PDA

View Full Version : Confessions of a joystick builder



rnzoli
04-09-2006, 10:26 AM
...here are my sins... http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif


<span class="ev_code_YELLOW">NOTICE: This thread starts with a very old post from 2006. To see more recent information, go straight to </span>page 3 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/3791063134?r=6521092685#6521092685)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/1-Overview-left.JPG (http://web.axelero.hu/rnzoli/JoyBuild/1-Overview-left.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/2-Overview-above.JPG (http://web.axelero.hu/rnzoli/JoyBuild/2-Overview-above.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/3-Throttle-and-trim.JPG (http://web.axelero.hu/rnzoli/JoyBuild/3-Throttle-and-trim.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/4-Pedal-front.JPG (http://web.axelero.hu/rnzoli/JoyBuild/4-Pedal-front.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/5-Pedal-side.JPG (http://web.axelero.hu/rnzoli/JoyBuild/5-Pedal-side.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/6-Stick-alone.JPG (http://web.axelero.hu/rnzoli/JoyBuild/6-Stick-alone.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/7-Stick-closeup.JPG (http://web.axelero.hu/rnzoli/JoyBuild/7-Stick-closeup.JPG)


After a long project lasting more than a year, I have finally built a very accurate HOTAS system with Force Feedback effect and pedals with toe-braking feature. I built all the mechanics around the electronics and the grips of 2 old Logitech FFB stick that were supposed to be thrown into the garbage long time ago. It is hard to describe how happy I am, good equipment is something that is essential for playing IL2, and at times, I wasn't sure I get this far with this project.

If someone is interested I can type up the long story behind it and/or the technical parameters, or post more pictures.

Summit17
04-09-2006, 10:33 AM
Nice

BueJack
04-09-2006, 10:40 AM
Velly noice.. my project has lasted 6 month's so far, just need my HUD helmet. http://forums.ubi.com/images/smilies/25.gif

tHeBaLrOgRoCkS
04-09-2006, 11:38 AM
mwuahahahah its alive its alive http://forums.ubi.com/images/smilies/53.gif erm nice post ...i'll get me coat

GSNei
04-09-2006, 11:50 AM
Wow - somebody suggested a sticky thread for stick mods - I think that's a great idea.

triggerhappyfin
04-09-2006, 03:47 PM
This was really interesting approach on the stick...any close ups on the mechanism?

Jatro13th
04-09-2006, 04:04 PM
GOOD GOD!!!!!! That's some amazing craftsmanship Rnzoli!!! PM me with the details plzzzzzzzzz!!!!!

rnzoli
04-09-2006, 04:10 PM
here they are (hi-res):

http://web.axelero.hu/rnzoli/JoyBuild/IMG_1924.JPG
http://web.axelero.hu/rnzoli/JoyBuild/IMG_1925.JPG

In summary, everything moves on ball-bearings. Soft, steady, doesn't wear, and doesn't become loose over the years. Yes, but heavy, heavier than plastic.

The angle detectors are slightly bent resistance strips from linear 100K pots... so simple concept, almost trivial, but the idea occurred to me only after months of headache and other sleepless nights http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

willyvic
04-10-2006, 05:10 AM
Amazing.


Bump

triggerhappyfin
04-10-2006, 05:10 AM
Tnx for posting close ups. Nice approach m8.

JG52-6High
04-10-2006, 05:28 AM
http://forums.ubi.com/groupee_common/emoticons/icon_eek.gif

Whoa-that is mechanical rock`n`roll at it`s best!!

6

triggerhappyfin
04-10-2006, 06:29 AM
Particularly iI´m fond of the idea of attaching the stick at the table. http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

BaldieJr
04-10-2006, 07:18 AM
The angle detectors are slightly bent resistance strips from linear 100K pots... so simple concept, almost trivial, but the idea occurred to me only after months of headache and other sleepless nights


very nice idea.

TgD Thunderbolt56
04-10-2006, 08:14 AM
That's too cool. http://forums.ubi.com/images/smilies/25.gif

Wild.Bill.Kelso
04-10-2006, 08:27 AM
Yes! That is very Cool!

I love seeing the stuff people build to play their sims.

A Sticky Thread is a great idea for these home built contraptions!!!

rnzoli
04-10-2006, 09:54 AM
Originally posted by Jatro13th:
That's some amazing craftsmanship Rnzoli!!! PM me with the details plzzzzzzzzz!!!!!
Thanks for the compliment. I think this runs in the family. My elder brother and our father are the same DIY men. I remember using their tools when I was a kid, and getting hooked on doing stuff for fun. Too bad my father died back in 1999, if he can see this from Heaven, I hope he is giving this a thumb up, too.

