconnection needs to be made for their monitoring circuitry to be aware it is a group effort and not go against each other. I didn't yet look into that.

Rico, interesting observation. I did see you initially mounting them wide and had a hunch that wider is better. but in a brainfart moment during assembly went completely the other way. Back flap pivots mounted close together, almost as opening a book which means im more pushed forwards instead of "falling into" the chair" on actuation.

Thanks for the feedback, looks like I will be moving pivots to almost near outer edges, which might require some sort of spring to balance the fact that motors now hold more of the rider's weight.

Frusheen, good stuff. I missed it the first post too. :P

Agrasyuk, exactly. The back ones are much better wide to allow you to sink into the seat. The bottom ones are tricky because your weight has to be supported on the pivot more, but the placement is also important because the way it aligns with your hip bones makes a big difference on comfort.

Regarding parallel power supplies, it might be best to have one power the left side motors and the other power the right side.

The power supplies I have to hand are Hewlett Packard DPS-600pb designed to be hot swapable for use in servers. They are the same type and rating and from the pin out they feature a 'current share pin'. Many use them in parallel to run chargers in RC. I had a set before connected in series :-o Not the safest thing to do as I had to have a floating DC charge on one and isolate it from earth on the AC side. Powering a 1300watt charger they sure charged my 12s lipo batteries quickly though lol

Well then, that sure sounds like it'll do the business! :thumbup:

1300W charger into a 12s pack?! That's absurd. What are you flying/driving? (sorry to derail, but I gots to know. :D )

Well then, that sure sounds like it'll do the business! :thumbup:

 

1300W charger into a 12s pack?! That's absurd. What are you flying/driving? (sorry to derail, but I gots to know. :D )

Lol. A couple of Mikado Logo 600 RC helicopters.

Ha, missed on redundant server supplies earlier. Sounds very beefy

While on topic of Supplies, Rico if I may inquire what amp draw you seeing while flying in your setup? I have a pair of 10A and single 15A 12V supply. I was told those are not cheap china but rather very respectable industry grade devices. Was thinking to power each bottom motor from dedicated 10 amp unit and both back ones from a 15A one . i know it is 5amp short from Bergison setup. do you recon it may be enough or would you say upgrade needed.

Ha, missed on redundant server supplies earlier. Sounds very beefy

 

While on topic of Supplies, Rico if I may inquire what amp draw you seeing while flying in your setup? I have a pair of 10A and single 15A 12V supply. I was told those are not cheap china but rather very respectable industry grade devices. Was thinking to power each bottom motor from dedicated 10 amp unit and both back ones from a 15A one . i know it is 5amp short from Bergison setup. do you recon it may be enough or would you say upgrade needed.

Anton a stalled wiper motor can pull 15amps or more. The supplies I listed above are readily available on eBay and are very reasonably priced. I'm sure Rico can give you an idea of normal amp draw in use but if using the supplies you have be sure to add fuses.

Frusheen, those are some really nice helicopters! I haven't had a 3d helicopter bigger than a 400, I can't afford to crash them at the rate that I do. :D

Anton, regarding the power supply demands, I never saw much more than 8 amps per motor during testing. I stall tested one up to 10A, and like Frusheen said, they're more than happy to do up to (and likely over) 15A. You'll likely never see it in practice, unless maybe the bottom flaps are carrying too much weight.

I just did about an hour of mixed flying with a kill-a-watt meter as a quick test. I did a good amount of A2A in jets, and there's nothing more violent than the thrashing that goes on when your shot up plane is tumbling through the air. Even with all that, I only peaked at 180W at the outlet (which would mean <15ADC) and the entire session averaged only 26W.

I have moved from the garage back to the PC to work on the wiring. Updated the first page with additional images.