PDA

View Full Version : New stick, with a car's uni joint and hall effect sensors



julian265
07-15-2008, 10:04 PM
Here it is:

http://www.jpfiles.com/hardware/all.jpg

It's based around a universal joint, part of a prop shaft from a car, $2 hall effect sensors (Allegro A1321EUA-T bought online from Farnell), and some 8mm cubic neodymium magnets for less than $1 off ebay. (And of course Leo Bodnar's BU0836 USB interface, which is mounted in the grey box)

The automotive uni joint is suited to use in a stick - it's tough, is a pre-made gimbal, uses roller bearings with unnoticeable slack, has a square limit of movement, has a mounting flange, and pole to put a grip on, and it was reasonably easy to attach the sensors. I don't think this will ever wear out!!

I measured the output from this hall effect sensor/magnet relationship, as fitted to this stick, and as you can see it's damn close to linear:

http://www.jpfiles.com/hardware/linearity.png


Details regarding the use of the universal joint, and the hall effect sensor arrangement which achieved the above output, are written up in this PDF:

http://www.jpfiles.com/hardware/uni_stick.pdf

But these pics summarise the essence of the angular sensing method. (hall effect sensor shown in blue, magnet is stationary for this axis, with one pole on the upper right face, and the other on the lower left)

Pitched forward:
http://www.jpfiles.com/hardware/pitch_forward.jpg

Pitched backward:
http://www.jpfiles.com/hardware/pitch_back.jpg

And the rudder pedals:
http://www.jpfiles.com/hardware/pedals2.jpg

And the rudder pedal's sensor:
http://www.jpfiles.com/hardware/pedal_sensor.jpg

This is same arrangement as used for the stick. The pedal bar pivots on the large bolt and bearing, and the magnet is glued to it's top. The sensor is mounted on the pedal bar, and rotates around the magnet's vertical axis. One pole of the magnet is on it's front right face, the other is on it's rear left face. So in this view, the sensor is parallel to the magnetic field lines flowing around the magnet, and reads centered.

This sensor/magnet arrangement is not new, it is an adaption of one I saw in a google e-book snippet, which used two magnets, with the sensor positioned between them. It turns out that there's no need for the second magnet, you can just use the field flowing around one of them, and mount it and the sensor on the axis of rotation.

Edit - changed file hosting

WTE_Galway
07-15-2008, 10:11 PM
good solid engineering ... well done

makes one wonder whether you could build a practical stick using strain gauges

ffb
07-15-2008, 11:46 PM
looks ideal ..... so when are you going to market them? http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

na85
07-16-2008, 12:27 AM
Awesome!

JG52Uther
07-16-2008, 02:01 AM
Originally posted by na85:
Awesome!
+1

WTE_Ibis
07-16-2008, 02:55 AM
I'll take two please.


.

rnzoli
07-16-2008, 03:01 AM
you can just use the field flowing around one of them, and mount it and the sensor on the axis of rotation.
This is a very important discovery, I think a lot of DIY people will appreciate it! http://forums.ubi.com/images/smilies/11.gif

squareusr
07-16-2008, 12:44 PM
"car uni.joint used for bu836 flight stick" - i knew this would eventually happen one day. Thanks for posting, great motivation for my own projects!

K_Freddie
07-16-2008, 01:17 PM
Very nice.. An added thought - you might want to add some toe caps and serrated teeth on the rudder pedals - you need these for those 'exciting' moments

.. as well as sensors on each pedal so they can be used for differential braking.

http://forums.ubi.com/images/smilies/25.gif

julian265
07-16-2008, 05:05 PM
Thanks for the complements!

Galway, I guess the simplest method of using them would be one strain gauge on the front, and one on the side of the stick pole. Doing that wouldn't give you any movement, so you'd then need a sprung base (a engine's valve spring might be ideal). It sounds like a good idea, could be very compact, and has no movement around the sensors!

Rnzoli, I have seen a few pics which might have used this method, but they lacked an explanation of it, and the orientation of the magnet, so I didn't understand it! Either way, it's been about a year since I started reading about hall effect sensing, and it's taken this long for me to to find a method I was truly happy with, albeit with two magnets!

Freddie, I've had no problems slipping off the pedals so far, my wheeled chair is held near the stick post, and my legs place all their weight on them. I guess some side and top 'fences' wouldn't go astray though, especially to aid in settling my feet on them.

I was thinking about brakes when I made the pedals, but IL2's lack of use of differential analog braking, and the desire to get flying again put me off including them! (It was hard enough doing the painting, and delaying flight for the sake of aesthetics!!) I'll build them in one day though.

rnzoli
07-17-2008, 06:37 AM
Either way, it's been about a year since I started reading about hall effect sensing, and it's taken this long for me to to find a method I was truly happy with, albeit with two magnets!
I was personally afraid of using magnets and their invisible fields, the temperature drift that I read about, or the power consumption of compensation circuits. But looking up the sensor you used, I can also say "wow".... http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif


Each device has a BiCMOS monolithic circuit which integrates a Hall element, improved temperature-compensating circuitry to reduce the intrinsic sensitivity drift of the Hall element,
a small-signal high-gain amplifier, and a rail-to-rail lowimpedance
output stage.
http://www.allegromicro.com/en/Products/Part_Numbers/1321/1321.pdf

Icc Max = 8 mA - BU 836 can drive a few of them without problems (it can drive up to 100mA, I think).


Great work, I am inclined to try this sensor on my own hardware.

However, I am interested about one important aspect. How much it is resistant against noise?

For example, if you put all your controls into neutral position, and start Leo's application DIView.exe, can you see any noise (small value changes) coming up in the digital signal?

I have the 12-bit version of Leo's controller, and with my infrared sensors it picks up a small noise (only 1 bit plus and minus) in this all-netural-and-untouched configuration. So I am wondering, whether the active output from the HE sensors would be more stable, and less vulnerable to interference from other electric signals. http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

RxMan
07-17-2008, 07:13 AM
Are you using any centering forces ?

Worf101
07-17-2008, 07:38 AM
Dagnabit, no pics anymnore. Can't see what you're talking about and the links dead. Sigh...

Da Worfster

julian265
07-17-2008, 07:54 AM
RxMan, no, the stick holds whatever position I leave it in, which I prefer.

Rnzoli, dependent on where each axis is left, (pots and hall effect sensors) I can get one bit of oscillation too, dependent on the position of each axis. ie, at one position they're still, at another they oscillate by one bit. IMO this is due to the signal being between the ADC's 'bands', and will occur with any ADC, at differing frequencies depending on sample rate and damping. The fact that it occurs with pots as well seems to me to confirm that.

With 4096 steps, or even 1024, I wouldn't be particularly worried about it, as one bit represents a tiny change in input!

I liked the average quiescent output voltage vs temperature graph of this sensor... I think it was originally one of your comments that triggered me to look at it!

julian265
07-17-2008, 07:56 AM
Originally posted by Worf101:
Dagnabit, no pics anymnore. Can't see what you're talking about and the links dead. Sigh...

Da Worfster

Hmmm, I still see the pics here. But I'm using my ISP's free webspace, and I've noticed that it isn't too reliable. Please try again another time!

JG52Uther
07-17-2008, 09:16 AM
I still see the pics.

FoolTrottel
07-17-2008, 12:09 PM
Great suff this, great discovery...
(I'm in the process of ordering these components... to 'upgrade' my homebuilt pedals and controlbox)

I'm just wondering... would it also work when one glues the magnet to the moving axis, and keep the sensor still? http://forums.ubi.com/images/smilies/blink.gif

Another question: Could one use one magnet, and two rotating sensors, one each side of the magnet?

julian265
07-17-2008, 05:19 PM
Moving magnet, still sensor works fine! As long as both the magnet and sensor are mounted on the axis of rotation, it doesn't matter which is where.

Take a close look at this:
http://www.members.optusnet.com.au/julian265/il2/unistick/both_sensors.jpg

The left hand pair monitors roll, and for this axis, the magnet rotates with the stick.

For pitch, it's the sensor that rotates with the stick. (though both the magnet and sensor move with the roll axis, they're mounted on it!) My old logitech extreme 3d pro had the same arrangement.

julian265
07-17-2008, 05:25 PM
Oh, and I'm puzzled as to why you'd want two sensors per axis!

TX-EcoDragon
07-17-2008, 05:27 PM
Incorporate some sort of centering sytem that is adjustable/FFB based, smooth (no transition flats) and you'll have a top notch, nice realistic stick!!

For that matter, do that, and I'll take one. . .or maybe two!

M_Gunz
07-17-2008, 07:20 PM
Originally posted by julian265:
Oh, and I'm puzzled as to why you'd want two sensors per axis!

There's two axes as a gimbal in that picture. FFB on that would be wicked.

WTE_Galway
07-17-2008, 08:00 PM
Originally posted by M_Gunz:

FFB on that would be wicked.

Hydraulic rams anyone http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif Let's simulate the REAL force needed to pull back on the stick in a high speed dive with the controls locking up.

na85
07-17-2008, 09:52 PM
Originally posted by FoolTrottel:
Great suff this, great discovery...
(I'm in the process of ordering these components... to 'upgrade' my homebuilt pedals and controlbox)

I'm just wondering... would it also work when one glues the magnet to the moving axis, and keep the sensor still? http://forums.ubi.com/images/smilies/blink.gif

Another question: Could one use one magnet, and two rotating sensors, one each side of the magnet?

You mean, one magnet, with one sensor doing Y axis, and the other sensor at right angles to the first doing the X axis?

Possibly yes, but it would be very complicated.

julian265
07-17-2008, 10:05 PM
Ah, ok, just got it.

That's a creative idea - and I can see it working if you mount the magnet at the intersection of the X and Y axes. On my stick, that'd be right inside the uni joint's spider, though it's work for with a custom gimbal.

One step further is this (http://www.melexis.com/News/MLX90333_Triaxis™_3D-Joystick_Position_Sensor_Recipient_of_Electronic_P roducts'_2007_Product_of_the_Year_Award_573.aspx)

One sensor (well, three in one), and one magnet.

FoolTrottel
07-17-2008, 10:57 PM
with one sensor doing Y axis, and the other sensor at right angles to the first doing the X axis?
Well, not specifically X and Y...

Just thinking about sliders.

Like for Power and Pitch, next to eachother, in alignment, but separate from eachoter... except for sharing one magnet.

But, thanks for the answer. http://forums.ubi.com/images/smilies/25.gif

I will go and try... once I've got all parts together that is... http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

Got yet another question.
My current controlbox is based on normal pots, using the electronics from older joysticks.
Some pots are only using 2 wires, the middle one, and one outer one. Would those Hall sensors need all three wires connected? (If so, I think I may be needing a Bodnarboard...)

julian265
07-17-2008, 11:39 PM
Yeah, the sensors need +5V, 0V and the other pin is the output.

I'm not sure about the modifications needed for the two wire pot circuits, however I would definitely NOT hook a hall sensor's output up to them without modification, even if you gave the sensor +5 and 0V externally.

WTE_Galway
07-18-2008, 12:13 AM
Originally posted by julian265:
Yeah, the sensors need +5V, 0V and the other pin is the output.

I'm not sure about the modifications needed for the two wire pot circuits, however I would definitely NOT hook a hall sensor's output up to them without modification, even if you gave the sensor +5 and 0V externally.