But back to the present: you don't think I have any blueprint, do you? http://forums.ubi.com/images/smilies/halo.gif I had a few sketch drawings, but no CAD/CAM stuff http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif Much more exiting that way http://forums.ubi.com/images/smilies/35.gif



Originally posted by triggerhappyfin:
Particularly iI´m fond of the idea of attaching the stick at the table. http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif
This is indeed very important, the stick base is rock steady this way. The downside is that the FFB effects rattle the entire table, so the family next room can always hear when I am in a heavy fight... http://forums.ubi.com/images/smilies/crackwhip.gif



Originally posted by BaldieJr:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The angle detectors are slightly bent resistance strips from linear 100K pots... so simple concept, almost trivial, but the idea occurred to me only after months of headache and other sleepless nights
very nice idea. </div></BLOCKQUOTE>
Also cost-efficient. Accurate angle detection is literally life-or-death question in IL2, but the professional angle detectors cost an arm, a leg and a head, too:
- high durability pot ~40 USD/piece
- Hall-effect sensor ~90 USD/piece, (http://www.hirevsims.com/01prod/fltcontrl/halleff/halleff01.htm, this is probably used for Cougar mods)
- optical PSD sensor ~150 USD/piece (http://www.sitek.se/1d.htm)
...and we need 3 pcs for roll/pitch/yaw axes...

On the other hand, a normal linear pot is just ~3 USD or less. Wow, I saved lots of money with this... http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

GreyFox5
04-10-2006, 10:22 AM
Very cool setup! Very professional look as well. Have some questions...

I take it your interface to the PC is USB. How are you setting up your Stick and pedals on the PC? Many questions here...

Chivas
04-10-2006, 10:48 AM
Very interesting. Your right FB requires a very precise joystick and I've been rippering apart all the joysticks I own trying to improve them. There isn't a joystick on the market thats any good.

AFJ_Locust
04-10-2006, 11:02 AM
Originally posted by Chivas:
Very interesting. Your right FB requires a very precise joystick and I've been rippering apart all the joysticks I own trying to improve them. There isn't a joystick on the market thats any good.

Chivas you tryed CH Products yet ?

rnzoli
04-10-2006, 12:45 PM
Originally posted by GreyFox5:
I take it your interface to the PC is USB. How are you setting up your Stick and pedals on the PC?
Yes, the 2 base units plug in with 2 separate USB cables into the PC. In the control panel the PC recognized 2 joysticks (with the same name in my case). IL2 assignment is fairly easy - Hardware setup/HOTAS section/click on an "axis" (similar to buttons)/move the corresponding slider/axis to register the assigment.

The only complication I had with joystick IDs. When I plugged in the separate throttle/trim, it 'stole' the assignments from my main stick, so I could only fly with the 'trim' sliders, not the stick http://forums.ubi.com/images/smilies/blink.gif But I only had to swap the USB plugs to correct this, so I retained my original settings for the stick, and added the throttle/trim controls for ID2 buttons/axis. meaning that IL2 will indicate the second joystick like "ID2 X-Axis", instead of the usual "X-Axis".

hugso
04-10-2006, 01:49 PM
That is superb. Your seat looks mighty hard, tho. Go on, admit it, you sit on a parachute, don't you?

I thought I was the only one crazy enough to spend months tinkering with hardware, have just about perfected my pseudo-G effect. Two strips of inner tube on seat and back rest, inflates as stick pulled back, pressing up into you, pressure drops when stick pushed forward, feels like seat falling away from you. Side to side tilt also. Really pretty good, but a lot more complicated than I anticipated getting sensitivity right, rate of inflation etc. Nice project over the dark winter evenings.

Just love your joystick, very tempted to do something similar.....

rnzoli
04-10-2006, 02:24 PM
My "seat" is quite unconfortable, so it works as a natural limiter on how much time I should spend with IL2 http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif Maybe the the rubber tubes of your pseudo-G simulation would solve that?

Seriously, I hope you will post more info about that, when ready. I just love tinkering with these things, it makes a creative past-time activity, especially during winter nights. Maybe this is the reason why suddenly there are so many threads about hardware mods....people are coming out with the results http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

Sokol__1
04-10-2006, 05:55 PM
but the professional angle detectors cost an arm, a leg and a head, too:

A promising substitute for pots in circuits USB.
Hall sensor of Honeywell SS490 series.

2 for 1 $
neodymium magnet: 3 for 1 $

Look this tread in RSnet forum:
http://forum.rscnet.org/showthread.php?t=248618

http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

Sokol1

WWMaxGunz
04-10-2006, 06:15 PM
I've bought electronics through Digi-Key in the US and they tend to have selections with
good prices. I see Hall sensors with built-in IC's to do different functions cheap but
the output may only be digital even though it has an amplfier stage. One such function
that seems common is geartooth counting and speed.

We have optical sensors in mice, the wheel spokes run through, they are double gates to get forward/backward. That could just as well be a notched edge on a length of metal. Is there
another cheap IC that can count and output analog from that, esp as resistance? You get 4
states per notch out of 2 gates alone, that multiplies the resolution of linear movement.
I wonder can a 555 do it?

SATAN_23rd
04-10-2006, 06:19 PM
Damn,that looks like a lot of work.

