Page 1 of 3 123 LastLast
Results 1 to 10 of 21
  1. #1
    75+ Posting Member



    Join Date
    Sep 2007
    Location
    Scotland
    Posts
    81
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    DIY Force Feedback Yoke

    ...have been working away on this for ages on and off - and it still isn't right yet!

    However, I've written up a couple of web pages on the project for anyone interested. I think the mechanics, software and control hardware is working well - but the motors aren't quite there yet.

    http://buggies.builtforfun.co.uk/Sim...ack-yoke-1.php

    ...does anyone else get so obsessed about solving a technical problem?

    Ian

    MARCH '11 ,PS Here's the link to the latest FFB yoke design..... the design above has been updated..

    http://buggies.builtforfun.co.uk/Sim/FFB-yoke-1.php
    Last edited by IanH1960; 03-29-2011 at 05:34 AM. Reason: added updated link...

  2. Thanks nax228, cz4hkr, Leo Bodnar, autocadplease thanked for this post
  3. #2
    1000+ Poster - Fantastic Contributor Bob Reed's Avatar
    Join Date
    Oct 2005
    Location
    Holley, New York U.S.A.
    Posts
    1,776
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Simple answer...... YES..... Like a bull dog..... I will not let it go once it stumps me....Sad really.....
    Bob Reed

  4. #3
    150+ Forum Groupie Roland's Avatar
    Join Date
    Apr 2006
    Location
    Netherlands
    Posts
    187
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Hi Ian,

    Another impressive piece of work! You really took a lot of things into consideration.
    Just a question: Does the yoke react on autopilot? Since you seem to take info from joystick directly, I think not. Correct? Is FS trim info useful in any way here?

    You still use DC motors with 2 brushes. Any trials on 4-brush motors?

    Great work!
    RR

  5. #4
    75+ Posting Member



    Join Date
    Sep 2007
    Location
    Scotland
    Posts
    81
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Hi Roland,

    Thanks, it's an interesting project....

    Autopilot - the advantage of using a software driven approach is that all sorts of effects can be programmed in. The yoke doesn't currently follow auto-pilot induced movements but it could be programmed to do so. I could monitor the auto-pilot status and if it is engaged use the FS reported control surface positions to drive the yoke positions. This could be done with the system driving in a closed-loop position controlled mode similar to the method I suggest to link pilot and co-pilot stations through the software. I'll need to check the effect of moving the yoke whilst the autopilot is engaged to see if this has any strange effects in FS.

    FS Trim - I'm not sure if the FS trim inputs are of much use here - I bypass them completely to get a more realistic trim behaviour from the yoke. However this wouldn't be a problem for a builder unless they particularly wanted to see the trim wheels moving in the flight sim display. The trim pot joystick axes could be disabled in FS but maintained in the force feedback software and would continue to affect the load balance of the yoke axes.

    Motors - The cost of new skewed rotor servo motors is an issue - and I imagine that 4 brush motors would be even more expensive than the 2 brush motors I've tested. Do you think though that 4 brushes would significantly reduce the commutation torque ripple? My feeling now is that to get a really good quality control loading the torque ripple needs to be removed almost entirely.

    I've just received a couple of the ironless disc-armature pancake motors so should be able to do some tests on them soon - once I figure out how to mount them! A question you might be able to answer for me is the effect of their low inductance on the PWM controllers. I know some of the EV controllers I've used can deal with low-inductance pancake motors but have heard a suggestion that not all might?

    Ian

  6. #5
    150+ Forum Groupie Roland's Avatar
    Join Date
    Apr 2006
    Location
    Netherlands
    Posts
    187
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Hi Ian,

    Too low inductance will increase ripple current in the motor current. I remember my motors are around 2.2mH. With H-bridge switching at 20kHz, 36V supply gives about 500mApp ripple. Bigger ripple will increase steadystate bridge dissipation. I'd say that 1 - 1.5App is still acceptable.

    I'll check my 4-brush types at home vs some 2 brush types. Not sure how to quantify the torque smoothness, have to think of a good test setup.
    I'm in Shanghai right now, I'll get back to you as soon as I'm back home.
    RR

  7. #6
    75+ Posting Member Leo Bodnar's Avatar
    Join Date
    Apr 2008
    Location
    UK
    Posts
    95
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Hi Ian,

    Brilliant project! Very well done!

    I have just finished working on a custom project for a racing simulator steering wheel. Even though it's only one axis (unlike yours) the design challenge was to get 25Nm of peak torque (which is about 30kg force on a wheel rim.) At such high forces the torque control resolution had to be high otherwise you would feel steps or grittiness at low forces. I have used 15bit torque control and it works great.

    I have used 3kW AC servomotor with custom controller and 3rd party servo amp. I am prototyping my own FFB controller/servo amp combo designed to drive up to 70V/30A AC servomotors. The firmware supports X-Y axes as in joystick but I had only one servo amp onboard. It complies with standard Windows FFB joystick interface but has 50MIPS DSP onboard so one can literally run the full dynamic forces calculation inside the innermost control loop (2-5kHz.)

    It would be interesting to produce an engineering samples run of a dozen X-Y FFB controllers and see what people like you can do with them.
    Quote Originally Posted by IanH1960 View Post
    ...have been working away on this for ages on and off - and it still isn't right yet!

    However, I've written up a couple of web pages on the project for anyone interested. I think the mechanics, software and control hardware is working well - but the motors aren't quite there yet.

    http://buggies.builtforfun.co.uk/Sim...ack-yoke-1.php

    ...does anyone else get so obsessed about solving a technical problem?

    Ian

  8. #7
    75+ Posting Member



    Join Date
    Sep 2007
    Location
    Scotland
    Posts
    81
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Hi Leo,

    Glad you liked the work - your 12bit joystick module works a treat.

