Open Source Joystick FFB / DIY FFB Joystick

Slartibartfast: I'm participating in a similar project, simpit in a glider hull. Pretty much at starting point, and summer (flying season) doesn't help. How's your project going?

I've kind of hit the wall when I got to designing the spring return mechanism I'm afraid.

You can see my progress in the SimHQ forum if you're interested though.

Nope, see on Reddit:

 

 

http://imgur.com/a/zfItq

This one have a linear motor. I have doubts regarding the precision of this motor to produce the exact effects, even if might be more resistant. The software part however is fully worthing to be analyzed.

Hi

I have some News for this Project.

First of all I think I found a wayto fix my Timing Problem.

I simply split the calculations to getmore time for the USB calls.

That costs only a few cycles more and itseems to work as planned.

(I will see if I can upload a new Hex nextweek, for everybody to test)

Having that sorted I have some Problems

First

I Need a Better test board. So Istarted to build a Demo stick.

I am experimenting with different Gimbaldesigns, but I am not very happy with the results. So I am open for ideas.

Second

I can't get my brain to come up with aformula to get this Sinus Function to work. So if someone is good at math and hassome spare time I will be thankful.

MetalGear_Honk

I am planning to start buying components to try help you on tests. However I am unsure what controler should I go for. Arduino 2560 is ok ?

Slartibartfast: Thanks for the link, always interested to see what people have done! Don't know if it's of any use, but Finnish Traffic Agency has a lot of data available on PIK-20 - for example, "Control of ailerons, elevator and rudder" at https://www.trafi.fi/en/aviation/airworthiness/pik-20_support/pik-20_and_pik-20b_parts_catalogue_and_specification_drawings might be interesting, control column and trim systems visible :) PIK-20 has the same kind of trim that LocNar told SZD-59 has, that is, a adjustable spring loading the control column. I saw some really high performance modern glider has the same arrangement, but can't recall the type - technical pictures of these planess are really hard to find :P

When planning a distributable, open joystick, the available tools define quite a lot of what is a "good" gimbal/centering structure. 3D printing, CNC machining, laser cutting & welding all prefer different kind of structures. But the mechanism, IMHO, stays the same. A scissor centering mechanism that you see in most commercial joysticks is never a good solution; it simply doesn't represent the forces in real planes. For a passive system the cam seems to be somewhat optimal; at least it can be tuned by changing the cam, springs and preload. Even speed-variable load and "correct" trim aren't out of picture. Electrical control loading gives a lot more options, but is also two-three steps more complicated.

(WHAT? Condor2 is not dead-dead-dead? :D )

Abburo: IIRC MetalGear_Honk said his software is on Arduino Leonardo, which uses AVR32U4, which means that 2560 wouldn't do. Arduino Leonardo, Arduino Micro, Teensy 2.0 and some others use the same chip.

MetalGear_Honk: PM the code or link to it, I'll take a look. Based on some mental calculations I think it should be doable on AVR. And if not, eg. STM32F103 -based development boards are dirt cheap ($2) and are a lot faster.

Slartibartfast: Found the high-perf glider picture. ASW-27B at wings and wheels; practically everything is visible:

 

Abburo: IIRC MetalGear_Honk said his software is on Arduino Leonardo, which uses AVR32U4, which means that 2560 wouldn't do. Arduino Leonardo, Arduino Micro, Teensy 2.0 and some others use the same chip.

 

.

Thanks Jusu :thumbup:

Hi

For the Controller: yes I use anATMega32U4 that is the one from a Pro Micro

Or teensy. I have a STK600 Board so therecan be some differences.

forthe Code:

I hope it is not to chaotic. As I am usedto work alone.

I have used great parts from FFB Adapterso all credits to them.

JOYSTICK_Test.zip

Thank you MetalGear. I am gonna buy some hardware next month... then I will start make some tests. I might require your assistance during that time.

I know that stepper motors are not desirable for this application, however during tests I am gonna use this type of motors then I am gonna switch to RC brushless outrunners.

Great work MetalGear! I've ordered some Teensy's and looking forward to trying out your code as well. I've got the motor hardware breadboarded ready for testing.