Sokol__1
04-10-2006, 07:50 PM
WWMaxGunz,

Optical encoders by mouses are good for things likes trims, sample:

http://images.electronicsinfoline.com/View/Photo/10001/end_stop.jpg
http://users.on.net/~merrifield/opto/index.html (http://users.on.net/%7Emerrifield/opto/index.html)
The pot case are only for axis an knob suport.

Used by FS9 cockpit builders for "radio" coms, too.

http://img66.imageshack.us/img66/7817/encodermouse4jy.jpg

A friend made this gadget for FS9, a mouse weel encoder (mecanical type found in cheap Chinese mouse) hook in PCB of cheap USB gamepad.
If u move weel up, press one button, in endeless steps.
If move move weel down, press another button, in endeless steps.
Trim in steps, like ones in R/C transmiters.
http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif
BTW - The captions are in the "boring" Portuguese.

But i think that for flying controls u need a more precise device, maybee likes this encorders (8000 positions per revolution):
http://www.motion-control-info.com/express_encoders.html
Proposed by LBodnar for BU0836 in Frugalsworld. http://forums.frugalsworld.com/vbb/showthread.php?t=98831

Well, more complicated too.

Rnzoli, excuse for the deviation of his original theme. Do you have schematic of the sensor used in their stick?
And congratulations for the nice work.
http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

S!

Sokol1

rnzoli
04-11-2006, 04:03 AM
Yeeehaaa, now it gets interesting! Accurate angle detection! http://forums.ubi.com/images/smilies/metal.gif


Originally posted by Sokol__1:
A promising substitute for pots in circuits USB.
Hall sensor of Honeywell SS490 series.

2 for 1 $
neodymium magnet: 3 for 1 $

Look this tread in RSnet forum:
http://forum.rscnet.org/showthread.php?t=248618

Interesting, indeed! I have looked at Honeywell sensors last year, and there was one issue that I was afraid of, namely the higher current required to drive the simpler (cheaper) sensing elements. In the specs I read that 7 mA needed from 5V, that corresponds to R=5/0.007 = ~700 Ohm resistance. Now, my original pots were 100 KOhm, taking far less current, so I wasn't sure, how the re-used Logitech electronics would cope with that increased load.

But this is worth to investigate and test, I would definitely go for this type of cheap sensing element/magnetic actuator, if I build the electronics myself, too. Maybe at a later stage? http://forums.ubi.com/images/smilies/88.gif



Originally posted by Sokol__1:
But i think that for flying controls u need a more precise device, maybee likes this encorders (8000 positions per revolution):
http://www.motion-control-info.com/express_encoders.html


This I also explored briefly before. For joysticks, the requirement is that the sensor should provide absolute values, minimum 512 positions over a 60 degree movement, so 360/60*512 => ~3000 positions per revolution would work. IIRC I was put off by the price (higher resolution gets more expensive), and the difficulty to getting the digital discreet signal into the available electronics (D/A conversion, only for allowing the electronics to do reverse A/D conversion again http://forums.ubi.com/images/smilies/16x16_smiley-sad.gif).



Originally posted by Sokol__1:
Do you have schematic of the sensor used in their stick?
If you mean my stick, no, no drawings, but maybe it is simple enough to explain on the below picture:
http://web.axelero.hu/rnzoli/JoyBuild/7-Stick-closeup-explained.jpg
Seen from a different angle (on the left side):
http://web.axelero.hu/rnzoli/JoyBuild/7-Stick-closeup2-explained.jpg


You may ask, why this complication and not wiping the resistance strip directly, as normally pots do? The answer is: because of durability. With the ball-bearing in between, the metal ring rolls over the fragile carbon resistance layer, no friction (only a little pressure from spring-loading). The metal wiper contacts the metal outer ring, so the friction takes place here (between metal and metal), not between the metal and carbon resistance strip.

There is one possible downside however: I am not sure, how much contamination will be taken by the open resistance strip. Maybe a textile dust-cover would come handy, but no problems so far (concept used in past 3 months).

WWMaxGunz
04-11-2006, 10:37 AM
A home made rheostat. You know that temperature affects the range? Also other things?
But it should work just fine.

I am not sure that high precision of the mechanism will translate into the same in the PC.
I've never seen other than a fixed range of integer values down in the code or PC hardware
specs. The calibration sets the translation last time I looked.

What I'd like to see if the range can be widened tremendously is the adoption of CD guts
to read a data track on a disc to determine angle. Even a track on mini-disc would put
most else to shame and the repeatability... how many bits per track close? Yeah, then
comes the Blue Ray DVD sensor for aiming at moon craters from ground.

DarkCanuck420
04-11-2006, 01:00 PM
impressive

rnzoli
04-11-2006, 01:30 PM
Originally posted by WWMaxGunz:
A home made rheostat. You know that temperature affects the range?
Yes, it is similar to a rheostat due to its large size. I hope however, that the temperature doesn't affect its range too much, because the resistive element is from a potentiometer, not resistive wire (sensitive to temperature).


