I would like to add my thoughts to this discussion:

When I was building my "Julian" stick (Made from a car U-joint utilizing hall sensors) I too struggled with the notion of incorporating a centering force of some sort. Of course, the logical design would have used some sort of spring tensioner to "re-center" the stick, right?

I was so eager to use this new build that I started to play the game with a stick that had no self-centering action built into it. (My idea was that I would need to plan the design before using it, and I became impatient)

What I discovered was that my new full-throw stick was much more intuitive than I thought it would be. Then it occurred to me...

Unless there is some game input attached to some sort of servo action that actually increases stick forces as forces are built up on the control surfaces of my computer plane, then it is all fake anyway!

A spring is a spring is a spring. As you move a spring centered stick away from center, the same amount of force is applied, pushing you back to center. It matters not what type of force would be on your computer generated control surfaces. I am sure, everyone is aware that stick forces are not always constant. The more force on a control surface, the more force on the stick (When dealing with cable and/or rod driven controls) That's why there is more force needed to pull out of a steep dive than to manipulate the same amount of deflection in normal flight.

So, what does a spring really simulate? Nothing. It merely puts the stick back in the middle. Whoopie... I can do that without a spring.

Springs only give a person a false sense of control force.

I could be wrong, but I am not sure if a servo driven stick could take inputs from the game and generate this type of true feedback into the stick. I am not a devicelink wizard, so I am not sure this is a correct statement.

So, with my stick, I have no centering force other than gravity. In level flight, I trim my plane so that my stick is in the middle of it's throw. Again, it is quite intuitive; even more so than I had dreamed. And because there is some amount of friction inherent in a U-joint, the stick tends to stay in the position that I put it unless it is leaning heavily to one side or another. That's how I determine the amount of trim needed by feel.

Perhaps there will be programming in the future that will allow someone to create a true force feedback stick, one that interacts with the flight game software and actually applied more force on the stick under the proper situations, and not just wiggle in your hand or just push back at the same rate all the time.

Now imagine THAT! Diving in your "Plane X" and having to exert 40 lbs of pressure on the stick to get "X" degrees of deflevtion on the elevators in order to pull out of the dive! And imagine only needing say 10 pounds of pressure to deflect the elevators the same "X" degrees in normal flight! THAT would be realistic and would be worthy of being called true "force feedback".

Springs alone cannot do that.

SO, really.. Discussion of using springs without some sort of game to stick interface that transfers the stick forces of the game into your hands is really a useless discussion.

quote:

Perhaps there will be programming in the future that will allow someone to create a true force feedback stick, one that interacts with the flight game software and actually applied more force on the stick under the proper situations, and not just wiggle in your hand or just push back at the same rate all the time.

I am not sure i get this part.

The MSFFB2 stick ( and its predecessors) already do a certain level of what you describe. Fly your aircraft at high speed and the force required to move control surfaces increases. Fly near stall speed and the forces required to move control surfaces is negligable or non existant, allowing flight by feel rather than guestimate.

Unfortunately the feedback sticks i have experienced sadly do not allow for feedback on the rudder axis. I don't believe any 3rd party rudder pedal does either. For me, this is what is missing in sims. Currently you can fly and trim by feel on elevators and ailerons, but without a feedback axis for rudder any hope of gaining an acceptable "seat of the pants" experience is failing.

Well see? Ya learn something new every day!

So you are saying that the stick forces in one of those FFB sticks actually increases as the simulated control surface force increases?

Well then all that is needed are servos strong enough to create 50-70 pounds of force, and the software to drive them I guess...

I always want to learn something new each day I live. I dont feel that my day is complete without it. But now that I have learned something new so early in the day... What the hell am I going to do with the rest of my time!!!

Thanks a lot! LOL!

No one has made a stick that is HARD to move from dead center in any of the 4 directions and that ramps up nicely giving a tactile feel to how much input your giving but doesn't become too hard to hold at max deflection.

Yes its always hard and doesn't change like a real planes stick would but you live with this fact like people with low tension sticks do(like CH users).

Moving a stick by releasing tension is way easier than moving a stick by muscle memory.Our bodies are in tune with force or I should say our muscle.

Do this test.Hold an imaginary joystick in air and push it forward ever slightly.Its not easy to move "slightly" since the body expects back pressure.You need alot of practice to be able to do this.

Now push against a desk with your fist until you have a nice backpressure and then imagine adding in a slight forward movement.You now have a tactile feel and by adding just a bit more force you get movement.

This allows one to not have to sit as steady as possible trying concentrating super hard on the sticks movement...instead you can relax and concentrate on your surroundings instead.

quote:

I was so eager to use this new build that I started to play the game with a stick that had no self-centering action built into it.

Fehler,

And if integrate a "deadman switch" in these spring-less stick?
That switch turn off X an Y response if you take your hand off the grip.