Hello everybody ,

Firstable thank you for your work for an OSJ i am searching for a long time now a way to fly with a more powerfull and precise ffb stick .

An interesting post for shure ! After reading all the posts i have one "stupid" question :

As the lack of joystick ffb firmware is the problem :

Why couldn t we build and plug two force feedback devices with arduino ( or another ) , uisng the Mmos as firmware And assign one wheel to the X axis and the other to the Y axis .

As the Mmos as a minimum of 90 degres of rotation that could be reduce easely to about 20 degre for flying stick with gear ?

For exemple i can plug 2 sidewinder ffb2 and assign one axis from number one and the other from number two in dcs and it work fine .

Disadvantage is that you have to build 2 Mmos divice but as we don t need necesseraly powerfull servo motors we could achive the setup with costless DC motors?

Hi

the first problem is that the protocol for ffb will send a direction where the force is going to go. so somehow we would need to determine how much force witch wheel should get. and when we have that sorted the rest is not much more of a problem.

the second problem would be to get both axis synchronised. for example to build a square wave form in 20° direction. and not one in0° and one 90° with different strength.

and last we would have two devices in the configuration but ffb will send only one protocol for both where we are back for the first problem.

at the end we will need to write a new firmware to sort out these problems.

or we can write a new one from scratch. (witch I am doing at the moment)

I uploaded a test version if you feel in for some experimenting.

Greetings

MetalGear_Honk

Thank you Metalgear for your lightening . Is your test firmware usable on aurdino ? is it a test or that it make already arduino or other similar cards recognised as a FFB device ?

I will have to go back on my venerable force feedback 2 .

Electrical question for you guys who seems to have a lot of knowledge ,not related directly to the subject but..:

- modding a sidewinder force feedback 2 stick to increase the current in the 2 DC motors seems a little bit tricky according to :http://www.simprojects.nl/ms_siderwinder_ff2_hack.htm

Is there a way to just pick up the signals arriving in the terminal DC motors of the ffb2 ( and keep them working ) and i via an amplifier power supplied increase the the current output (ampliefed but the exact same signal )to use stronger DC motors .

I don t understand why Roland Van hoy (ffb hacking ) made all this complicated hack instead and if this idea , he certainly have good reasons but perhapse it is possible ?

Hi

The Test Firmware is turning an ArduinoPro Micro into a FFB Joystick already.

But not all Force Effects are implemented.Sinewave form is missing for example.

For the Hack:

I would not recommend doing that withoutenough knowledge and Routine in soldering PCB.

For getting the PWM signals (Motor Directionand strength):

Looks possible but I am not sure.

Could be that the Controller shuts itselfdown when it recognizes that the Motors are missing or not responding withthe expected current.

In that case you would need to feed it withamplified Current readings from your new Motors.

MetalGear_Honk

Nope, see on Reddit:

 

 

http://imgur.com/a/zfItq

It's a prototype designed by Iris Dynamics, which currently has a force feedback yoke in production. Here's their page on their force feedback joysticks:

https://vs4.irisdynamics.com/magnetic-force-feedback-joysticks/

There's a mailing list that you can sign up for:

https://vs4.irisdynamics.com/magnetic-force-feedback-jostick-mailing-list/

I e-mailed them to ask for details, they said that they're working on a FFB joystick for a commercial client, but are considering selling to the general public. They're hoping to release it sometime this year. The price point is unknown, but if it's anything like the yoke, it's gonna be $1,000+.

Thank you again Metal Gear !

Arduino is ordered , your firmware downloaded !

Impatient to have a try .

For the ffb effects in dcs i am pretty shure that there are only spring + - ( for speed + - ) for warbird , spring non related with speed for jets , helicopter , shaking for stalling , force trim for helicopter ( wich is spring effect in a certain position ...No friction or damping

So spring and shake ( on my force feedback 2 ) . If not all the effect are implemented on your firmware that is not realy a problem for dcs !

For information the only 2 ffb "professionnal " and powerfull force ffedback flying stick "on the market" are one from Brunner ( they make a very good yoke two ) :price about 8000 dollars ;( and the one with a magnet ffb from iris wich is under developpement..