Originally posted by WWMaxGunz:
I am not sure that high precision of the mechanism will translate into the same in the PC.
I've never seen other than a fixed range of integer values down in the code or PC hardware
specs. The calibration sets the translation last time I looked.

The fixed range is basically the resolution of the analog->digital converter. Most simple joysticks (e.g., MS Sidewinder Precision) employ an 8-bit converter, so the joystick sends maximum 256 discreet values to the PC via the USB bus.

In my case, the Logitech electronics send ~170 discrete values to the PC. (Out-of-the-box it sent only about 90 different values, I nearly have double resolution now). The throw lenght of the main trigger button is ~22 cm, so the physical resolution is about 1.2 mm per different value, which seems to be adequate (my hand isn't much more precise than that). Although I suspect that the sensor would provide steady and smooth analogue signal for ~500 discrete values as well (measured with a digital voltmeter from 0.00-5.00 Volts).

Another thing I checked was update frequency while moving the stick. The joystick electronics sent a change no faster than once in every 8 ms normally, so with fast movements, there were samples "missed out". I think IL2 can extrapolate, but there was something really strange: at one point, there must have been some congestion on the USB bus, because the updating missed samples for more than 100 ms. Which brings in an interesting issue: with high resolution joysticks like the Cougar (4000 discrete values?), frequent position updates take a lot of bandwidth, and maybe even the game isn't able to handle all the samples individually. Food for thought, just in case someone wants to build electronics as well http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

hugso
04-12-2006, 02:59 PM
Are you using FFB motors from a dis-membered commercial FFB joystick? If so, are they powerful enough, seeing as your joystick is 3 or 4 times as long as a shop-bought model?

What I really want is separate signals from in-game, one to give appropriate stiffness, and maybe some vibration at stall, to the joystick, and another set to give bumps, thumps and vibrations to feed to Bass-shakers in my seat. My crude system of filters works quite well, and it's great to feel the bumps of touchdown and rolling on grass strips, vibration when flying with flaps down, turbulence etc, but it would be better to have the signals derived in- game. In fact there would be scope for even more variety of jolts and vibrations. They really add a lot of immersion when felt thro your seat. Unfortunately, I am not a computer bof.

Keep up the good work.

Pig_Mac
04-12-2006, 04:51 PM
rnzoli, do you have any 'blueprints' of them rudders? I'v been trying to get a pair of decent ones made for me .. but the simpeds guy seems to be sick atm, and have a really long wait. Can you make me a pair or toss me the instructions on how to make them?

My 'mecanicalminded' mate can hardly move his right arm ATM, after a little accident. So I guess I have to borrow his workspace and try create something. I'm sick and tired of using pedals from a steering wheel set! and too poor to toss out more then 150$

I really like your setup!

WWMaxGunz
04-12-2006, 09:22 PM
Sokol_1, it is not hard to increase resolution on that device in your top picture.
Only needs a larger wheel simple to make with inkjet transparency sheet and printer
and optical reader moved outwards. The wheel only has to interrupt light and be
less precise than photo printers do now. How big a wheel is needed when the 2 gates
of a mouse encoder light bridge return 4 states = 4x resolution of the spokes?

Hard part for me is find another cheap IC or circuit that counts up and down and makes
resistance or voltage from that count. Or perhaps something that dispenses with A/D
stage altogether though I don't know the serial protocol the A/D uses? The latter is
the cleaner and more dependable by far, will always know absolute position within the
resolution of the wheel.

Linear sensing, just print a strip instead of a wheel and travel strip or encoder along.

What I'd most love to see would be drivers for homemade devices. Something that can
read serial/USB characters and PC sees that as stick input. So far we only have stick
via A/D port of soundcard though I think MIDI is capable of much more. We also have
SerialKeys designed just to interface but I see no joystick acting commands there.

rnzoli
04-13-2006, 09:51 AM
Originally posted by hugso:
Are you using FFB motors from a dis-membered commercial FFB joystick? If so, are they powerful enough, seeing as your joystick is 3 or 4 times as long as a shop-bought model? .
Very good observation. Yes, same motors. With this setup, they aren't able to keep the stick in center position, so there is a static centering spring as well, under the stick. However, the motors are powerful enough to shake the stick well (gun recoil, stall warning, turbulence etc.) I considered the possibility to use gearbox assembly, thereby increasing the 'strength' of the FFB, but since the electronics don't provide good resolution for centering spring forces (only ~16 discrete forces along the axis travel http://forums.ubi.com/images/smilies/bigtears.gif), I haven't bothered.


Originally posted by hugso:
What I really want is separate signals from in-game, one to give appropriate stiffness, and maybe some vibration at stall, to the joystick, and another set to give bumps, thumps and vibrations to feed to Bass-shakers in my seat.

I complely agree with you, this 'combo' FFB is not realistic, the guns were not shaking only the stick, but rather the entire aircraft.