With traditional gameport analogue joysticks varying the pot usually controlled the frequency of a timer somewhere on the soundcard or mb. In fact on very old PC's quite literally a pair of good old 555 timers with the external pots in the joysticks being slotted into the timing circuit. One on each axis.

rnzoli
07-18-2008, 02:01 AM
Originally posted by FoolTrottel:
(I'm in the process of ordering these components... http://forums.ubi.com/images/smilies/34.gif He he! So I am not the only one roaming the Farell site, huh? http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

Julian, one question about the magnet. From a local distributor, I can see 2 types of 8mm x 8mm x 8 mm cubes, one with N35 and the other with N48 "grade" markes. Do you know which one you use?

Another question: from where did you get the grip/handle on your stick?

julian265
07-18-2008, 02:47 AM
You probably know this, but to confirm, the N grades refer to the magnet's strength, or density of it's magnetic field. (but make sure that they're both neodymium or rare earth)

Firstly, both grades will work (I used N45). With everything else constant, the stronger magnet will increase sensitivity to angle. In other words, for a given amount of joystick deflection, the stronger magnet will cause a larger change in voltage from the sensor.

Since you can also adjust sensitivity by moving the sensor toward or away from the magnet, I'd buy the N48 kind. You can decrease sensitivity by any amount you want by positioning, but if your sensor won't saturate at your joystick's angle extents, when using the N35 magnet and a close sensor, you can't increase it.

julian265
07-18-2008, 02:55 AM
Oh yeah, and the grip. A squad member (453_Dutneall) works with aircraft, and was replacing cracked grips. I was lucky enough to get one of the un-airworthy ones! (it has an ineffectual crack in the handle)

It has a two stage trigger, 4 way hat, two thumb buttons and pinky switch... and is damn comfortable! (when you get used to the tough buttons!)

Attaching it to the driveshaft required a bit of thought, but ended up being easy and solid. I cut a section of shaft into three (lengthways), flattened them slightly, then used a large hose clamp from a car's intake to clamp them onto the shaft and stick base.

rnzoli
07-18-2008, 05:13 AM
Hmm, so you have a real figher stick http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif
Man, I am green with envy.
Who is this Dutneal guy, and how can I befriend him? http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif

BTW, I am trying to change my order to N48 magnet (yes, it's NdFeB) as you suggested. I was afraid of having too little stick travel before saturation, because I need roughly +/- 30 degrees with my hardware, a little bit more than you have. But you are right( and I am a little dumb http://forums.ubi.com/images/smilies/53.gif), increasing the distance between the sensor and the magnet should take care of that. I will probably try to put together a mechanical arrangement, where I can asjust the distance/angle with positioning screws, because I will need very steady holding due to FFB shaking, rattling etc.

However, another question: you write about the dangers of shorting the +5V to GND, and sending the compy to eternal heavents. Is that really so dangerous? I was always under the impression, that USB interfaces are fool-proof, their power supply is monitored and cut-off automatically when shorted or exceeding the current limit (500 mA? me thinks). Did you actually blew a computer that way (sorry for scratching open wounds in that case http://forums.ubi.com/images/smilies/halo.gif), or it was just a "be on the safe side" type of warning?

julian265
07-18-2008, 07:46 AM
Yeah, I was lucky to get this grip!

Those magnets should be cheap enough to buy both types, and pick the one that works best for you.

Yeah, it was that type of warning! Although despite any over-current protection, I still wouldn't be willing to short my USB!

Tully__
07-18-2008, 08:40 AM
Originally posted by julian265:
....but IL2's lack of use of differential analog braking....
IL2 does have differential analog braking, but it is not implemented with the modern pedal system.

British and Soviet aircraft of the WW2 era had a single brake actuator (generally using compressed air) and a valving system attached to the rudder axis to control how much pressure went to each side of the braking system.

The game works the same way, you can even see the hand operated brake lever moving on the joystick in some of the aircraft in the game that were equipped with this system. I've had mine set up this way on an axis and it does work as advertised.

Sokol__1
07-18-2008, 10:25 AM
Julian265,

Nice work.
Thanks for sharing info about use of HALL. I have two Suncon F-15 waiting for "mods". http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

I found a similar use of u-join here:

http://a77.org/sap/joy/gallery_joy.html
http://translate.google.com/translate?u=http%3A%2F%2Fa7...&ie=UTF8&sl=ru&tl=en (http://translate.google.com/translate?u=http%3A%2F%2Fa77.org%2Fsap%2Fjoy%2Fgal lery_joy.html&hl=pt-BR&ie=UTF8&sl=ru&tl=en)

Sokol1

julian265
07-18-2008, 06:36 PM
Thanks Tully, I had no clue... I'll change the pdf.

Ah, Sokol, you found another nutter!

rnzoli
07-23-2008, 03:37 AM
OK, the magnets are here (N48), but the sensor is on "back order" status at Farrel since last Thursday... no estimated delivery whatsoever http://forums.ubi.com/images/smilies/blink.gif

Does it really take that long to start shipping? Or did someone (FoolTrottel maybe!)order a 2000 pcs. batch? http://forums.ubi.com/groupee_common/emoticons/icon_mad.gif

FoolTrottel
07-23-2008, 05:19 AM
but the sensor is on "back order" status at Farrel since last Thursday
Hm... I recognize this from somewhere... could it be they've had some increase in sales in these things recently? http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif

rnzoli
07-23-2008, 07:04 AM
http://web.t-online.hu/rnzoli/ohnoes.gif
julian265 bought ALL the stock for experimenting, and now there is nothing left for us! The sky has fallen, the sky has fallen!

julian265
07-23-2008, 07:16 AM
MUHAHAHAHAAAAAA

All the world's 1321 are belong to me!!!

You will all be forced to pay prices set by my cartel!!

I hope you get them soon... You did order from the website of your own country, right? Mine arrived at my doorstep in about one week, including postage from the only stock available, which was in the UK.

na85
07-23-2008, 08:49 AM
Originally posted by FoolTrottel:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">with one sensor doing Y axis, and the other sensor at right angles to the first doing the X axis?
Well, not specifically X and Y...

Just thinking about sliders.

Like for Power and Pitch, next to eachother, in alignment, but separate from eachoter... except for sharing one magnet.

But, thanks for the answer. http://forums.ubi.com/images/smilies/25.gif

I will go and try... once I've got all parts together that is... http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

Got yet another question.
My current controlbox is based on normal pots, using the electronics from older joysticks.
Some pots are only using 2 wires, the middle one, and one outer one. Would those Hall sensors need all three wires connected? (If so, I think I may be needing a Bodnarboard...) </div></BLOCKQUOTE>

The sliders bit would be significantly easier than X and Y.

As for your control box question, I can't say for sure.

Pots and hall effect sensors are drastically different devices, though. The pot's 3 wires are so that you can use the full resistance if required. You never need to use more than 2 wires.

For hall sensors, I think (correct me if I'm wrong) that they require their own 5V power source, because there is a small integrated circuit in there.

na85
07-23-2008, 08:50 AM
Originally posted by rnzoli:
OK, the magnets are here (N48), but the sensor is on "back order" status at Farrel since last Thursday... no estimated delivery whatsoever http://forums.ubi.com/images/smilies/blink.gif

Does it really take that long to start shipping? Or did someone (FoolTrottel maybe!)order a 2000 pcs. batch? http://forums.ubi.com/groupee_common/emoticons/icon_mad.gif

Can you provide a link to the magnets you bought?

rnzoli
07-23-2008, 08:52 AM
Originally posted by julian265:
I hope you get them soon... You did order from the website of your own country, right? Mine arrived at my doorstep in about one week, including postage from the only stock available, which was in the UK.
well, the funny thing was, when I was preparing the order, the item was marked as "awaiting shipment" and still is, for new orders

so either I don't grasp the process at Farnell, or their online data is rather misleading

I wouldn't make a big fuss about this, but upon the "awaiting shipment" indication, I gave my workplace as delivery address, however, I am on vacation from next week and the Farnell site doesn't let me change the delivery address anymore for the placed order http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

So the sensors will be sitting at my workdesk alone for 3 weeks, dark, while I am going to be sitting at home, unable to experiment them when I would have the time - what a tragic combination http://forums.ubi.com/images/smilies/bigtears.gif

Maybe I will sneak in to my workplace and pick them up after they arrive http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif

rnzoli
07-23-2008, 08:57 AM
Originally posted by na85:
Can you provide a link to the magnets you bought? I was pleasantly surprised to find a manufacturer/distributor in my own country, I bought from them: http://www.euromagnet.hu/ probably not an option for you, unless you live here...

rnzoli
07-23-2008, 09:17 AM
Originally posted by na85:
The pot's 3 wires are so that you can use the full resistance if required. You never need to use more than 2 wires.
I think there are several cases when all 3 wires are used, e.g., when you use the pot as a voltage divider. Then you put one end on GND, the other on +5V, and the sweeper in the middle gives you a voltage in between, depending on the mechanical angle (or distance in case of sliders)


For hall sensors, I think (correct me if I'm wrong) that they require their own 5V power source, because there is a small integrated circuit in there. I think it is more correct to put it this way: Hall effect sensors will require considerably higher current/power from the +5V lines than ordinary pots. A joystcik pot can be in the range of 10-100kOhm. The Hall sensor draining 8mA will look less than 1 kOhm. You don't need separate power supply for 8mA, it's practically nothing, can be driven from the USB as well, however, it may be considerably higher than what old joystick circuits are designed for. I suspect that my old Logitech circuitry had 1-2 kOhm resistors before and after the +5V/GND, because I could short them without punishment http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif and it still worked http://forums.ubi.com/images/smilies/typing.gif However, putting a HE sensor would not, the joystick circuit would not provide enough power to the sensor. This is not a problem howwever with the BU836, according to its specs, it provides up to total 100mA directly from the 3 pins per sensor (no separate wires, no separate power supply needed).

FoolTrottel
07-23-2008, 09:36 AM
This is not a problem howwever with the BU836, according to its specs, it provides up to total 100mA directly from the 3 pins per sensor (no separate wires, no separate power supply needed).
Yeah, I think I'll be needing one of those...
Took apart a Trust stick... measured some voltages... only found values of 3V... oops...

FoolTrottel
07-23-2008, 09:42 AM
Originally posted by na85:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by rnzoli:
OK, the magnets are here (N48), but the sensor is on "back order" status at Farrel since last Thursday... no estimated delivery whatsoever http://forums.ubi.com/images/smilies/blink.gif

Does it really take that long to start shipping? Or did someone (FoolTrottel maybe!)order a 2000 pcs. batch? http://forums.ubi.com/groupee_common/emoticons/icon_mad.gif

Can you provide a link to the magnets you bought? </div></BLOCKQUOTE>

This is where I got 'm: SuperMagnete.de (http://www.supermagnete.de/eng/index.php?switch_lang=1) (Well, still awaiting delivery, ordered them last weekend, goods should be here tomorrow)

@rnzoli, as for the Hal sensors, I only ordered 25 of 'm http://forums.ubi.com/images/smilies/16x16_smiley-happy.gif

rnzoli
07-24-2008, 02:10 AM
Uhh, I think I found what the problem is with the delivery of these sensors:

From the manufacturer's web site:

Due to <span class="ev_code_YELLOW">warehouse relocation</span>, samples will be unavailable until August 1, 2008. Contact your sales rep or purchase from distributor.
http://www.allegromicro.com/en/Products/Part_Numbers/1321/

Fehler
07-24-2008, 02:21 AM
I have been following this thread and the cockpit building one for some time.

