gfesim
02-17-2017, 11:42 AM
Today I would like to show off my force feedback yoke.
I have been flying for some years with cheap CH and Saitek stuff, and later on with a (not so cheap) PFC yoke, but even with the massive metal high quality gear, the concept of spring loaded centering has its limitations. A real yoke shakes and feels mushy on stall imminent, vibrates with the engine rotating, elevator axis are moves while trimming, it feels heavier with higher airspeeds and so on.
Low power FFB systems like the Thrustmaster or Sidewinder are a far cry from realism anyway.
The force feedback presented here is built around the commercially available driver cards form Ian Hopper (BFF) and the concepts presented there. It features realistic loads (up to 40 pound-force, or 200daN, depending on gear translation and hardware setup). Unlike most other sim yokes, the wheel rotates by 180 degrees and has a travel of 20 cm (only limited by my design goals).
A few other builders here on the forum have posted their great work with these cards in their type specific sims.
As I’m running a generic cockpit, I planned for a generic design, which could be transferred to any future cockpit update.
I changed the original mechanical construction proposal from wooden to a metal design and made a few more alterations which I will cover later.
A detailed Sketchup construction drawing is available in the download section.
http://www.mycockpit.org/forums/attachment.php?attachmentid=12114&stc=1
But before I start, let me stress that this project is definetly not for beginners. It requires some skills in mechanical and electrical engineering. The loads can be quite high, so mechanical connections have to be stiff and solid. As in a real plane, sudden movement of the controls can do physical harm to your wrists. The PSU is an unprotected industry part, and the builder has to take care about electrical safety himself. An emergency stop button is common on commercial FFB systems.
And the parts are not exactly cheap. Total cost for the parts was about 900 EUR. But this is still inexpensive compared to industry force-feedback systems which cost 5000 and more, or a quality all-metal spring system which is roughly the same price.
Basically, the system consists of two brushless DC motors, reduction gear to increase torque, and a driver card for each motor. Using BLDCs with field oriented control is todays industry standard for controlling torque. Implementation in software is quite complex but that is all provided in the firmware of the driver cards microcontrollers.
Software for FSX and X-Plane is provided with the cards.
The motor for the roll axis drives the yoke through a gearbox. This motor sits on a carriage running on rails, which is belt driven by the pitch axis motor. Encoders mounted on the motor axis send position data to the driver cards, the cards calculate aerodynamic forces from sim data and send position data back to the sim.
First question was, would I stick with the original construction based on wood, or would I redesign everything in metal. Going with the original plans would have saved me a lot of work, since there is a neat drawing for each and every part, and detailed instructions on how to build.
I’ve been working a lot with wooden materials, but I believe that systems with moving parts, that require some precision, are better done in metal. Additionaly this would make my design more compact and heavier (which is a good thing, given the forces this yoke can produce).
Second decision to be made was about the roll motor. The original design uses a pulley-belt gear assembly, which leads to a relatively wide and high carriage. As compact size was a design goal, I decided for a motor with a planetary gear. This adds the benefit of rotational forces acting in-line with the yoke axis.
Another thing to consider was the placement of power supply and driver cards. As we are working with high currents here (around 7 amps for each motor), and low voltage sensing circuits, I preferred short cables and a compact all-in-one design. The PSU is a Meanwell 24V/15A model, it has a cooling fan and helps with a slim form factor.
One more thing I changed were the rails and bearings. The original design features two rails mounted in the wooden side walls. I found this somewhat difficult to build at the required precision. We don’t want any noticable slack or friction when moving the wheel, therefore the rails have to be precisely aligned vertically and horizontally. So I decided for supported rails, that sit on the ground plate, so vertical alignment is inherently assured.
The whole system, without cover, with a used Piper yoke mounted to it:
http://www.mycockpit.org/forums/attachment.php?attachmentid=12115&stc=1
Front view, yoke unmounted, with the supported rails and bearings and the pitch axis end stops (rubber bumpers)
http://www.mycockpit.org/forums/attachment.php?attachmentid=12116&stc=1
Side view with yoke column dismounted. The small metal block in the front attaches the carriage to the tooth belt.
http://www.mycockpit.org/forums/attachment.php?attachmentid=12117&stc=1
View from left side, with PSU and two driver cards
http://www.mycockpit.org/forums/attachment.php?attachmentid=12118&stc=1
The system has been running stable now for more than a year, and has been a great enhancement for my pit.
Currently I’m planning for a rudder pedals box. This will require even higher forces in the range of 80 to 100 pound-force. Hopefully I will have something ready until end of this year.