Originally posted by hugso:
Keep up the good work.
http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif You better not say this. http://forums.ubi.com/images/smilies/10.gif I promised my wife to stop for a while... http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

rnzoli
04-13-2006, 10:07 AM
Originally posted by Xallo_72:
rnzoli, do you have any 'blueprints' of them rudders? [...] Can you make me a pair or toss me the instructions on how to make them?
No, unfortunately this was a trial-and-error approach, no blueprints or instructions were created. This is also a problem for making them, it's way too time consuming to copy it from the first one (instead of manufacturing it from proper design documents).


Originally posted by Xallo_72:
My 'mecanicalminded' mate can hardly move his right arm ATM, after a little accident. So I guess I have to borrow his workspace and try create something.
Speaking of accident, beware that DIY with metal is less forgiving than working with wood. I had a few small accidents during this project, involving small amount of blood and a large amount of heavy cursing.


Originally posted by Xallo_72:
I really like your setup!
Good to hear, me too! http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif I was absent for 5-6 weeks from online combat, because of doing this, as well as test-flying it. Recently I came back for the real test, and it is a dream that came true. It is a big psychological difference, when you change flying from your wrist (hand) to flying from your shoulder (whole arm). Sure, this is no ace-maker, but opened the way for more learning, like long-distance deflection shots, proper tactics in a low/high altitude combat etc. And when a real good guy tells you that you were fighing like an ace in an engagement, that's nice. Helps me accepting that I get shot down 9 out of 10 times anyway http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

notamuppet
04-13-2006, 11:06 AM
Hello have a look in the simpit & stck mod sticky in community help http://forums.ubi.com/images/smilies/25.gif

hugso
04-14-2006, 12:40 AM
It should be possible for some computer programmer genius out there to separate the "stiffness" information from the "bumps and thumps" information in FFB. Occasionally my graphics card (ATI 9800 Pro) crashes. Usually it re-sets itself after a few seconds but, here's the thing, when the games continues the stiffness has gone from FFB - stick completely floppy, but flies OK - but the bumps and thumps still function.

What I want is a way of doing this to order and being able to use the two separate signals. Way beyond me, but theoretically possible.

WWMaxGunz
04-14-2006, 12:57 AM
If there are not two sets of wires to the motor (one for stiffness, one for bumps) then the
problem would be in the driver and good luck there.

rnzoli
04-14-2006, 01:35 AM
interestingly enough, the FFB joystick tension and the rumble parameters are separete commands to the joystick firmware via the USB

so i also consider this possible, by writing a custom joystick driver and corresponding firmware, sending only the tension parameter to the stick, while sending the rumble parameters to the seat

more information in this doc: http://www.usb.org/developers/devclass_docs/pid1_01.pdf

(search for "spring")

WWMaxGunz
04-14-2006, 04:32 AM
Mzoli, you have the development capability to write drivers? Down on the 0 level?

Look, devicelink provides data for the bumps but to be able to read optical up-downs and
translate to stick position --- that kind of thing would be a boon to hundreds if not more
hobbyists.

I can only hope you can do W98 as well as W2000/XP. If you can get at the parallel port,
oh the bits both in and out there are many. I've never gotten to them with Windoze but
back in the simpler DOS days that was my favorite interface!

BTW, I did find cheap digital pots but unfortunate the ones with simple up-down I/O have
low resolution (64 bits one is 45 cents US) while the 1024 settings ones use I2C or SPI
and run US$2 and up. But with a proper driver there is no need.

rnzoli
04-14-2006, 07:25 AM
Originally posted by WWMaxGunz:
Mzoli, you have the development capability to write drivers? Down on the 0 level?
Sadly, no. My knowledge comes from my research I did, more than a year ago. Originally, I considered building and programming the joystick electronics as well. But I realized that I lack the time and experience. I do have access to Visual Studio, so could program and complie in C and C++ languages, I have seen articles about "how easy to write a Windows device drivers in 10 minutes", but that's all.


Look, devicelink provides data for the bumps but to be able to read optical up-downs and translate to stick position --- that kind of thing would be a boon to hundreds if not more
hobbyists.
I am not sure about this, because devicelink has an illogical limitation to off-line play, which is such a major discouraging factor in my book, that I "boycott" ( http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif) any work with it, although the separate instrument panels availabe for it already, are beautiful. Still, not much value, if only good for offline, because I would get spoiled in off-line and then terribly miss the feature on-line.

Bearcat99
04-14-2006, 08:58 AM
Zoli that is some great stuff there.

Taylortony
08-29-2007, 05:05 PM
Originally posted by Chivas:
Very interesting. Your right FB requires a very precise joystick and I've been rippering apart all the joysticks I own trying to improve them. There isn't a joystick on the market thats any good.

Actually there is...... And I have it http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif all i need to do is spend some time in figuring out the wiring, then I will eventually rewire a plug to connect to it and into the parts off my old x45 http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif but it wont be force feed back.

You need to think out of the box on this......