I have to say that Julian's application of a U-joint as a gimble is sheer genious! I work on cars in my spare time as a hobby, and I am quite disappointed in myself that I didnt think of this! http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

I had an old drive shaft laying around that I will be using for this project (As soon as I decide which hot Texas day I want to go do some sandblasting to remove the rust! http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif )

I received my hall sensors today (I ordered them from digikey.com.) Oh - I was surprised to see how small they are. I was expecting something a little bigger for some reason.

I have a couple of old hard drives that I will probably use for magnets. The fields on a hard drive magnet are pretty strong, so I will see if they work before I go order ones elsewhere. I can cut them to the desidered size with a band saw. Just got to be careful not to breath in the magnetic dust or I might stick to the refrigerator when I walk by it... Haha!

I have a BU086 I purchased a few years ago but didnt do anything with because I took some time off from flying (Job related).

So, I am going to start my project this weekend. I have access to plasma cutters, welders, and a bunch of stuff for fabrication, but I am at a loss for finding a suitable grip.

I found a nice piece at Tarmac Aces, but my wife would probably kill me if I spent $500 US on a flight stick handle!! LOL! But wow, it's a nice piece! (Wonder if she will notice $500 missing from the savings account or the kid's college fund... Hmmm.... LOL)

I will post pics as soon as I can.

Thanks Julian! This is going to be a lot of fun! http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

bolox00
07-24-2008, 02:47 AM
grips?
http://www.guardian-electric.com/html/grips.htm

not very WW11 but might be of use

Esel1964
07-24-2008, 04:56 AM
julian265-Looks like a bulletproof build,except the wiring.
With that setup like that,you could let a silverback gorilla take the controls-and get p***ed off in the process,if you reinforced the wiring a bit.

I may be talking overkill,but if you used heat shrink-wrap ,then plastic,spiral wrap on the wire bundles,and large heat-shrink wrap to cover the 'circuit boards' holding the sensors(or a small plastic box) to prevent shorting,in case something metal caused a short.

The mechanical construction looks bulletproof,so tough looking it makes the wiring and sensors look vulnerable.I know it's 2 driveshaft yokes and a U-joint,so it has to be tough;if you reinforced the wiring it'd be tough all the way 'round.I'm not criticizing your work in the least,it's darned good looking work,the wiring just looks vulnerable to accidents.

Anyway,I'm through 'suggesting',beauty of a job M8! Congrats on a good job. http://forums.ubi.com/images/smilies/25.gif

julian265
07-24-2008, 07:28 AM
Fehler, I can't claim to be the first to use a U joint for this, but I can happily say that I discovered it for myself! (after almost barking up a more difficult tree with a ball-jointed gear shifter)

Looks like you've got almost all the bits - though you might be in for some fiddling with those HD magnets. What made the angle sensing method in the PDF so forgiving, was the nice area of nearly parallel field lines along the faces of the magnets. I don't know the pole positioning of HD magnets off the top of my head, but if their poles are at either end you'll be fine.

Bolox, I found my grip on the site you linked!
http://www.guardian-electric.com/html/b_8_control_grip.htm

Dunno what they charge for them... I was lucky to be given a used one!

Esel, I certainly agree about the wiring and sensors. Hopefully I'll be able to get the sensors encased in resin or similar... I just wasn't able to come up with anything better whilst I knew that my new stick was so close to done!

I've been considering moving the interface box off the main post, to prevent all the wires from passing under one leg. Something has to be done to call it "finished", that's for sure!

Thanks you all for your interest and compliments. It's great to see the write-up being used!

na85
07-24-2008, 07:42 AM
Originally posted by rnzoli:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by na85:
The pot's 3 wires are so that you can use the full resistance if required. You never need to use more than 2 wires.
I think there are several cases when all 3 wires are used, e.g., when you use the pot as a voltage divider. Then you put one end on GND, the other on +5V, and the sweeper in the middle gives you a voltage in between, depending on the mechanical angle (or distance in case of sliders) </div></BLOCKQUOTE>

You're right. I should have said "for this application you never need all 3 pins"

Sokol__1
07-24-2008, 08:06 AM
From Frugalsworld, a examplo of how put spring in U-Jont:

http://users.accesscomm.ca/9thumbs/u-joint.jpg
http://users.accesscomm.ca/9thumbs/u-joint+electronics.jpg
http://users.accesscomm.ca/9thumbs/floor-mount3.jpg

Sokol1

JG52Uther
07-24-2008, 09:08 AM
Fabulous stuff,and its giving me itchy fingers!
One question:
Can you connect this stuff through USB or is it gameport?

Urufu_Shinjiro
07-24-2008, 11:58 AM
Originally posted by Sokol__1:
From Frugalsworld, a examplo of how put spring in U-Jont:

http://users.accesscomm.ca/9thumbs/u-joint.jpg
http://users.accesscomm.ca/9thumbs/u-joint+electronics.jpg
http://users.accesscomm.ca/9thumbs/floor-mount3.jpg

Sokol1

That's awesome! I'm no good with my hands, I wish I could get someone to build on of these for me, lol.

Sokol__1
07-24-2008, 03:22 PM
Two more U-Join with springs, from AviaForum:
http://img296.imageshack.us/img296/3972/ujoinspringlz6.jpg

Strange concept:
http://img517.imageshack.us/img517/3319/ujoinspring2bg7.jpg


Sokol1

M_Gunz
07-24-2008, 03:58 PM
I've been doing some experimenting and parts gathering. I have IR LEDS and IR sensors that run
under $0.25 ea hooked into a basic circuit. The LED has only 4Ma current at 5V, still the
sensor picks it up just fine. 4Ma is far less than normal operating current for this LED.

I had tried reflecting the IR off paper with gradient white to black printed on it but there
is no difference to 900nm IR. I went a long time looking for inkjet printable clear sheets
in less than a $40 pack when finally I printed gradients on white and for $0.50 had Office
Depot laser print those on clear. And.. it works, the light going through is cut.

The sensor is a basic photo-resistor, has two leads only and the less IR the more resistance.

Final steps.
The gradients are printed in strips to go around the edge of a plastic lid for pedals but if
I make sliders then the strip is kept straight. I have short strips to make small dials work.

It's been a while. I wanted something very fault-tolerant that can take looseness or less
than machine tolerance or even a bit of wear and still work. The gradient strips are 12.5mm
wide plus some edge to glue and lines for keeping somewhat straight. Little bit off will not
matter as long as the part does not jitter. At this point I feel I can even build sliders
box of cardboard and still make it work. :-P

However I thought I would share how it is possible to go low cost and still work. The LED
and sensor are very cheap, it still needs (I have now) a trim pot for adjust, wire and a
15 pin connector for the gameport and there is the electronics for position-sensing, total.

It is even possible to get fancy with the gradient to create custom profiles this way.
And yeah, I have been a fan of Hall sensors in other applications.

julian265
07-24-2008, 05:40 PM
JG52Uther - the BU0836 is a USB interface, giving you 8 analog axes, and 36 buttons. It's also quite simple to deal with.

Fehler
07-24-2008, 11:44 PM
Well, I found a little time today to solder up a sensor and test out these hard drive magnets I have.

Since my sensors didn't come with a datasheet, I assumed that the wiring configuration was +5v/Sensor/Grnd (Like the average Pot) And likewise, I assumed +/- didnt matter (Just like the average pot) I was wrong.

As soon as I plugged in the BU086, the hall sensor got so hot it burned my finger! My immediate thoughts were... SH!T

I got on the internet and located a datasheet. The sensor was supposed to be wired +5/Grnd/Sensor. Note to self: Remeber to follow Ben Franklins words about an ounce of prevention beating a pound of cure!

I fixed the configuration of my pins and was impressed when the bugger still worked! :P Tough little bastages I guess...

The sensor was able to read the magnetic fields of the hard drive magnets, so I played around with moving the sensor closer and farther away from the surface of the magnet, and it does indeed change the sensitivity as Julian described in his .pdf.

Finding the centerline of the field is easy with a simple voltage meter, and on the controller interface in Windows I noticed how precise the movement was with the 1024 resolution of the BU086! Much more precise than my old buggered Saitek X45 to say the least!

Hope I find some time to start cleanup on the driveshaft this weekend!

julian265
07-25-2008, 12:17 AM
As Leo says, those chips are tough. I smoked one of my sensors too, by accidentally shorting the output pin to ground, with my multimeter's probe, and the 0836 was still fine to my amazement!

Oh, and Leo offers 12 bit, or 4096 steps!!

Fehler
07-30-2008, 09:25 PM
Well, I found a little time this weekend (In between work and getting a horrible sinus infection) to do a little work on this project.

This first pic is of my U-joint all blasted up and some black primer sprayed on it. I tried the drill bit, then the grinding stone on the hardened cap of the U-Joint. I said P!$$ on it and blew a hole through it with a plasma cutter. Took 2 seconds... haha!

http://webpages.charter.net/cuda70/J1.jpg

Then I decided to use some two-part epoxy to hold everything in place on the perf board. I love epoxy. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

http://webpages.charter.net/cuda70/J2.jpg

Well, that's it for now. I plugged in the controller and moved the sensors to adjust the sensitivity to my liking. Maybe I will have more time to work on it again this weekend. I have to go pick up some plate and 1x1 box tubing to make the framing for my pit. I really can't build the stick handle until I know exactly where I want to mount the thing (Got to get the right throw or I will be banging my thighs all the time when I roll the plane)

I have to say, I bet the sugar-cube magnets work better than these Hard Drive magnets. The field is a little flakey on these hard drive magnets because of the shape of the magnet, and the cubes are a little more compact. I will probably be getting some and re-doing the dowel pins. Oh, I used epoxy to stick the magnets to the pins too... Did I mention that I love epoxy? Haha!

julian265
07-31-2008, 01:50 AM
Fantastic!

lol... plasma cutter... I didn't have ANY success with a drill bit either.

I'm looking forward to hearing your impressions when you get those magnets sorted.

Chivas
07-31-2008, 11:55 AM
Love your idea Julian. If I remember correctly yours doesn't have a centering force. I kind of like that idea where the controls stay where you leave it. How does it compare to flying a real aircraft? Is the only centering force in aircraft of the era the air resistance forcing the ailerons and elevators to a more neutral position?

FoolTrottel
07-31-2008, 01:11 PM
Working on setting up a set of sliders using them magnets and HAL sensors...
(Still awaiting delivery of the sensors though http://forums.ubi.com/images/smilies/halo.gif)
http://img.photobucket.com/albums/v668/fooltrottel/Slider001.jpg
Using a 50mm long aluminium U-profile, its width is 15mm. 5mm bolt/nuts, and a 7mm magnet.
(Dunno if the 5mm magnet will be sufficient... got some 10mm ones too)

The locknuts are tightened up to a point there's a nice resistance felt.
The plastic washers make it feel smooth.
http://img.photobucket.com/albums/v668/fooltrottel/Slider002.jpg
http://img.photobucket.com/albums/v668/fooltrottel/Slider003.jpg
(Magnet's not yet glued up, it's holding its own there...)

rnzoli
07-31-2008, 02:23 PM
I am also waiting for the delivery of the sensors, goddamit. If my vacation is over before they arrive, I have to wait until the Christmas holidays to start experimenting http://forums.ubi.com/groupee_common/emoticons/icon_mad.gif

My priority is to fix a small wheel on my home-made throttle, which I use for prop pitch. It uses a little pot, but started to spike badly, and the small dimensions of the magnet/sensor combination would be perfect for replacement.

