RiftFlyers G-Seat

[RiftFlyer]

Hi all,

I've started to make some progress with my Bergison inspired G-Seat. This will be used primarily for helicopter sims with the Oculus Rift.

I've never fabricated anything in steel before so this has been a great learning experience. I bought myself a cheap welder and some steel and after lots of YouTube videos I set about joining some metal! The results aren't too bad but I won't be posting any close ups :-p. I've backed up my welds with some pop riveting. I have the main seat frame complete. This will be mounted on vibration isolation mounts to a base which holds my flight controls.



I'll be using Arduino and monster moto shield to control the flaps and belts. I need to source some better mounting brackets for the motors and pulleys though. Progress is slow due to time constraints and I'll have plenty of questions before I'm finished. Any comments, advice or questions are welcome

[Bergison]

Hi RiftFlyer!

Nice to see another build inspired by my G-seat! This should be interesting with an all metal construction. Don't forget to ground-isolate your motors and feel free to contact me for any upcoming questions!

Good luck and enjoy building!

[RiftFlyer]

I've started to make some progress on the actuators. I've assembled one as a test.

I tried the official moto monster sketch viewtopic.php?f=39&t=1547 but the motor doesn't move. I loaded the sketch posted by racingMat in his 2dof build viewtopic.php?f=37&t=943 .I have movement and can stop and reverse the motor with the pot.

I still need to figure out why the motor moves with either of the arduino analogue pins listed in the sketch. I would have expected each pin to be tied to an individual motor output. More testing needed before I connect the pot to the motor.

[RiftFlyer]

Ok I have made a little more progress.

Here is a video of my manual feedback test:

http://youtu.be/BC0miR7b-Tg



I connected the pot to test automatic feedback positioning. I'm having issues though. I'm not sure if it's a PID problem or if it is a setting that I missed. I'd appreciate anyone's thoughts.

http://youtu.be/LxlYdvD_LW8

[RiftFlyer]

I made some more progress on the actuator. It seems like the default PID were causing the oscillations. Still not perfect results but the oscillating has been reduced. It will require further adjustment when fitted in the seat and tested under load.

http://youtu.be/n7_KzX5j_1U

[RiftFlyer]

I played around with the PID and pwm % and I think I now have much better results. Much less overshoot and oscillation. Sorry for the poor lighting.

http://youtu.be/7xoHRhkp4E8

[RacingMat]

Hello RiftFlyer!

nice to see a new gSeat project coming to life!!

what code are you using at the end?

I hope your motor will be fine under some load as it looks like standard wiper (from what car model is it?)
yours
Mat

[RiftFlyer]

Hi Mat,

I used the code from Sirnoname in the end due to the ease of tuning with the gui (viewtopic.php?f=39&t=1547). I want to take this opportunity to thank you for your thread though Mat, I learned a great deal from it. I had to lower the max pwm to 80% and remove all power offset . I also ended up with P=2, I and D at 0.

The graph shows good response with little overshoot and no jitter.

I'm at a loss to understand why I needed to go so far from default values. I have geared carefully so I have full travel on my pots. The only reasons I can come up with are perhaps:

- Belt driven pot causes slightly different response versus gears,
- I am only driving the motor over an arc of about 90 - 100 degrees,
- interference (I don't rate this too highly though).

The motors are from VW Caddy vans (Bosch 9390332376). I have isolated the internal brushes from the housing. As you say under load the tuning may be very different so I may find default values are less of an issue. Based on Chris Bergisons experience I don't think the motors will have any issue with the load.

I've added heat sinks to the moto monster boards today and also picked up some 15amp fuses for the motors. I just wish I had more free time to get it finished. I haven't got my head around the math setup yet. I think it will require some hands on tinkering. I have completed a second actuator so will spend some time running them in unison before inserting them in the seat.

[RacingMat]

you're welcome! I'm glad my tutorial was helpful to you

yes, the graph can be more handy but PID tuning stays difficult.
If you ended with I and D = O, it's no longer a PID

With the code I use, it's straightforward: no PID tuning as it's a simple P driven regulation.
Eager to see more from your sim!

[RiftFlyer]

I've the majority of the work completed on the base. I need to tidy up some welds and prep the base for painting. The cyclic and anti torque pedals are quickly detachable to allow me easily reconfigure for my G27. The seat is vibration isolated from the base. I have fitted casters to allow easy movement when not in use.

The steel inner sections of the seat base and back rest are not yet secure. When the flaps are fitted these will need some final adjustment for proper movement sensation.