I bought one of these, (not the one shown)

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=160143472430

It is a Tornado Item I believe, and has all the gimble and sensors for it built into the grey section..... NOW THIS IS THE REAL THING so there is next to no play, the thing is accurate, it has to be, and it isn't gonna fail anytime soon or you would have Tornados dropping like flies........... I simply have to work out the wiring to the plug on the base and knock up a stand for it, it could even be on a post if needed and floor standing http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

http://i14.ebayimg.com/03/i/000/ag/5d/10d9_1.JPG

I just wish I had bothered to buy more than one.

I could slip the cuff off and take a pic if you want to see how a real gimble is done..... its actually quite clever and not that many parts

BaldieJr
08-29-2007, 06:12 PM
Pics please. I'd love to see how the real-deal works.

ddpairborne59
08-29-2007, 10:42 PM
Very, very nice Mr. rnzoli, http://forums.ubi.com/images/smilies/11.gifhttp://forums.ubi.com/images/smilies/11.gifhttp://forums.ubi.com/images/smilies/11.gif I see alot of excellent potental here to keep the flight sim pit ideas alive with your brilliant setup.

Amazing what you can learn on this forum, from all the sharing and ideas. ~S~ http://forums.ubi.com/images/smilies/25.gif

cmirko
08-30-2007, 03:58 AM
Originally posted by Taylortony:
Actually there is...... And I have it http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif all i need to do is spend some time in figuring out the wiring, then I will eventually rewire a plug to connect to it and into the parts off my old x45 http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif but it wont be force feed back.

You need to think out of the box on this......

I bought one of these, (not the one shown)

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=160143472430

It is a Tornado Item I believe, and has all the gimble and sensors for it built into the grey section..... NOW THIS IS THE REAL THING so there is next to no play, the thing is accurate, it has to be, and it isn't gonna fail anytime soon or you would have Tornados dropping like flies........... I simply have to work out the wiring to the plug on the base and knock up a stand for it, it could even be on a post if needed and floor standing http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

http://i14.ebayimg.com/03/i/000/ag/5d/10d9_1.JPG

I just wish I had bothered to buy more than one.

I could slip the cuff off and take a pic if you want to see how a real gimble is done..... its actually quite clever and not that many parts

wow http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif - that a nice stick - can't find any more though http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif - maybe its just my clumsy ebay search capabillity http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif - if you provide another link i will for sure bid on that http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

ploughman
08-30-2007, 05:21 AM
Wow, TT, what a neat idea. Of course those sticks have now become as ubiquitous as rocking horse poo. I notice you can buy a Sea Harrier on ebay for 28,000 (no joystick in her though, no surprise there).

rnzoli
08-30-2007, 06:01 AM
Originally posted by Taylortony:
I could slip the cuff off and take a pic if you want to see how a real gimble is done..... its actually quite clever and not that many parts
This also crossed my mind (I was wondering, how come that Airbus pilots don't complain about joystick jitter, and other troubles http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif), but I was under the impression that with aviation-grade materials, this should be easy to manufacture, however, at a vey high price (new part). But getting this from old planes is a clever idea and maybe the best alternative for reasonable-priced, still aviation-grade joystick.

Could you please take some photograph about the mechanics? http://forums.ubi.com/images/smilies/inlove.gif

Also, do I understand it correctly: it's the mechanics only, no electronics are included in the stick base?

Taylortony
08-30-2007, 02:54 PM
I will take some pics on the weekend and post them, he did have several for sale, i picked up the last one, so cost me more http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

the Type and number gave it away.

RAF Aircraft Joy Stick MBB CO415

EX RAF MOD Surplus

Type mrca nhc adv


MBB is the German Company Messerschmitt-Bolkow-Blohm

Under type MRCA was the designation for the Multi Role Combat Aircraft, which became the Tornado, ADV was the Air Defence Variant.......
So thats what I went by..

Rnzoli, as it's a self contained item if it's done on hall effect sensors they should be in there too or even switches......... As an Aircraft Engineer sourcing a plug to connect to it is not a problem, indeed one is coming to me from an airline for free.... so it's just a matter of finding what input does what on the pins and cross connecting them to the board from a commercial gaming stick.......


Hmm I dont want to give to much away, as another may appear on the market and I do not want to lose out.......

What I will say is on another Fighter Aircraft the RAF bought an In house trainer for their fighter...... this consisted of a joystick mounted in a floor based column and a seperate controller, these were properly connected via a USB so they could use them on PC's, indeed that was their purpose. they were the genuine item. Even though the cables had been cut I bid up to quite a lot of money on them then bottled out..... re making these cables is not a problem for me, heck I can drop it off to mates in the industry and get it rewired in a couple of days for free, or simply do it myself........... still kicking myself over not completing on it http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

Taylortony
08-30-2007, 03:06 PM
Still thinking out of the box there are companies in the UK and indeed worldwide that sell parts and the original items............ this is where cost as said goes through the roof, BUT companies like this

http://www.mason-electric.com/products/control_grips/default.stm

actually have the drawings online for their sticks if you were to say erm machine one http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