After that, the trottle itself would follow suit...

FoolTrottel
07-31-2008, 03:03 PM
If my vacation is over before they arrive, I have to wait until the Christmas holidays to start experimenting
Darn, you got a busy life! http://forums.ubi.com/images/smilies/16x16_smiley-surprised.gif
(Where are you at? Want me to send you some sensors should they arrive at my doorstep earlier than yours?)

julian265
07-31-2008, 06:18 PM
Chivas, I can't comment on how real it feels, I haven't flown in a while and haven't flown anything with a stick (not column) anyway.

I'd be guessing that some centering forces would be most realistic, however for flight sims, I'm very happy with the frictional set up, and will leave it as is.

I've realised that the feature that makes the largest difference to me is the long throw of the stick, and due to this, the grip doesn't change angle too much. The weight of the stick makes a good difference too.

julian265
07-31-2008, 06:24 PM
Looking good Fooltrottel! I was pondering how to get some decent and adjustable friction for a throttle, and that seems to be it. No bearings either!

I hope you get your sensors soon guys... It's taking a disappointingly long time.

Fehler
08-01-2008, 02:33 AM
Interesting slider idea...

I had already thought of something similar using the design of the Saitek throttle in mind...

This is a very crude drawling of what I was thinking about making...

http://webpages.charter.net/cuda70/T1.jpg

Using a piece of cloth strapping as shown in the figure and screw and washer through it as a tensioner, I can adjust the tension on the pivot bar to my liking... Just an idea I have in my head. This is very similar to the method Saitek uses to control the tension on their throttles...

I really don't want to use bearings because of they could get a little pricey. Although, I could over-engineer the hell out of this with a nice budget! Hahah!

rnzoli
08-01-2008, 03:56 AM
Originally posted by FoolTrottel:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">If my vacation is over before they arrive, I have to wait until the Christmas holidays to start experimenting
Darn, you got a busy life! http://forums.ubi.com/images/smilies/16x16_smiley-surprised.gif
(Where are you at? Want me to send you some sensors should they arrive at my doorstep earlier than yours?) </div></BLOCKQUOTE>
Thanks a lot, I am afraid we are going to get ours roughly at the same time due to the warehouse move of the manufacturer. I was thinking of swapping some of my magnets with Fehler in exchange of a few spare sensors, if he got any http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif but since he seems to be in the US and I'm in continental Europe, I figure that I just have to be very patient. Maybe the situation won't be that bad at work, and I can snatch half-a-weekend once in a while even before Xmas.

Meanwhile, I keep myself entertained looking at the various ideas coming up from you guys http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

FoolTrottel
08-01-2008, 04:15 AM
and I'm in continental Europe
Looks like you are right, I'm probably suffering from that same Warehouse move too...

Oh well... http://forums.ubi.com/images/smilies/halo.gif

Chivas
08-06-2008, 04:27 PM
I've been looking around for a suitable car u-joint and just thought of using a socket universal joint. Maybe a heavy duty 3/4 or 1" universal socket. I'm also wondering if you could use a single 3" wide coil spring slid over top the the universal joint with enough room for the hall sensors. Just thinking out loud. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

FoolTrottel
08-06-2008, 06:01 PM
In the meantime I've developed the slider construction, and found out I do not need a U-shaped base, an L-shaped one does fine.

Construction is basically the same, if you look at my first pic, from left to right:
- Removed the locknut, the washer, and the aluminum vertical part.
- The next nut has been replaced by a 16mm long bolt, next the slider strip, and the nut (which is tightened to the bolt, thus fixing the strip to the bolt)
- Plastic washer, vertical part of aluminimum L-shaped strip, plastic washer.
- Locknut, tightened so tension/friction is fine.

Thus saving in materials, and (more important) space!
(Does this make sense? No pics yet, sorry)

Fehler, the locking nuts / plastic washer combination is not expensive at all, and rather nicely adjustable to get the tension one likes. Whether it'll do in the long run, I dunno... yet.

(Still waiting for the sensors to arrive...)

julian265
08-07-2008, 12:15 AM
Originally posted by Chivas:
I've been looking around for a suitable car u-joint and just thought of using a socket universal joint. Maybe a heavy duty 3/4 or 1" universal socket. I'm also wondering if you could use a single 3" wide coil spring slid over top the the universal joint with enough room for the hall sensors. Just thinking out loud. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

I just looked at my 1/2" drive U-joint. Mine has *a lot* of free play, so you'd need to buy a really good one with closer tolerances to make a good stick. The pivots also looked very difficult to attach magnets/sensors to.

If you want to avoid a full size joint, I suggest a steering U-joint.

The coil spring sounds good and simple... If you can get one wide enough!

WTE_Galway
08-07-2008, 12:30 AM
Originally posted by Chivas:
Love your idea Julian. If I remember correctly yours doesn't have a centering force. I kind of like that idea where the controls stay where you leave it. How does it compare to flying a real aircraft? Is the only centering force in aircraft of the era the air resistance forcing the ailerons and elevators to a more neutral position?

The most significant difference that I have noticed between real aircraft and any game level simulation is the position the stick tends to "center" around varies a considerable amount (as in several inches) depending on trim and flight dynamics. Basically you move the stick to get the aircraft flying the way you want then hold the stick steady in that position and wind on trim until stick force disappears.

Ironically one of the reasons i bought my MS-FFB2 was I naively assumed that FFB sticks would model this behaviour.

Arguably a non centering stick with reasonably friction is a BETTER simulation then one that centers at some arbitary position.


Julian 265: fascinating project, keep up the updates.

rnzoli
08-07-2008, 01:12 AM
Originally posted by WTE_Galway:
The most significant difference that I have noticed between real aircraft and any game level simulation is the position the stick tends to "center" around varies a considerable amount (as in several inches) depending on trim and flight dynamics. The last time I looked that the FFB specification over USB interface, I found it surprising that theoretically the game sofware can tell your FFB stick, to what position should it center. So it doesn't have to produce centering tention only towards the geometric center of play, but to any position. However, I don't think the games/simulators use this possibility.


Basically you move the stick to get the aircraft flying the way you want then hold the stick steady in that position and wind on trim until stick force disappears. Yepp, the lack of correct modelling for this is a general problem in flight sims, it's not the first time I hear RL pilots complain how trimming is more difficult and takes more time in the sim than in real airplanes.


Ironically one of the reasons i bought my MS-FFB2 was I naively assumed that FFB sticks would model this behaviour. There is an embryonic FFB mod thread over the evil red forums, maybe it can be implemented in IL-2. But I would rather start with separating shake effects from stick centering forces into 2 separate devices - only centering tension should act on your stick, the rest (e.g., weather effect) should shake your seat.


Arguably a non centering stick with reasonably friction is a BETTER simulation then one that centers at some arbitary position. The position will always be the same in IL-2 (the geometric center position), and I still prefer feeling the FFB centering forces, which depend on speed. Especially in zoom climbs, one can get recognize the contrast of heavier elevator at the beginning of the climb, and the lightening of control as speed diminishes.


*edit* still got no sensors, still on "back order" status at Farnell since 17th July.

WTE_Galway
08-07-2008, 01:24 AM
Originally posted by rnzoli:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">
Arguably a non centering stick with reasonably friction is a BETTER simulation then one that centers at some arbitary position.

The position will always be the same in IL-2 (the geometric center position), and I still prefer feeling the FFB centering forces, which depend on speed. Especially in zoom climbs, one can get recognize the contrast of heavier elevator at the beginning of the climb, and the lightening of control as speed diminishes.


*edit* still got no sensors, still on "back order" status at Farnell since 17th July. </div></BLOCKQUOTE>

The other issue with games like IL2 is that if you run a stick profile that changes the response around that center point you would need a dynamic profile that moved with the center point. Probably impossible unless it was hard coded into the sim from scratch.

M_Gunz
08-07-2008, 02:21 AM
If you attach your magnet to an iron or steel part (bolt) that the magnet will stick to then
that part and all similar attached will soak up field from your magnet.

Not saying it won't work, it probably will anyway. Just that mounting on a non-magnetic
stainless steel or other non-magnetic material part will leave you with the strongest field.

julian265
08-07-2008, 02:59 AM
That's an interesting idea Galway. I used to have a non-linear response from my previous stick, and when lining up a shot, the stick is frequently away from center, making the 'slow zone' useless and also resulting in an asymmetrical response. A dynamic center point might solve that problem. Fortunately now I don't need a dead zone or slow zone, as my stick has no slack, and the throw is long enough (ie lower angular sensitivity) to allow fine adjustments even with the linear response!

You're right M_Gunz, to add to your point, you can either weaken or strengthen the magnetic field in a particular area by placement of ferrous material. The best way to think of it is that ferrous materials channel more magnetic field through them than the surrounding air.

My rudder pedals have the magnet sitting on a bolt, and the sensor is on the top, and they work perfectly despite the interference.

When making a sensor for a speedometer, I didn't have a suitable magnet, so used a small bolt to channel the field into a smaller area, then through the sensor. When a ferrous object is close to the other side of the sensor, even more magnetic field flows through it, thus increasing it's output by a measurable amount.

Fehler
08-07-2008, 03:23 AM
Originally posted by M_Gunz:
If you attach your magnet to an iron or steel part (bolt) that the magnet will stick to then
that part and all similar attached will soak up field from your magnet.

Not saying it won't work, it probably will anyway. Just that mounting on a non-magnetic
stainless steel or other non-magnetic material part will leave you with the strongest field.

That was also a concern of mine. I ordered N42 magnets to replace the hard drive ones I was going to use. (Should arrive tomorrow by UPS)I understand that the attached ferrous metal will pick up some magnet properties over time, and slowly degrade the magnets. Thus, a wooden or plastic dowel would be the best mounting option. Using epoxy to attach the magnets seems to be the best solution with the materials that I have available.

I have lots of epoxy... I love the stuff!

As a side note: I finally built the base for my flight chair out of 1x1 boxed tubing and mounted the passenger seat from a Honda accord to the rails by welding the bottom brackets to the frame. This gives me the full adjustability of the seat as if it were bolted into a car. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif I mounted the U-Joint gimble to the framing and made the stick shaft to suit the throw I was looking for.

I should have pics by the weekend. My wife took my camera on her vacation... http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif


Posted by Julian:

My rudder pedals have the magnet sitting on a bolt, and the sensor is on the top, and they work perfectly despite the interference.


I noticed this on the pictures you posted and was wondering how your pedals worked. I also wonder if the field of your magnet would decay over time. But then again, how long will it take for a magnet to decay? Probably not in your lifetime... haha! I don't know enough about magnets to say.

M_Gunz
08-07-2008, 05:48 AM
Non-magnetic stainless steel hardware will do the trick nicely.
I like using hot glue. If I don't like the join, I just heat it up and redo.

I learned from a friend how to use crimp connectors, cardboard and hot glue to make custom
PC connectors. He needed something for the 23 pin D on his Amiga. But later on he got a
25 pin body and cut it down to make something more professional.