Sokol1

quote:

Originally posted by M_Gunz:
Fly with a light touch.

instead of using the higher tension spring, i stuck a hard rubber washer at the base of the stock spring. The tension itself is very consistent, there is just a bit more of it which gives it a nice feel. There is no increase in tension as i increase throw.
I really don't think the springs in some sticks were meant to simulate tension, but instead just makes sure your stick pops back to center.

Since logitech Attack 3s are so cheap and have so many buttons i often wondered if another could be used as a throttle quadrant.
Something like: Y axis = power and X axis = pitch
Taking the spring out of the stick means the stick no longer returns to center, but its not enough to hold the stick in place.

Are there ways to add tension to hold the stick in place and give it a nice smooth throw with out making it pop back to center?

Bill maybe a big glob of non conductive grease(for plastic also) jammed in would hold the stick up.Maybe also take off the handle so its just the knob and you may not need grease or whatnot.

BillSwagger

Better made "Joy-Thottle" with some modification:




The horizontal wood rail make friction to Y lever. X is put in pot with knob in left side of handle.

Sokol1

Sokol...try using the actual throttle pot.Its different quality on Sidewinder joysticks.I have had mine for so many yrs and while the pitch and roll pots spike the throttle pot is still rock steady.Its also coincidentally a different color.

I think a persons liking and use of a joystick and for that matter pedals are very subjective. If I understand you, wolf-striked, you are saying that you would have better control over the game if the joystick
acted more like the brake pedal of a car.

Because you would be using a strong opposing force to push against you would have greater precision than relying on greater movement with weaker forces.

This is undoubtably true. I also play race simms - GTR2 and Grand Prix Legands - and intend to build my own wheel, pedals and gear shift. I have purchased the strain guage version of Leo's board and intend to use a load cell for the brake. I will use a heavy spring - a car valve spring perhaps between brake lever and cell to give a small movement to transmit the force.

For me, the whole point of playing a sim is to make as much effort as possible to try and kid myself that I really am a Jim Clark racing driver or a WWII fighter pilot locked in mortal combat. For that reason I try and make my input devices feel and behave as near to the real thing as possible. For racing this is not so hard - I drive every day in real life and have even done a little motor sport at club level.

I have never flown a plane let alone a WWII fighter so I only have my imagination and second hand accounts to work with. On a purely subjective level I am very happy with my project to date and feel it goes some little way to replicate the movement, forces and contol I might experience in real life. Of course it's just a game and even if I went so far as to build a replica cockpit perhaps with motion built in and multiple displays etc. it would still fall a mile short of the real deal.

For me there are a couple of issues regarding input devices. Fistly they need to at least aproximate to the feel and behaviour (real or imagined) of the real life control. In my experience most run of the mill devices are pretty poor. You need to spend more than most care to - or build your own.

The second is the degree of input precision they have over the game. Most controllers have analogue output - potentiometers or hall sensors. This output has to be converted to digital to be used by the game ie. sliced up into descreet values. The game has a resolution or a set number of values it can recognise. Like a digital camera, the higher the resolution the better the picture.

The board in the device also has an output resolution 8, 10 or 12 bits for instance. The higher the resolutions the greater the potential for precise control.

Most devices pivot around an axis and have a degree of rotation. In the case of joysticks its usually less than 90 degrees. Most joysticks usually attach the pot or sensor directly to the end of the axis. This means only a fraction of the potential range is available to be sliced up into digital values hence the attraction of gearing to increase the range available. This is as true for hall sensors.

Joysticks are built to a budget to maximise profit. I don't know any that use gearing because it would need a high quality of build but then most simpler sticks suffer from some play or soon become become worn enough to lose precision over the limited range of movement. The Logitech Wingman was dire.

I would have thought the best and simplest solution to obtain wolf-striked's effect would be to attach a beam load cell directly to the end of the joystick shaft limited to one plane of movement with a second at right angles similarly limited to one plane. I don't know how much movement a beam load cell gives if any. It might need a slightly flexible joystick shaft?

The Saitek Evo has a mouse type wheel turning in a sensor which eliminates the problem of wear but because of the limited rotation it only moves about 16 windows through the sensor from lock to lock giving a very low resolution. Again this would benefit from gearing to give multiple turns - like running a mouse across a table rather than moving it 1/4 inch.

The best commercial stick I've used for precision is the Sidewinder which feels much like a gamepad joystick.

quote:

Originally posted by wolf-striked:

quote:

Originally posted by M_Gunz:
Fly with a light touch.

I guess you don't have the slightest idea who wrote that.

quote:

Originally posted by M_Gunz:
LOL, which is better, the optical mouse or the one that uses the mechanical ball and wheels? No contest!