Cough, cough like this

http://www.mason-electric.com/products/catalog/controlgrips.pdf

Taylortony
08-30-2007, 03:22 PM
Originally posted by Ploughman:
Wow, TT, what a neat idea. Of course those sticks have now become as ubiquitous as rocking horse poo. I notice you can buy a Sea Harrier on ebay for 28,000 (no joystick in her though, no surprise there).

err can wipe the floor with that price, can even source a complete scout helicopter for under 10 grand or a Jag for under 8000 see
http://www.everettaero.com/

birdy1964
09-01-2007, 07:06 AM
Originally posted by rnzoli:
...here are my sins... http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/1-Overview-left.JPG (http://web.axelero.hu/rnzoli/JoyBuild/1-Overview-left.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/2-Overview-above.JPG (http://web.axelero.hu/rnzoli/JoyBuild/2-Overview-above.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/3-Throttle-and-trim.JPG (http://web.axelero.hu/rnzoli/JoyBuild/3-Throttle-and-trim.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/4-Pedal-front.JPG (http://web.axelero.hu/rnzoli/JoyBuild/4-Pedal-front.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/5-Pedal-side.JPG (http://web.axelero.hu/rnzoli/JoyBuild/5-Pedal-side.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/6-Stick-alone.JPG (http://web.axelero.hu/rnzoli/JoyBuild/6-Stick-alone.JPG)

http://web.axelero.hu/rnzoli/JoyBuild/Thumbs/7-Stick-closeup.JPG (http://web.axelero.hu/rnzoli/JoyBuild/7-Stick-closeup.JPG)


After a long project lasting more than a year, I have finally built a very accurate HOTAS system with Force Feedback effect and pedals with toe-braking feature. I built all the mechanics around the electronics and the grips of 2 old Logitech FFB stick that were supposed to be thrown into the garbage long time ago. It is hard to describe how happy I am, good equipment is something that is essential for playing IL2, and at times, I wasn't sure I get this far with this project.

If someone is interested I can type up the long story behind it and/or the technical parameters, or post more pictures.
~S~ All, especialy rnzoli
This may sound abit stupid of me, but here goes anyway...
I was realy impressed with the creation by rnzoli and like the look of the joystick set up. My question is this, I have a logitech extreme 3d pro stick but would like to replicate a stick in this style. My stick has a gimble arrangement, as most or all do, and this gimble has to axis which have what I believe to be called pots? which are rotational devices which the stick reads. Are they capable of being replaced with another device that can simply be wired into their sockets on the base of the joystick which inturn can be used in the creation of a stick as is the one pictured in this post. This would then allow the pc to treat or recognise the base of the stick as it did before modification. Only a thought and I am probably making loads of you reading this with my lack of knowledge but I don't know if this is even possible.

rnzoli
09-01-2007, 05:28 PM
which have what I believe to be called pots? which are rotational devices which the stick reads.Are they capable of being replaced with another device that can simply be wired into their sockets on the base of the joystick which inturn can be used in the creation of a stick as is the one pictured in this post. This would then allow the pc to treat or recognise the base of the stick as it did before modification.
Yes, it IS possible http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif What you describe is how you can re-use the electronics of an existing stick and build your own hardware around it.

But you really need to be careful when you re-wire the cables. First of all, you need to replace the original pot with a device, which has roughly the same overall resistance, so you will not overload the stick elecronics inadvertenly. Secondly, the wiring should be of quality - boards can burn when short-circuited. USB stuff is fairly okay (overcurrent protection is there by default), but still, I managed to burn out one board with a stupid short circuit. This is not a problem when happens in factory (they just take a new board from the shelf), but quite painful when happens at home, far away from that factory http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

Taylortony
09-09-2007, 07:20 PM
Sorry for the delay I have made a thread showing its construction. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

TX-EcoDragon
09-10-2007, 12:47 AM
Very nice work!!!

That's something I've been wanting a long time. . .I miss the travel of an actual aircraft stick, particularly for aerobatic flying and combat stuff. Sure there is more body landguage needed, and many simmers would have to get used to using mroe than their wrists. . .I guess that's what I miss most when using my joysticks, well that and no detents in the action.

Here's a nice example of real stick travel:
http://www.kbvp.com/node/104

rnzoli
09-10-2007, 06:27 AM
@Taylortony, many thanks for the other thread, I am looking at it right now with a magnifying glass, LOL.

#EcoDragon, It really makes a nice difference to work from arm (lower arm), and not from wrist. I tried my hands on an X-52 recently in a shop, I just can't imagine myself going back to a "small" side-stick. http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif Also, having the stick in front of me allows me to swap hands and fiddle with my map and perhaps write up a few enemy sightings http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

Many thanks for the video clip, impressive flying there. I have one question though, about stick forces. Do you happen to know (from your flying experience), whether the elevator and aileron deflection requires roughly the same, or different forces normally?

I would guess that aileron movement requires noticeably smaller force on the stick (smaller area is deflected).

TX-EcoDragon
09-14-2007, 12:41 PM
I think I'm going to set out today and start building mine. I've got enough old sticks laying around, it's time to do it.