Yeah, you can form parts with the stuff and it's good for potting circuits. I guess I could
do the same with epoxy and cloth or even cardboard, even make stronger that way but you have
to move fast and you better like what you get!

Chivas
08-07-2008, 08:40 AM
Originally posted by julian265:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Chivas:
I've been looking around for a suitable car u-joint and just thought of using a socket universal joint. Maybe a heavy duty 3/4 or 1" universal socket. I'm also wondering if you could use a single 3" wide coil spring slid over top the the universal joint with enough room for the hall sensors. Just thinking out loud. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

I just looked at my 1/2" drive U-joint. Mine has *a lot* of free play, so you'd need to buy a really good one with closer tolerances to make a good stick. The pivots also looked very difficult to attach magnets/sensors to.

If you want to avoid a full size joint, I suggest a steering U-joint.

The coil spring sounds good and simple... If you can get one wide enough! </div></BLOCKQUOTE>

Thanks for your ideas julian265. Yes the pin could be difficult to attach the magnet too, and some of the pins didn't move properly depending on the type of universal joint. I was going to go to a local tool shop to see what sizes and tolerances are available.

Also in your u-joint how did you adjust the tension so that the joystick didn't flop, but held any position you left it in.

Fehler
08-07-2008, 09:40 PM
Use a new U-joint. They are plenty stiff. They really dont start to break-in until torque and a few hundred thousand rpm's are applied... Hardly what you will be doing to the joint when flying (Or so I am guessing)

Greasing them will also make them stiffer, but with the hole drilled through the cap, I recommend not using a grease gun or the grease will push your dowel pins right out...

Instead, apply a generous amount of grease to the needle bearings before installing the new joint.

I finished my stick today and cant wait to complete the whole project so I can start using it!

Sokol__1
08-08-2008, 08:07 AM
I suggest a steering U-joint.


I get in junkyard a section of steering column from a unknow car. These have two U_join.

But, diferent the one from transmission, these dont have internal holes (for grease) in axis. http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

I have to try with these alternative way to put magnets (9thumbs in Frugal's):
http://users.accesscomm.ca/9thumbs/green-ujoint2.JPG

Sokol1

julian265
08-10-2008, 01:54 AM
Fehler - I'm not sure how long these magnets will last, I'm hoping they live up to the name "permanent magnet"! Though if not, I'll just find a non-interfering bolt, or mount it more remotely.

Chivas - as Fehler said, I tightened mine by buying a new bearing set (AU$30 for my joint), and using grease. I also pressed the bearing cups in quite far/tightly, to get the desired amount of 'hold'.

Re the steering U-joint, that's a good point Sokol, I didn't think of the lack of holes. To get around that, I'd probably use a small grinding wheel to make a slot in the spider's ends, epoxy something like an aluminium rectangular rod into that, which can then mount the magnet. That's if the steering bearings are as hard as the drive-line bearings... if they're not, just drill them.

Fehler
08-12-2008, 05:42 PM
Well, I have been busy at work this last week, so I only got a few more modifications made on my setup. No wiring yet, this is just a work in progress...

The seat/joystick mounting. I welded the bottom of the seat tracks to the boxed tubing. I probably went overboard on the tubing though.. it's 1x1x1/8th tubing. I think I could park a truck on the thing and it wouldnt affect it. The reason I welded the seat track to the frame was so I could retain the adjustable seat. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif I plan on raising the unit off the floor with some larger boxed steel and casters to give it some mobility. Although, I have to say, this thing is getting a little heavy with everything I used to construct it...

http://webpages.charter.net/cuda70/S1.jpg

A pic of the stick in it's mounted location. I used some barrel nuts to mount the stick base to the framing. This way, nothing sticks out on the bottom of the unit.

I have to borrow a friend's sheet metal break and shear to make a nice containment box which I *hope* I can do this next weekend.. Again, depends on work...

http://webpages.charter.net/cuda70/S2.jpg

http://webpages.charter.net/cuda70/S3.jpg

http://webpages.charter.net/cuda70/S4.jpg

This panel box I made will house my throttle, prop pitch lever, and all trim wheels as well as my flap lever. It's made out of aluminum so it won't interfere with the magnets and hall sensors I plan to use. You can't see it in the pictures, but the top of the box is covered by a very thin (2mm thick) piece of plexiglass. I plan on printing out the button labels on thick paper and mounting the paper under the plexiglass. This will look very professional, and the glass will protect the paper from damage. Also, if I decide to change things up, I can simply print another sheet and re-label the buttons to whatever configuration I have. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

http://webpages.charter.net/cuda70/P1.jpg

So there you have it so far. I have the bases for the throttle/Prop Pitch levers made already. Once I have the units assembled, I will show how I did this. They are, like the joystick, quite indestructible...

So far, this has been a fun project, although quite a bit of work fabricating in the 105 degree weather of my back yard!

I didnt keep a record of how much I spent in hours and money on this project so far. I really don't think I want to know! When my wife looks at it she just shakes her head. But hey, she knows I don't go drinking with the fellows or anything else like that, and other playing some recreational softball, this is the only non-paying hobby I have. I do have other hobbies that actually bring money in, so she is fine with those... haha

M_Gunz
08-12-2008, 11:13 PM
When you think about pedals .... here's the best pivots I've found to date.
Centering springs should be a breeze compared to the stick.
The one with the 1000 lb load limit might fit the strong construction theme. ;^)

http://www.nextag.com/Shepherd-Hardware--2700353/lazy-susan-bearing/brand-html

* 12" round plate
* 1/4" ball bearings
* 1000 lbs. max
* Zinc plated

http://www.acehardware.com/product/index.jsp?productId=...searchId=30272550563 (http://www.acehardware.com/product/index.jsp?productId=1296642&cp=&sr=1&origkw=lazy+susan&kw=lazy+susan&parentPage=search&searchId=30272550563)

* 6" sq plate
* 1/4" ball bearings
* 500 lbs. max
* 15" to 30" turntable dia
* Zinc plated

I have a garage door opener bearing you can have... it's taller and doesn't have attachment
holes but it's good to something like 700 lbs.

Fehler
08-13-2008, 07:11 PM
Hey Max, using the Lazy Susan is a great idea! I have chewed on several different ideas for pedals. And yours looks really easy to use!

My CHPro pedals are still working fine, and I may end up mounting them to the frame for now. I really want to get this project "flying". It gets frustrating looking at it and knowing you are so close, but still have work to do...

On a side note, a friend of mine came over today and looked at it. He thought I was out of my mind devoting time and effort into such a project.

This made me step back and a good look at myself and reflect...

http://systemofsystems.files.wordpress.com/2008/03/good-grief-charlie-brown1.jpg



"Am I really a weirdo? Should I be spending my valuable time on, say.. watching the olympics or washing the car or something like that? Perhaps a good book would really be more relaxing; after all..."


Bah... It's fun to make stuff. And that's all that really counts... Haha!

M_Gunz
08-13-2008, 08:44 PM
I'm weird too, and I deal with these things as well. http://forums.ubi.com/images/smilies/88.gif

I have a lot of parts but they haven't come together right. The garage door bearing and pulley
looked good enough to buy but on the practical end there isn't much clearance under the desk
for knee action so I need a lower profile pivot and why not easier to mount as well? Other
problem with that thing is that whatever it bolts to has to be solid enough to take being
wobbled and I don't have a metal shop handy. BTW if you are doing sheet metal and don't know
what bend deductions are, I can give you some tips on how to get very precise around corners.

I have a tested method using IR through gradient printed clear plastic rather than hall sensors
which I also have. I'll be using chip sockets for circuit boards, I have a cheap source and
the LED's and sensors (photoresistor, less light = more resistance, hello gameport!) leads just
plug into the holes making replacement a no-solder affair. I figure to glue the gradient inside
the round edge of the bearing so it's up and curved and moves with the pedal bar then mount the
light and sensor over that so it shines through.
Return springs should also be easy enough. Two sets on one post inside the bearing hole and
one on a pedal bar to hold the spring taut. If the pedal bar moves, the springs stretch more
and want to return to center. No need to balance springs or forces. Before that I was going
to use gravity, the pedal bar would ride up when turned and always return down to center but
that's way more work, too complicated. I still have to get carriage springs and mounts... I
still have to set up a crossbar!
I think my light through the gradient will adapt for sliders really easily. The LED's and
sensors were cheap through allelectronics.com (clearance house) as are sockets and all kinds
of parts. It took a while to figure having the gradients printed up at an office store to
save $ over buying a whole pack of inkjet printable clear sheets. One 50 cent copy makes a
lot of gradient strips.
I've even figured out how to do a multiturn trimmer using led and sensor. I need a turn wheel
or wide knob with fine threaded mounting bolt. As that turns into a threaded hole the end
inside will move nice and slow and linear to block more or less light. The bulb and sensor
are only 5mm wide though and only 4mm of that can be used for linear response so no 20 turns
with any useful thread I know but it's still not a 60 degree move from end to end. Even a
1/4-25 UNF will give 4 or 5 full turns to cover the gap. OTOH I could figure out how to make
the axle bolt move the led and sensor, mount them on a bolt prevented from turning perhaps....

If you like any of that then go ahead and use it. IR led is like 25 cents in singles and the
sensor is likewise cheap. A 2.2k resistor works for the led given 4.5V while the gameport
supplies 5V but in time the led will dim perhaps 10-15% so I've got 1.1k resistors and 1k trim
pots on the way, which are also cheap. Hall sensors alone ran me $1.50 ea for good ones with
nice long leads... first ones I bought were cheaper but surface mount I can't solder or socket.

I could mount gradient and sensors to U-joints even without center pins that turn with them,
and I don't have to be close tolerance to get good readings! Funny how last year I was all for
hall sensors, too. They beat turn pots easily.

julian265
08-13-2008, 10:42 PM
Looking good Fehler. That's going to be a comfy "work" station! Unfortunately I'm restricted to using a normal chair and table http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

The people who think we're weird either have their own hobbies which we wouldn't see the attraction of either, or they're part of the majority who haven't solved problems and MADE something, and not experienced the satisfaction and independence of one's own solutions. Also, choosing what you do on the basis of what is 'normal' will only result in being boring and predictable, and never satisfied.

Gunz, that lazy susan bearing looks good - does it hold itself together (ie, will it come apart if you pull it?), or does it only take compressive loads?

Also, how would you mount the gradient/sensor on a U-joint without holes and pins? I can't figure it out http://forums.ubi.com/images/smilies/blink.gif

M_Gunz
08-14-2008, 12:18 AM
Until I get one that's not part of something else, I dunno just how they stay together but I
have lifted such turntables by the top before and the bottom came up too.

Maybe I can't get the arrangement to work on all universals. If it's two yokes and a cross
piece then it's a matter of clearance to the crosspiece. I don't need to nor can I work this
at the axes of rotation themselves but rather at a radius. If it's two yokes and a circle or
square frame I can still do it again only with clearance. Those little universals for tools
don't have much clearance. But then you can do the same with the hall setup on any of those
too.

And BTW, before I got into coding I was a production engineer at a small precision metal shop.
I made stuff work that made stuff work, that's how I got into programming. It's a hell of a
lot easier when you have the tools and materials right there, I miss having access to all that!