Fly safe!
Gerald
I have been flying for some years with cheap CH and Saitek stuff, and later on with a (not so cheap) PFC yoke, but even with the massive metal high quality gear, the concept of spring loaded centering has its limitations. A real yoke shakes and feels mushy on stall imminent, vibrates with the engine rotating, elevator axis are moves while trimming, it feels heavier with higher airspeeds and so on.
Low power FFB systems like the Thrustmaster or Sidewinder are a far cry from realism anyway.
The force feedback presented here is built around the commercially available driver cards form Ian Hopper (BFF) and the concepts presented there. It features realistic loads (up to 40 pound-force, or 200daN, depending on gear translation and hardware setup). Unlike most other sim yokes, the wheel rotates by 180 degrees and has a travel of 20 cm (only limited by my design goals).
A few other builders here on the forum have posted their great work with these cards in their type specific sims.
As I’m running a generic cockpit, I planned for a generic design, which could be transferred to any future cockpit update.
I changed the original mechanical construction proposal from wooden to a metal design and made a few more alterations which I will cover later.
A detailed Sketchup construction drawing is available in the download section.
http://www.mycockpit.org/forums/attachment.php?attachmentid=12114&stc=1
But before I start, let me stress that this project is definetly not for beginners. It requires some skills in mechanical and electrical engineering. The loads can be quite high, so mechanical connections have to be stiff and solid. As in a real plane, sudden movement of the controls can do physical harm to your wrists. The PSU is an unprotected industry part, and the builder has to take care about electrical safety himself. An emergency stop button is common on commercial FFB systems.
And the parts are not exactly cheap. Total cost for the parts was about 900 EUR. But this is still inexpensive compared to industry force-feedback systems which cost 5000 and more, or a quality all-metal spring system which is roughly the same price.
Basically, the system consists of two brushless DC motors, reduction gear to increase torque, and a driver card for each motor. Using BLDCs with field oriented control is todays industry standard for controlling torque. Implementation in software is quite complex but that is all provided in the firmware of the driver cards microcontrollers.
Software for FSX and X-Plane is provided with the cards.
The motor for the roll axis drives the yoke through a gearbox. This motor sits on a carriage running on rails, which is belt driven by the pitch axis motor. Encoders mounted on the motor axis send position data to the driver cards, the cards calculate aerodynamic forces from sim data and send position data back to the sim.
First question was, would I stick with the original construction based on wood, or would I redesign everything in metal. Going with the original plans would have saved me a lot of work, since there is a neat drawing for each and every part, and detailed instructions on how to build.
I’ve been working a lot with wooden materials, but I believe that systems with moving parts, that require some precision, are better done in metal. Additionaly this would make my design more compact and heavier (which is a good thing, given the forces this yoke can produce).
Second decision to be made was about the roll motor. The original design uses a pulley-belt gear assembly, which leads to a relatively wide and high carriage. As compact size was a design goal, I decided for a motor with a planetary gear. This adds the benefit of rotational forces acting in-line with the yoke axis.
Another thing to consider was the placement of power supply and driver cards. As we are working with high currents here (around 7 amps for each motor), and low voltage sensing circuits, I preferred short cables and a compact all-in-one design. The PSU is a Meanwell 24V/15A model, it has a cooling fan and helps with a slim form factor.
One more thing I changed were the rails and bearings. The original design features two rails mounted in the wooden side walls. I found this somewhat difficult to build at the required precision. We don’t want any noticable slack or friction when moving the wheel, therefore the rails have to be precisely aligned vertically and horizontally. So I decided for supported rails, that sit on the ground plate, so vertical alignment is inherently assured.
The whole system, without cover, with a used Piper yoke mounted to it:
http://www.mycockpit.org/forums/attachment.php?attachmentid=12115&stc=1
Front view, yoke unmounted, with the supported rails and bearings and the pitch axis end stops (rubber bumpers)
http://www.mycockpit.org/forums/attachment.php?attachmentid=12116&stc=1
Side view with yoke column dismounted. The small metal block in the front attaches the carriage to the tooth belt.
http://www.mycockpit.org/forums/attachment.php?attachmentid=12117&stc=1
View from left side, with PSU and two driver cards
http://www.mycockpit.org/forums/attachment.php?attachmentid=12118&stc=1
The system has been running stable now for more than a year, and has been a great enhancement for my pit.
Currently I’m planning for a rudder pedals box. This will require even higher forces in the range of 80 to 100 pound-force. Hopefully I will have something ready until end of this year.
Fly safe!
Gerald