My first ideas about building a custom stick actually revolved around using the sensor of an optical mouse for measuring stick travel. Would have been a very awkward construction (with some awkward software to go along with it). With those >1000 dpi mice we get these days one could get impressive stick resolutions, given the right mechanics. But in the end hall sensors and the BU368 are by far the more reasonable solution.

quote:

Originally posted by Fehler:
Perhaps there will be programming in the future that will allow someone to create a true force feedback stick, one that interacts with the flight game software and actually applied more force on the stick under the proper situations, and not just wiggle in your hand or just push back at the same rate all the time.

Good FFB Sticks like the Microsoft ones do not have any centering besides the motors and can be set up so that there is little or zero wiggling and only the game-controlled amount of centering.

I'm currently in the process of using this to control the amount of recentering force by inlining FFB motors between two springs, so that the motors will add more or less force to the pull of the springs. It's supposed to work like this:


Little motor force:

  
  #--fixed-----|/|/|/|/|/|-------------.
                                  /Motor\
                                 |   0   |
                                  \     /
<--to-stick----|/|/|/|/|/|-------------´

More motor force:

  
  #--fixed-----||||||||----------------.
                                  /Motor\
                                 |   0   |
                                  \  -> /
<--to-stick----|/\/ \ / \/\/|----------´

The motors are driven by regular FFB electronics that are tricked into applying recenter force by constantly putting the pots into an off-center position (will be interesting to get around auto-calibration)

The motor configuration looks like this. The string is a dummy representation of the wires connecting the springs, the ball bearings are used to take load off the axle.

Right now i only need to find time to get me a useful collection of fresh springs, because it will certainly take some experimentation to get the best balance between springs and motors.

quote:

Originally posted by wolf-striked:
No one has made a stick that is HARD to move from dead center in any of the 4 directions and that ramps up nicely giving a tactile feel to how much input your giving but doesn't become too hard to hold at max deflection.

What you describe is not centering force but inertia. You might just use a heavy long throw stick that is made from steel instead of plastic and you are set. Moving parts on my bicycle-based stick are probably about 2 kg total, i guess many of the unijoint sticks are considerably heavier.

I wouldn't know how to get a mouse to be seen as a stick axis. To mount the mouse steady and move a section
of a disc past it would be clumsy-big even if the cover was stripped from the mouse.

I have experimented with IR LED and IR sensor, passing the light through clear plastic printed with B/W gradient
at a local office store. 50 cents gets a whole sheet copy laser printed on clear plastic, LOTS of strips that way.
I tried first to reflect the IR from white paper with black-ink gradient but LOL the IR reflected the same from
white or black which I didn't expect! Maybe the 900nm near-IR is exceptional, I dunno but the LEDs only ran me
20 cents each and the idea of being able to vary the gradient to change the sensor output allows flexibility.
The sensors were under 30 cents and the current draw for the whole setup is very low, I got good results on the
meter running about 10mA at 4.8V (3 AAA's through 470 ohm resistor plus the LED) through the LED. Haven't gone
past breadboard and anything I'd use regularly though.

Using a LED through a square aperture and sensor I can measure total movements of an arm down to 2 or 3mm pretty
close to linear. If I could use a lead screw to block the light then very small rotations could be measured.
That makes me think... many old floppy and CD drives do have small lead screws in mounts.

Solenoid for FFB actuator?

No, they're not all suitable.

Last bit, you -might- have the stick springs press on piezo discs and pick up the potential on that with a FET.
I'd bet they use piezos of some kind in the new Saitek, but no springs. It is nice tight solution and you can
measure even very small forces that way.

A simple demonstration.

Not sure how to describe this, maybe this is what you are already talking about, but what about the idea of using something similar to sensors on a digital scale.

Essentially the stick hardly moves, but it might be better to have some give and maybe load it with a very heavy spring, so at least it gives the feel of stick and not just this handle that you push and pull against. Pulling on it will compress on a sensor that measures how much force is being used. So it could give a measurement in grams or ounces, but instead the signal would go to the driver or software where the information is calculated into stick movements.

So if you can imagine taking a digital scale and pressing on it, the higher the number that shows, the further the stick movement would be.
This way would be more relative to forces achieved during flight. For example, in a game you could program 80lbs of stick force, which could be measured and calculated by this type of stick.
Currently in Il2, you might notice on some planes at higher speeds there is little elevator authority because the pilot is programmed to use 50lbs of stick force.
Imagine if the stick forces were measured by the actual strength of the players?
So maybe it takes 60lbs of stick force to recover a 109 from a dive, but if i'm able to pull 90lbs using both hands, the recovery might be much sooner than what the current engine simulates.

The other possibility with this stick is that regular maneuvers might only require 30lbs of force, while at higher speeds you might need more strength to get the same results. Trimming off forces would then be necessary or your arms would tire.


This type of stick is probably way ahead of its time, and would require gaming/sim developers to also work a format into their code that would allow the stick forces to actually be measured and used properly in the game, instead of having pre programmed stick forces like in Il2. Full real settings, could also mean full real stick forces.



Bill