Urufu_Shinjiro
08-14-2008, 09:05 AM
I wish I had half the talent you guys do. I'm horrible at making things with my own hands, I'm forever doomed to buy cool stuff off of guys like you, lol. Once some of you guys get this stuff worked out how much do you think you would charge for a u-joint stick with all the stuff installed and ready to go, assuming you would want to repeat the experience, lol.

M_Gunz
08-14-2008, 02:56 PM
I don't have access to tools needed to even think of that u-joint stick myself but I can say
that the shipping alone would be a good bit. Perhaps someone like Red Star would be cheaper
than UPS, them's pallet load kind of things.

TX-EcoDragon
08-14-2008, 05:00 PM
Originally posted by Chivas:
Love your idea Julian. If I remember correctly yours doesn't have a centering force. I kind of like that idea where the controls stay where you leave it. How does it compare to flying a real aircraft? Is the only centering force in aircraft of the era the air resistance forcing the ailerons and elevators to a more neutral position?

Centering forces in flight are quite heavy.

Different designs have different amounts of "breakout force" and different amounts of centering force, as well as differences in how much the force ramps up as the stick is moved further towards it's limits. This is especially obvious when comparing newer designs to older ones, or designs that have different goals - dedicated aerobatic aircraft vs transports for example will be on different ends of the spectrum with respect to stick forces.

Many WWII era aircraft would need 45-60 pound forces on the pitch axis for 3G pulls, and more for more spirited flying or at higher speeds. Some, like the 109 at high speed needed much more.

So to be realistic at all, yes, there must be centering forces, and some degree of breakout force. In a case where the stick sits where you leave it, not only will you be unable to determine the center position by feel (critical during maneuvering flight) but trim will no longer apply since you will simply have no froces to contend with, and no hands off center position - just leave the stick wherever. Certainly a couple light springs in a joystick is a far different experience than a real stick feels, but it's better than no springs without a doubt.

That said, a centering system for this stick should be easy enough to incorporate, even after you've already built it.

Another thing that people doing these types of sticks need to keep in mind is that if you build a stick with realistic length, travel, pivot point, and placement, you will also have to mindful of the fact that most PC rudder pedals sit at a more flat angle than real rudder pedals, and are very narrow, as a result your legs will be in the way of the stick even at mid range travel. Most WWII era designs result in me hitting my legs even with them on the sides of the cockpit, and that's with rudder pedals that are often twice (or more) as wide as CH pedals.

M_Gunz
08-14-2008, 09:30 PM
Originally posted by julian265:
http://www.members.optusnet.com.au/julian265/il2/unistick/both_sensors.jpg

There's two yokes there with a crosspiece in the middle.
If you can attach a very strong tension spring between the flat top of the crosspiece and the
middle of the U shaped yoke above then any time the upper yoke swivels that spring would stretch.
That's because the flat is not on the axis.

A similar treatment on the underside of the crosspiece to the base yoke center would do that axis.
1" coil of 1/8" wire... might just work? (http://www.thomasnet.com/catalognavigator.html?cid=403137&cov=NA&which=prod&what=Springs%3A+Extension&heading=77660801&searchpos=3&cnurl=http%3A%2F%2Fhardwareproducts.thomasnet.com% 2FCategory%2Fextension-springs&prodpos=1&searchpos=3)

Compressions springs for punch and die work can run up over 300 lbs per inch of squeeze.
Example -see H75-100 (http://www.daytonprogress.com/catalogs/pdf/22.pdf#page=8)

A compressive spring solution, I think would have two springs per axis needing to balance at
center but could still fit inside the joint. It would require adding base studs to yokes and
crosspiece and could be a _real_ bear to put together!

But... it doesn't look like there's two inches between yoke and crosspiece. Be neat if you could
since there'd be nothing external to bump against.

Fehler
08-15-2008, 06:27 PM
Sokol added this as one method of self-centering...

http://users.accesscomm.ca/9thumbs/u-joint+electronics.jpg

Looks pretty simple.

M_Gunz
08-15-2008, 10:55 PM
I get the feeling those u-joints are smaller than I thought, or at least this one is.

julian265
08-15-2008, 11:14 PM
FYI Gunz, there's about 20 mm between the top of the cross-piece and the yoke above it on mine.

I won't be adding springs to it though... I can't say I miss setting trim!

M_Gunz
08-16-2008, 05:08 AM
That is tight and I can easily understand not wanting to have to deal with IL2 trim,
trimming IRL is a bit different procedure and experience.

Well I'll have room in the pedal setup.

Still, tapping in trim and then loosening up on the stick does simulate the lessening
of stick (or yoke) forces, but moving the hand is wayyyy opposite my experience and the
game trim is missing the whole part of being able to stop when the force is relieved.

What I'd *like* is to hit a button for one axis and then okay, I'd have to move my hand
towards joystick center but all through that movement the movement itself would be taken
as the trimming until the joystick centered. For one thing, there would be no trimming
while pulling back further 1337-bat-turns.

squareusr
08-16-2008, 09:16 AM
Originally posted by M_Gunz:
What I'd *like* is to hit a button for one axis and then okay, I'd have to move my hand
towards joystick center but all through that movement the movement itself would be taken
as the trimming until the joystick centered.

I've thought of this too, and it should be easy enough to implement using glovepie (if you use a strong nonlinear joystick curve in il2 you might have to replace that with a similar transformation in glovepie).

Move stick in position for desired attitude, press "sample and hold" button (script saves position), move stick to center (script continues sending the saved position), release button (script adds difference between stick position at button-release and stick postition at putton-press to all subsequent stick postitions).

I have not tried it myself yet because:

1.) i'm not entirely sure if the IL2 flight model would really treat a stable off-center stick position the same as a stable center+trim

2.) it adds trim axes to planes that did not have it it irl and/or do not have it in IL2, this is not very attractive in my book (even when it's just "cheating on the AI", which is probably controlling it's dumbed down flight model with 32 bit floating point precision)

3.) i'm still occupied enough of my IL2 time with building/improving my controller hardware

4.) i have put enough time and effort into my current trimming solution (software to connect MIDI controller hardware to devicelink) to make replacing it feel like a loss, even if the replacement was superior

rnzoli
08-16-2008, 11:02 AM
Originally posted by squareusr:
1.) i'm not entirely sure if the IL2 flight model would really treat a stable off-center stick position the same as a stable center+trim
On this note, I think IL-2 treats the two situations in the same way, if the joystick sensitivity curve is flat 100%. Trim is just an offset of the stick, as far as I undestand, and you may notice that your in-game stick/rudder changes position with trimming, even when you don't move the controls.

Which is probably incorrect modelling, as in case an aircraft has separately movable trim tabs, the trimming should change the force on the stick, and may have a slight effect on the drag of the control surface (maybe miniature amount only).

rnzoli
08-16-2008, 11:09 AM
Originally posted by Fehler:
When my wife looks at it she just shakes her head. LOL, MY wife also shakes her head when I show YOUR pictures to her. http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif But it's a great help, to be honest, because this way, my wife is not so upset when I do some similar things at home, in smaller scale.

The interesting thing is that my kids are quite amused about the amount of time and effort I put into my own hardware and my elder daughter, who has absolutely no affinity with aviation, squarely said she was proud about seeing this sort of enthusiasm from me http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif Who can't handle a bit of "wife's head shake" now and then, when you can also get such compliments? http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

M_Gunz
08-16-2008, 03:58 PM
A squadmate and I had done some things with trim way back in the end of Dec 2001 and Jan 2002
and tracks were sent to Oleg. There was a lot of question about just what the trim was doing,
the 109G-6-A/S would do fantastic climbs when trimmed with rotary.
The answer we got was that holding joystick by hand there is small amount of flutter on the
digital level that translates larger and makes flight less efficient. And there were changes
made in the successive patches, Oleg would have the solutions to hardware limits improved.
The hardware limits though, and the time limits and code limits are there -- IMO 4.07 was a
big positive step in the controls interface that like others was taken by so many as FM change.

In a real plane you hold the controls and fly as you want to set trim then you wind in the
trim until the controls are not putting force on your hand. We have no way to replicate that
entirely with PC hardware, perhaps because IL2 is set up for conventional sticks.
What we do have is good for getting a diminished feel of the forces involved which I like
better than what I had before IL2 which was entirely positional, without trim and arcaded in
the sense that the game helped out with both stick and rudder as hardware limit solution.

ADD: does the longer stick get rid of the haveta-be-trimmed-for-best problem?

julian265
08-17-2008, 06:33 PM
I'll have to test attainable level speed with trim adjusted vs offset stick.

When I moved from my original logitech stick to a BU0836 equipped one, I noticed an improvement in the way the planes held their speed, and the ease with which they reached their maximums.

I put it down to two things - the BU has four times as many steps as the logitech, so each step is four times smaller. When both sticks flutter between two steps, IL2 would have been receiving an input four times larger from the logitech than the BU, hence depriving more speed.

Secondly, the logitech had a spiky output, which was most noticed when an early carburettor equipped engine would cut out if the plane was pitched forward *as gently as possible*.

M_Gunz
08-17-2008, 08:27 PM
I had taken to using a lot of filter on the stick sensitivity page. It dampens flutter but
hey less is still less. The filter, and I use about 40%, also smooths control moves.

I had a Logitech when my old sidewinder didn't work on my new PC back in 2001 and I wished I
hadn't bought it. That sidewinder was optical encoded and nicely precise.

rnzoli
08-21-2008, 09:00 AM
I don't know what this means, but my >1 month long order at Farnel changed status from BackOrder to Processing recently... maybe there's hope? http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

FoolTrottel
08-21-2008, 09:28 AM
Originally posted by rnzoli:
I don't know what this means, but my >1 month long order at Farnel changed status from BackOrder to Processing recently... maybe
there's hope? http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif
Status has changed over here too... can't be long now... http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

M_Gunz
08-21-2008, 09:58 AM
You guys might want to check out Digikey.com next time. If they say they have em in stock then
the order goes out in a day or so and yes they do ship to over 100 countries listed there.

FoolTrottel
08-27-2008, 04:31 AM
Received them sensors today, finally!

rnzoli
08-27-2008, 05:44 AM
Same with me, got it on Monday. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif
Just when the sky fell down at work and will be working over the weekend too. Travelling abroad this week and next week as well. *sigh*

julian265
08-27-2008, 07:11 AM
wow... that's pretty poor, but at least you have them eh? I hope you ordered enough to last you a while!

FoolTrottel
08-27-2008, 01:36 PM
Can't get it to work...

Have connected one sensor replacing a pot on a Trust joystick. Used the same three wires as connected to the original pot, making sure +5V and GND were connected OK, then the remaining wire must've been 'Output from circuit'.

The Game Controllers / Test item 'sees' the sensor, as in disconnecting the sensor results into a different signal.
However, holding it next to a magnet, rotating the sensor or magnet does not result into any change in the signal...

Weird stuff...

- Used A1321EUA-T "” ALLEGRO MICROSYSTEMS "” Hall Effect Sensors & Switches from farnell.
http://img.photobucket.com/albums/v668/fooltrottel/Magnet.jpg
- Used magnets from Supermagnete.de (http://www.supermagnete.de/eng/magnets.php?group=cubes&switch_lang=1) (W-05-N, W-07-N and W-10-N, 5,7 and 10mm resp.)

- With the sensor disconnected, there's 5V+ on Power vs. GND
- With the sensor disconnected, there's 2.66V+ on Power vs. 'Output from Circuit'
- With the sensor connected, there's 1.78V on Power vs. GND
- With the sensor connected, there's 0.88V on Power vs. 'Output from Circuit'
- Tried 3 sensors out of a batch of 25 ...