Do you happen to know (from your flying experience), whether the elevator and aileron deflection requires roughly the same, or different forces normally?

I would guess that aileron movement requires noticeably smaller force on the stick (smaller area is deflected).

There is no correct answer to that question other than to say that it is design dependant. There are methods to reduce or eliminate stick forces (in normal forward flight) in cases where that is the goal, but in most aircraft that are designed for more straight and level flight stick forces are higher, and may be far from harmonious. In those cases aileron is often lighter than pitch, at least around center, but with further stick travel aileron forces ramp up as well. Recognize that CG influences pitch but not roll heaviness, so this will influence harmony as well within a particular design.

In the high end advanced/unlimited class aerobatic aircraft, control harmony is critical, and generally stick forces are pretty much equal in any direction when in normal flight. Surfaces are hinged often at around 30% of the chord (so that the leading edge of the surface assists in moving the entire surface), spades (shovel like aerodynamic counterbalances) do a similar thing, etc. Getting it all perfect is rare though. . .it's an art really, but those designs that get it right really are wonderful to fly. In aircraft like the Extra 300, stick forces are so light in conventional flight that the stick almost seems to move away from your hand as you move it, they are precise enough that your pulse seems to lead to a response in the aircraft.

Many older designs might require a stronger push forward for a given G loading than the same G loading upright, but aircraft like the Pitts S-2C (photoship in my sig pic), Extra200/300, Edge etc have what are called "squared sticks". . .equal pressure in all four directions, and symmetrical responses upright or inverted-pushing or pulling. The previous models of the Pitts where very nice in this regard, but did require a harder push forward for a given G than when pulling, and this makes things like rolling circles etc. even more challenging than they usually are. The Pitts S-2C did away with these issues, increased roll rate, and got rid of the spades with an interesting aileron design that includes a hingeline that is moved back from teh leading edge of the aileron. If you look at the elevator and rudder these surfaces have aerodynamic counterbalances on them to lighten loads back at the tail.

All this changes when the aerodynamics change due to aircraft attitude. As an example, even the Extra with its practically non-existent stick forces becomes heavy in pitch when pushing the stick forward to accelerate an upright flat spin (or pulling in an inverted spin etc), the rudder tends to deflect with the spin direction, so the left and right pedal feel different, and inspin vs outspin aileron is a little different too. The ultimate divergence is during a tailslide when you are actually flying backwards. . .then you are working rather hard to keep the stick from slamming against the stops as you reverse all the control inputs for a given response.

249th_Maico
09-22-2007, 12:40 PM
SiNNNNERRRR!!! http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif Really nice stuff. The comunity is real proud of your efforts.

rnzoli
09-23-2007, 01:59 AM
EcoDragon, many thanks for all these details. It seems that force feedback technology has to come a long way to simulate in-flight control forces on the same detail level as flight modelling or visual rendering. The example about the controls trying to slam to the sides during the moments of "backwards" flight is a very interesting point and a very good example why current FFB implementation appear to be very far from the fully realistic feelings. FFB gives good addition sensory input in many cases, but not in every case yet.


As this thread is found time and again, sometimes resurrected, here is a bit of update what happened since I posted first.

The angle detection sensors were converted to infrared optical, to avoid signal instability due to wear and dust contamination. I found that the old resistance strips had to be replaced every 4-5 months - albeit cheap, still a hassle.

The black things are housing the angle detectors, and protects from interfering light sources.

http://web.t-online.hu/rnzoli/JoyBuild/optical-sensors-with-housing.JPG


Picture taken during the assembly shows the internals of the angle detector (already posted in the main HW modding thread):

http://web.t-online.hu/rnzoli/optical_sensor_1.JPG


I also re-designed and reconstructed the centering augmentation springs. The original FFB motors aren't strong enough to re-center the long stick from fully deflected status, therefore a centering mechanism helps this at larger deflection. Unfortunately the single spring system was easy to manufacture, but yet again, lateral forces on the spring cause material fatigue and spring failure every 3-4 months. The new contruction uses 2 separate springs and additional ball bearings to minimize lateral tention on the springs. The end result is quite nice: very minimal centering force in the middle area, this is where FFB shines, and a smoothly building up centering force near the edges of the movement.
What I learnt from this exercise is that
"If it rolls, it's smooth movement. If it slides, it will stick and move inaccurately." Except when you have access to military-grade materials, of course http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

http://web.t-online.hu/rnzoli/JoyBuild/dual-centering-augmentation.JPG


As usual with these things, once you complete something, immadiately the next ideas come to mind. I will plan to do 2 more things

Mechanically, it would be nice to use a lower seat and lower the stick as well, under the level of the desk. That would result in a more realistic cockpit-type seating arrangement (and a more comforable one, too http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif)

Electronically, I already have one of LBodnar's 12-bit ultra precise A/D converters, and I would try to incorporate it into the design in such way, that stick input would be provided by that 12-bit precise A/D converted, and relagate the 8-bit Logitech electronics to do the FFB part of the work only.

Stay tuned... LOL.