    Your AC brushless controller sounds very interesting - do you have a write up anywhere for it? I tried a proprietory brushless drive but I had to keep the cost down and felt higher quality drives were just too expensive for me to risk with my limited knowledge.

    My preference is to keep the freedom to do the force calcs separately from the usual FFB provision so as to be able to execise some real control over the force generation and the facility to do them within your controller is interesting. However I would also be interested to know if the combo unit would be instructable from the PC in a torque demand mode either using RS232 or USB HID. This would offer the possibility of driving it directly from my existing software without extensive re-write.

    Do you have a feel for possible cost of the combo unit?

    Roland..... Shanghai! Boy this new job is getting you around!

    Ian

  9. #8
    75+ Posting Member Leo Bodnar's Avatar
    Join Date
    Apr 2008
    Location
    UK
    Posts
    95
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Quote Originally Posted by IanH1960 View Post
    Hi Leo,
    Your AC brushless controller sounds very interesting - do you have a write up anywhere for it? I tried a proprietory brushless drive but I had to keep the cost down and felt higher quality drives were just too expensive for me to risk with my limited knowledge.

    My preference is to keep the freedom to do the force calcs separately from the usual FFB provision so as to be able to exercise some real control over the force generation and the facility to do them within your controller is interesting. However I would also be interested to know if the combo unit would be instructable from the PC in a torque demand mode either using RS232 or USB HID. This would offer the possibility of driving it directly from my existing software without extensive re-write.

    Do you have a feel for possible cost of the combo unit?
    Hi Ian,
    Which servo drives have you used? Unfortunately It's almost impossible to find a drive with accurate torque control. So far I have used servo drives form a small startup company in Finland which I have modified to accommodate higher currents and managed to control via SPI. It's awkward but it worked. But I want to integrate the drive and controller into one unit.

    You can trust effects like damping, inertia and spring to be controlled internally and just modify the parameters when necessary. I.e. in first approximation elevator load is a spring where null point is determined by trim and angle of attack and elasticity by IAS, etc. Parameters will tolerate much lower update rate then actual torque control calculations. AC servos need high resolution position feedback so I use 2000-4000CPR integrated encoders, I am not planning to support potentiometers.

    Parameters control is via USB HID with 500Hz update rate.

    I have no idea about the cost yet since it's only a prototype and a lot will depend on power stage final specification. What is you preferred bus voltage and maximum current?

  10. #9
    75+ Posting Member



    Join Date
    Sep 2007
    Location
    Scotland
    Posts
    81
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    Hi Leo,

    The brushless drive I used was a low cost Chinese built unit with hall effect feedback and what looks like sinusoidal commutation. I though originally the current proportional torque detents were due to a trapezoidal commutation method given the lack of external position feedback but measuring the voltage in the windings over a shaft revolution shows a sinusoidal voltage profile in each winding. It was clear that an AC brushless drive with high res external feedback would be better - but the costs were just too high.

    I think it makes some sense to do velocity and acceleration related torque component calcs where the refresh speed is high and the velocity and accel profiles are smooth - however the limits I've come up against in my application aren't due to poor refresh speed in the torque calcs - they are due to the inherent lack of smoothness in the torque output of the various motor types I've tried. An effective low cost AC brushless drive might sort this.

    Up to 48V/30Amp ratings would be okay for my applications. Just a thought - make the current a wee bit higher and I could drive it from my motion platform software - this would be particularly interesting if it had an internal PID position control loop and only needed to be fed position demands (3 channels would help but two 2-channel units could be used on a 3 DOF platform). A good low cost brushless platform drive might interest some DIY motion platform builders.

    I think a lot would hang on the cost. Keep us posted if you do more development.

    Ian

  11. #10
    10+ Posting Member cz4hkr's Avatar
    Join Date
    Apr 2009
    Location
    Prague
    Posts
    11
    Contribute If you enjoy reading the
    content here, click the below
    image to support MyCockpit site.
    Click Here To Contribute To Our Site

    737 dual yokes

    Hi,
    well done!
    I have the same plan how to implement force feed back to Boeing 737-800 dual linked yokes. Right now we are doing mechanical part.
    Btw does anyone here know if there exist some chart or formula for boeing load control?
    I have some info that there is some chart in B737 maintenance manual.





    Quote Originally Posted by IanH1960 View Post
    ...have been working away on this for ages on and off - and it still isn't right yet!

    However, I've written up a couple of web pages on the project for anyone interested. I think the mechanics, software and control hardware is working well - but the motors aren't quite there yet.

    http://buggies.builtforfun.co.uk/Sim...ack-yoke-1.php

    ...does anyone else get so obsessed about solving a technical problem?

    Ian

Page 1 of 3 123 LastLast

Similar Threads

  1. Software for DIY flight control loading (force feedback)
    By IanH1960 in forum Cockpit Parts and Motion Platforms
    Replies: 1
    Last Post: 07-04-2010, 05:13 PM
  2. force feedback idea
    By gokhotit in forum Cockpit Parts and Motion Platforms
    Replies: 12
    Last Post: 10-09-2009, 12:18 PM
  3. Rudder pedals with force feedback
    By Roland in forum Cockpit Parts and Motion Platforms
    Replies: 2
    Last Post: 06-15-2008, 02:40 AM
  4. Turning Logitech Force Feedback 3D into something more "yoke-ish."
    By Christian in forum Cockpit Parts and Motion Platforms
    Replies: 10
    Last Post: 08-29-2007, 04:03 PM
  5. An interesting Idea: Joystick Force Feedback to a Platform
    By phidgets in forum Phidgets & Cockpit Simulator Builder
    Replies: 0
    Last Post: 04-30-2005, 04:30 PM