Where did I go wrong? :-(

Sokol__1
08-27-2008, 04:35 PM
FT,

Made a calibration in Windows Control Panel, aply and test again.

Sokol1

M_Gunz
08-27-2008, 05:27 PM
Originally posted by FoolTrottel:
- Used A1321EUA-T "” ALLEGRO MICROSYSTEMS "” Hall Effect Sensors & Switches from farnell.
http://img.photobucket.com/albums/v668/fooltrottel/Magnet.jpg


From the datasheet, see page 10 of 11 (http://www.allegromicro.com/en/Products/Part_Numbers/1321/1321.pdf)

Power goes on the left, ground on the right and signal out is center.
Best to check operation with a meter, you are sure then with nothing else to be a problem.
If not a meter, a resistor and red LED or small light bulb can do.

I have an old projects kit that makes hookups easy with many parts mounted and ready.

FoolTrottel
08-27-2008, 06:18 PM
Originally posted by Sokol__1:
FT,

Made a calibration in Windows Control Panel, aply and test again.

Sokol1
Thanks for the tip...
Tried it... but.. that one axis I converted does not register... so calibration gives no result at all... the bugged thing don't move...

julian265
08-27-2008, 06:20 PM
Originally posted by FoolTrottel:
- With the sensor disconnected, there's 5V+ on Power vs. GND
- With the sensor disconnected, there's 2.66V+ on Power vs. 'Output from Circuit'
- With the sensor connected, there's 1.78V on Power vs. GND
- With the sensor connected, there's 0.88V on Power vs. 'Output from Circuit'
- Tried 3 sensors out of a batch of 25 ...

Where did I go wrong? :-(

*POST EDITED*

It's possible that the joystick circuit is incapable of supplying the small amount of current needed for one of these. Test them out of the circuit using an external *regulated* 5V supply and multimeter.

If this solves your problem, you'll still be able to use them in your stick, you'll just need to use the 5V USB (or other) supply in your stick directly, rather than the existing potentiometer supplies.

The sensors are meant to draw around 8mA at 5V.

FoolTrottel
08-27-2008, 06:21 PM
Best to check operation with a meter, you are sure then with nothing else to be a problem.
I did use a meter, that's where I got the V values from...

There's no variation in the values... when moving magnet-sensor....

M_Gunz
08-27-2008, 07:13 PM
Yeah, you're right, you do show voltages.

I see a huge drop in power to ground volts between sensor disconnected to sensor connected
in your numbers and that tells me maybe the supplied power isn't up to the current demand
from your device. I find that odd, those sensors are supposed to draw at most 8 mA.

However the joystick circuit may be doing its own thing like testing in pulses.
The data sheet does show example with a bypass cap between power and ground, something normally
used in devices that can soak power when run in clocked pulses to reduce circuit noise.

You might try using the recommended .1 uF capacitor across pins 1 and 3 as shown unless that
is the exact replacement part and the old one had none?

Have you verified these things work outside the joystick circuit? Could they have been
damaged in some way?
Consider that 3 AA batteries in series will give 4.5-4.8 V at 100 mA.

FoolTrottel
08-27-2008, 11:27 PM
Have you verified these things work outside the joystick circuit? Could they have beendamaged in some way?Consider that 3 AA batteries in series will give 4.5-4.8 V at 100 mA.
Now that's a great idea, draw +5V from a different source / different place. Will try. Ty.

M_Gunz
08-27-2008, 11:48 PM
By the spec in the data sheet, 4.5V will work.
Fresh batteries will supply over 1.5V each if you want to go the quick route.

ADD:

I wouldn't try using power from the PC just on general paranoia though I will hook working
circuits up. You can get robust sourced 5V from printer port, they are made to take a lot
but one bad screwup and it's bye-bye printer port!

Alternative would be a wall-wart type power adapter if you can find one at 5V. With a 5V
regulator chip you could use more but after seeing the trouble you have getting parts, it's
your call!

If you've got a spare PC power supply then be sure to plug something that draws decent power
into one of the plugs, like a spare hard drive. Switching power supplies should have some
more than miniscule load to run and regulate properly, or at least they did in the 90's.

FoolTrottel
08-28-2008, 01:26 AM
it'syour call!
Lol, I'm certainly aware of that!

Risking a joystick (though it's a Trust, taken apart) and a powered hub... it's all cool... I hope http://forums.ubi.com/images/smilies/partyhat.gif

FoolTrottel
08-28-2008, 05:48 AM
Well, taking the +5V from a different wire does work.

One thing kinda worries me: There's quite a noticeable deadzone, at center.

Is this because I need a higher resolution in the electronics?
The Trust stick just goes 0-255, with that zone at 127, in which the slider is moved, but value stays stuck at 127 for some 2 or 3 dgrs I guess.

Must get me a BU-Thingy...

M_Gunz
08-28-2008, 08:08 AM
You need to verify the sensors even work as a step in troubleshooting the problem.

Big part of that is separating the sensor from more complicated things like the joystick,
you call Trust means Thrustmaster by any chance?

You have 5V and ground, hook power and ground then a resistor maybe 2k ohm between output
and the same ground and watch for *current* coming from output across that resistor.

Wait! You need a resistor between power and your 5V source! You *have* to limit the
current going into the hall sensor! Spec sheet says 8 mA maximum current source!
Amps = Volts / Ohms, to limit 5V current to 5.6 mA takes 893 Ohm resistor!
Resistor doesn't have to be exactly 893 Ohms, anything that allows up to 8 mA is good,
from much more down to 625 Ohm (probably want more at minimum like 700) is okay.
Without that your sensor _may_ draw enough current to burn it out... I hope not already.

It is the current that is the signal, not the voltage in these digital devices.

Sensor works or it doesn't is step one.

IF it works then on to step two, putting it in the maybe other problems joystick.
--- what part in that are you replacing? You wrote pot.. hall sensor is sort of, sort of not.

It is the not I now worry about and perhaps I should. Those voltage drops you show when
connected are exactly the kind of thing that happens when current drawn is higher than
the power supply provides, it cannot keep the voltage up.

You can use a meter, do you also know basic DC electric -- Ohm's Law and like?
From what else you write I think you do but then there are these idea shadows that happen.

julian265
08-28-2008, 08:23 AM
FoolTrottel are you sure your magnet is orientated correctly? Is it also possible that the deadzone is set in the joystick's hardware, or even in an old calibration in your computer?


Wait! You need a resistor between power and your 5V source! You *have* to limit the
current going into the hall sensor! Spec sheet says 8 mA maximum!

It is the current that is the signal, not the voltage in these digital devices.

Gunz, these sensors should not have any resistor between their power pin and power supply, nor are they current-signal devices. They are more than a variable gain transistor, they've got an internal circuit which takes in ~5V and outputs a voltage signal.

M_Gunz
08-28-2008, 08:28 AM
I'm trying to type and calculate and read the device docs at the same time.

By the datasheet, supply current should be 8 mA MAXIMUM for that device.

Check it out, top of page 3. (http://www.allegromicro.com/en/Products/Part_Numbers/1321/1321.pdf)

ADDITIONAL SO YOU KNOW I EDIT:

The HALL SENSOR is not what I refer to as current-signal device.
The current is the signal to what the sensor is connected to.
TTL inputs measure current flow.

PPS:

I = E / R --- vary R and you vary I

julian265
08-28-2008, 08:45 AM
Ah ok, I did think you were referring to the HE sensor. However if FT's stick did use current sensing, there's a large chance that the sensors would simply not work with the circuit. Also the pots wouldn't need three wires if it did. IMO it's possible, but unlikely.

The 8mA is the maximum draw of the sensor - ie, what it's internal circuit will draw if given ~5V. It doesn't need to be limited externally. (with the 5V supply I use (1 amp 7805 reg), I'd know if they did require current limiting by now!)

FoolTrottel
08-28-2008, 08:56 AM
http://forums.ubi.com/images/smilies/blink.gif
Hey guys, thanks for your interest and support, but either my typing is not very clear, or a post of mine must be invisible or something, as IT IS WORKING !!

+5V from a different spot on the stick's electronics, and bingo, it works okay.

Apart from that deadzone I mentioned...

(Did my earlier post even show up for you? The one date/time 'Posted Thu August 28 2008 04:48' ??)

I've even tried a sensor on my homebuilt pedals, and it works fine there too, the deadzone seems no problem there...

So, I will check alignment of the magnets, though I think I got it right...

M_Gunz
08-28-2008, 10:23 AM
Originally posted by julian265:
Ah ok, I did think you were referring to the HE sensor. However if FT's stick did use current sensing, there's a large chance that the sensors would simply not work with the circuit. Also the pots wouldn't need three wires if it did. IMO it's possible, but unlikely.

The 8mA is the maximum draw of the sensor - ie, what it's internal circuit will draw if given ~5V. It doesn't need to be limited externally. (with the 5V supply I use (1 amp 7805 reg), I'd know if they did require current limiting by now!)

I sure hope so cause if he's supposed to limit that current then he may have junk by now.
The spec sheet says Supply Current and shows 8 mA max, I'd use the resistor just in case since
it won't hurt.

There are voltage fed ADC's but the cheap stuff just fills a capacitor and then uses the time
that takes to bleed out through a resistor as the value, takes milliseconds to do and costs
very little. At the resolution and speed of even the device you're using it's practical
though the BUO <sp?> may use something less primitive... it looks like a microcontroller
with simple A/D circuits though.
The Wiki on Analog to Digital Converters (http://en.wikipedia.org/wiki/Analog_digital_conversion)
has interesting notes down in the ADC structures section but it should be noted that many of
those require more chip than the BUO device has and are far-far overkill for the application.

Still, I don't right now _know_ what's in that microcontroller he's using.

Since voltage and current are so related anyway I guess even with the cheap stuff you really
read watts anyway. I just think of it as current from the mindset of how I made other things
work long ago -- now that I think of it!

M_Gunz
08-28-2008, 10:39 AM
Sorry FT, I saw "it's working" but still there was problem.

At least you don't have to have an extra resistor or even a bypass cap.

FoolTrottel
08-28-2008, 11:11 AM
Well, that's just it...
There's a deadzone.
But, is it a problem? http://forums.ubi.com/images/smilies/16x16_smiley-surprised.gif

Maybe it's just the Trust's electronics.
As it did work for the pedals (Based on a Thrustmaster stick) I will keep fiddlin', and in the end get me a BU-board...

M_Gunz
08-28-2008, 11:16 AM
That board has superior digitizer to gameport and more of them.
You could hook to gameport if you wanted. It is 0-255 only, but free if you already have one.

Sokol__1
08-28-2008, 02:22 PM
Maybe it's just the Trust's electronics.

With pots these deadzone still?

Same cheap joys have a deadzone using pots.

Sokol1

FoolTrottel
08-28-2008, 10:32 PM
With pots these deadzone still?
Hm, good question.
I'd have to hook up that pot again then, I never tried it... http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

FoolTrottel
08-29-2008, 08:41 PM
Same cheap joys have a deadzone using pots.
http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif
Hooked up the original pot... and there it is... the same deadzone.
Thank you Sokol!

M_Gunz
08-29-2008, 09:46 PM
FT, you have gameport on your PC?