It is definitively important to be able to build our own OSJ at a fraction of the cost !

If you have recomandation for the base mechanism of the stick ( 3d printable plan ) , the hardware needed ( gear or belt ) and the best choice for the Motors in therm of torque and power we can purhapse make a step bu step tutorial for other simmers !

Regarding the mechanism I have a question. What is the required minimum output RPM of the motor+gear combo to have short enough reaction time and proper ffb effects.

Hi

A quick reference for the Hardware

1. I use Small Brushless DC Motors. (I hadthem around) for testing.

2. I use 3d printed Gear (1:25 stepdown)for strength

3. The Motors are 12V powered by an ATXPower Supply.

4. Motors are controlled through anH-Bridge (L298 )

The L298 has 2 H-Bridges so you can wire both Motors to it

For the Gimbal:

I am still working on that part. I don'thave any FFB device so I cannot simply replicate something. And my test Gimbalsare not working very well.

Maybe someone with a working FFB can sendsome pic’s from the mechanics?

For the Force effects:

Spring, Damper, friction, constant, Rampand Square are implemented and hopefully working properly.

Sin, triangle, saw up, saw down and customare set in the HID but not working.

If no FFB Signal is send from the PCinternal spring should be active.

I am still working on the effects but at the moment the Gimbal is more in focus.

For optical tests ii use some LED andPotis:

4Red for Direction of Motor two foreach. And two green for strength. (PWM Visual on LED is Nonlinear)

That way I can test without always usingthe not so functional Gimbal.

MetalGear_Honk

Hi MetalGear,

Please find below several high resolution pcitures from both G940 and MS Sidewinder mechanism.

Let me know if you've been able to download them succsefully.

https://1drv.ms/u/s!AlRLImVJonHgiMdUdOunEQfMsbz-Xg

Hi

Thanks abburo got them.

MetalGear_Honk

Hi

I like the idea of a DIY FFB stick (had the same idea while I was writing a DirectX DirectInput plugin with FFB support for UE4)

Do you also plan to use\write a DirectInput interface for your firmware?

Bixio

Gesendet von meinem SM-G930F mit Tapatalk

Hi ,

For the Gimball : as soon as possible i will take my venerable Sidewinder ff2 appart and scan the mechanism , gears to make a 3d printable version .

Advantage : 1) after 14 years of punishement mine as not a single lack in the mechanism

2) i think that we can easely putt 2 time more powerfull dc motors without a problem on the mechanism using a hard plastic

3) cost effective method

4) modification easy to do on a 3d plan ( adapt the warthog stick for exemple)

5) the engenering as already be done and well done manu years ago ...

Pb : don t know if it is possible to scan every pieces .. In my country .

Hi

@Bixio:

I am not entirely sure what you are asking.

if I get you right:

Direct input is the DirectX command set for using the DX build in interface to communicate with PID devices(FFB for example).

As I am trying to get the firmware PID conform it will natively be able to receive Direct input signals send through DX. (as long as DX supports it).

pedestals:

I don't think it is necessary to scan the parts (what might be a violation in Legal terms) but to get to know how exactly it is working.

I am studying the Pics from Abburo at the moment. the G940 Gimbal looks exactly like my last try. only that mine is locking when going diagonal.

so I must have missed something.

I am under the impression that one Axis connection must be flexible in "distance" can someone confirm this idea?

@VO101_MMaister:

sorry missed you post.

the best gear ratio-> RPM is still not cleared, and open for discussion.

MetalGear_Honk

@MetalGear :

"the creation, dissemination and offering to the public of the 3D replica risks to be considered an infringement of the design right, except for a private and non-commercial use, for an experimental use and for citations or education uses."

As long as this is private and non comercial puphose or in a protoype convept there is no problem apparently ..

And once again all the engenering has alreday be very well done ... If you just need to scale some parts for stronger behaviour for exemple that could be far easy .. I don t say this is the only solution but in therm of reliability and assembling..

I will have a try if i can replicate the ffb2 gimball and will share my experience .

Hi

Good to know. I am not used to legalrights. So I take it as is.

Jpdesvals, can you share some pic of the disassembledparts?

MetalGear_Honk