FoolTrottel
08-30-2008, 02:07 AM
Yeah, I do have a gameport...

rnzoli
08-30-2008, 02:46 AM
Originally posted by FoolTrottel:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Same cheap joys have a deadzone using pots.
http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif
Hooked up the original pot... and there it is... the same deadzone.
Thank you Sokol! </div></BLOCKQUOTE>
Good call indeed. The BU-thingy will solve that, but don't be surprised to get a little bit of ripple instead due to the higher resolution (and the apparent lack of digital filtering in it).

M_Gunz
08-30-2008, 08:46 AM
FT, you can get 25 pin connector for not a whole lot of $ and hook to that port.
I don't see them online but I used to buy ones with crimp on the wire ends pins that would
insert through holes in the plug body, not just for hobby but for business use I did.
They are quick and work fine without need to even solder.

The gameport is 0-255 and no deadzone.

Here are the kits for sale. (http://www.showmecables.com/D-Sub-Connector-Kits.html)

At least you can see them, Radio Shack here used to carry them but they don't any more.

It's a real pain finding the right search words. I can find all the pieces at many places
but not all in a simple kit, I ended up buying solder-types and hoods at not much better
prices at allelectronics.com.

M_Gunz
08-30-2008, 09:01 AM
Originally posted by rnzoli:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by FoolTrottel:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Same cheap joys have a deadzone using pots.
http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif
Hooked up the original pot... and there it is... the same deadzone.
Thank you Sokol! </div></BLOCKQUOTE>
Good call indeed. The BU-thingy will solve that, but don't be surprised to get a little bit of ripple instead due to the higher resolution (and the apparent lack of digital filtering in it). </div></BLOCKQUOTE>


Unshielded wires allow the current flow along one wire to be sensed on a close neighbor, that
is how sensitive more than 8 bit resolution can give. It depends on how long the cable is and
how much current is being run in circuit.

At the BU0836 site: Good wiring helps. For ultimately clean signal use shielded wires and ground the pot's case if it's metal (http://www.lbodnar.dsl.pipex.com/joystick/)

rnzoli
08-30-2008, 10:14 AM
Been there, done that, very useful advice but will not solve the problem "ultimately". In my tests, no matter how clean a signal was (no cables, direct soldering onto inputs with shielded pot house), - if it fell between the detection thresholds, the flip-flopping ccurred anyway. Not a big issue, but can produce some strange graphical effect in-game (shaking stick). Nevertheless, the input to the flight model is very smooth.

M_Gunz
08-30-2008, 11:00 AM
You will get some dithering anyway, the thing to do is add Filter on the stick sliders page.

Did you really use individually shielded wires? What I was on about is if moving one axis
affects another to the small degree, not random spazzing. There's shielded cable that is
shielded from outside fields but all the wires inside are unshielded from each other. I've
used plenty of that but it's not the ultimate at all.

Sokol__1
08-30-2008, 03:45 PM
FT, you can get 25 pin connector for not a whole lot of $ and hook to that port.

Ops! (DB)25 is for LPT (printer) port.

For gameport use DB 15 (male). http://forums.ubi.com/images/smilies/25.gif

Sokol1

M_Gunz
08-30-2008, 07:29 PM
Oops is right! DB-15 is the gameport one and it's the one I have solder-type plugs for.

Mea Culpaaaaaa! I haven't looked often enough at the backs of PC's in a few years now.

DB-25 is also full size RS-232 but in opposite gender of printer port, IIRC.

The gameport pinout is easy to find on the web, ya would have spotted that gaff in a second!

FoolTrottel
09-07-2008, 05:58 PM
Bump!

And: I've replaced the rudder-pot on my pedals with a HAL sensor + magnet (5mm cube)

Result is good, much better control. More accurate. http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

This rudder control is based on the X-axis of a Thrustmaster Top Gun Fox 2 Pro Joystick.
Other axis on this stick are in use for Brakes (Y, right rudder pedal), flaps (Z) and Aileron Trim. (Slider, probably 'Throttle'). Next I'm gonna try to HAL-up the Aileron Trim.
(flaps + brakes do not need the more accurate control in my opinion... http://forums.ubi.com/images/smilies/metal.gif)

After that's finished, I should try to upgrade my Throttle control, based on the X-Axis of a Trust QZ-501 Predator Joystick. However, this may be tricky to impossible, as an initial test showed that 0-100% happened within 0-3 dgrs of deflection... http://forums.ubi.com/images/smilies/16x16_smiley-surprised.gif that's no good! (I could try and use a different axis for that though... http://forums.ubi.com/images/smilies/53.gif )

Oh well, I'll just take it one step at the time! (Which is basically -I'm certainly aware of this- postponing the ordering of the BU0836X ... http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif)

M_Gunz
09-08-2008, 07:09 AM
Have a look at this connector kit at $1 each. (http://www.cableclub.com/db15-male-8p8c-modular-adaptor-p-2145.html)

I've seen the like before, the modular adapter is on the back of that hood, not shown and
listed as 8P8C (8 position, 8 contact) which on lookup is eithernet cable, big phone jack.

Same place has cables cheap, even 10ft (3m) for $4.
DB15 male goes to gameport which has 5V and ground pins.

FlyingColander
08-28-2009, 01:06 PM
One year has passed since you worked on your terrific control column. I am going to give it a try thanks to your detailed posts and How To.

Just wanted to see if a year on you have an extra insight about your project? Are there things you'd do differently? Improvements? How is the stick holding up to use?

Thanks again for the time you put in to share this with us.

Cheers,

the Flying Colander

Sokol__1
08-29-2009, 09:47 AM
One squad mate build this desktop joystick, based in Julian tuto:

http://img40.imageshack.us/img40/8652/gripfinal3.th.jpg (http://img40.imageshack.us/i/gripfinal3.jpg/)

http://img113.imageshack.us/img113/558/hodas002.th.jpg (http://img113.imageshack.us/i/hodas002.jpg/)

http://img242.imageshack.us/img242/1231/modconj007.th.jpg (http://img242.imageshack.us/i/modconj007.jpg/)

http://img27.imageshack.us/img27/3879/conj002.th.jpg (http://img27.imageshack.us/i/conj002.jpg/)

Throttle:

http://img26.imageshack.us/img26/9118/manete007.th.jpg (http://img26.imageshack.us/i/manete007.jpg/)

Fiberglass grip, HALL sensor, BU0836...

Sokol1

julian265
08-30-2009, 01:22 AM
^ nice! That's a good sized joint to use. Not too big, not too small to make things fiddly.


Originally posted by FlyingColander:
One year has passed since you worked on your terrific control column. I am going to give it a try thanks to your detailed posts and How To.

Just wanted to see if a year on you have an extra insight about your project? Are there things you'd do differently? Improvements? How is the stick holding up to use?

Thanks again for the time you put in to share this with us.

Cheers,

the Flying Colander

I'm glad you like it! You've made me think... The stick still works perfectly, and there's nothing that I've needed to re-design or repair. The dowels became loose once, but I just pressed them in further and they haven't loosened since. The joint has lost some friction too, but not enough to make me tighten it (I'd just have to press the cups in a bit). If you use springs this issue wouldn't arise, but I'm quite happy to be spring-less.

Adjusting the sensor position and magnet angles can be a bit fiddly, but because the BU0836 won't lose calibration, this process should only have to be done once, and it's not *that* bad. So I'm not sure that the extra effort of designing an easier to adjust mounting system is worth it.

I guess an alternative to wooden dowels would be carefully selected bolts - you could hammer them into the cross-piece grease holes (their threads would have to be just a little larger than the holes), then cut off their heads, and use them to mount either the sensor or magnet on lock nuts, or some other internally threaded pieces (aluminium is nice to work with).
This method would provide a few advantages: The bolts wouldn't ever come loose, you could get them out by tightening a nut against the cup, and of course they'd provide an adjustable base for either the sensor or magnet.
There may be plastic bolts that would suit the purpose too, but if metal ones are too tough you can always soften them with some heat treatment (just get them glowing, then let them cool slowly).

Maybe some of those wall plugs would fit the cross-piece too, which you could then screw into, if they grip well enough.

I'm happy to answer any further questions if you've got any, and would be interested to see your build.

FlyingColander
08-30-2009, 11:24 AM
Sokol and Julian,

Thanks for your replies. Both of you have done so much to encourage others in our small community.

Please congratulate your friend, Sokol, on his excellent craftsmanship. I'd be curious to know how his calibrating of the hall sensors went. His hall transistors seem mounted much further away from the magnet than Julian's were.

Julian, your post-construction analysis will serve me well. I must admit I was a little skeptical of the dowels upon first look. But you've shown proof of concept and proven the old addage, "if it ain't broke, don't fix it!" Good tips for other routes though.

I look forward to sharing my own project once I'm done converting my old CH gameport pedals. For those of us without a good rubbish tip, do you know any place off hand to find universal joints for order?

Cheers,

Flying Colander

julian265
09-01-2009, 03:25 AM
Please excuse the verbose post, I'm not sure what you know already, so I went through most of it.

Short answer: any mechanic, or wrecker, or DIY wrecking yard (pick-a-part, u-pull-it, etc). I suggest you buy used parts, and/or pull them yourself from a DIY car wrecking yard. If you're going to get them off the shelf though, you'll need to know what make, model and year the joint is for, and which position in the drive line it's from.

I used the rear-most uni joint and yokes from a 1980s volvo 240 (244, 245, 264 and 265 will use the same joints too). Or you could visit the nearest DIY wrecker just to look under the cars, to see which one is best, then remove or order the parts for that car.

http://www.jpfiles.com/misc/rear_shaft.jpg

The one I used (before I shortened it...)
http://www.jpfiles.com/misc/shaft.jpg

Note that the one shown in the first picture probably isn't suitable, as it doesn't have a flange (and is therefore harder to attach to a base). Also, most cars aren't as easy to get under as this one! The section you need isn't necessarily the rear-most one, it just needs a flange on one side and a shaft on the other (must be at least as long as the stick you want). Often sections of drive-shaft are joined by fitting one into the other (to allow forward/backward movement of the axle), so you just need to pull them apart, once you've unbolted the flange and knocked it to one side.

The whole uni joint consists of a cross piece (aka spider), bearing cups, bearing rollers, two yokes, and grease. The yokes are usually welded onto the drive shaft or a flange. Normally when a uni joint wears out in a car, the mechanic takes the drive shaft out of the car, removes the cross piece and bearings, then fits a new cross piece and bearings to the original yokes. Hence a uni joint kit consists of just the cross piece and new bearings, but not the yokes.

To buy the whole lot new would likely be too costly for a project like this one. Hence I suggest you either obtain your parts from a DIY car wrecker (the kind where you walk into a large car yard and unbolt the parts you want), or buy used ones from a mechanic or other wrecker. Finally, if your used joint has too much slack for your liking, buy a new uni joint kit (mine was AU$30, be sure to get a cross-piece that has a grease nipple - this indicates that there's a hole through each axis), and have that fitted to your yokes, or fit it yourself.

If you want to remove the parts from a wrecking yard yourself, you should be able to get away with some spanners or large shifters (between 10mm and 19mm bolts usually, but it'll be different for each brand). I usually put a spanner on both ends of the bolt, then bash one with a hammer to jolt it loose. It may help to jam the other one against the car body or other parts before you start hitting. Always give the car some good shoves before you get under it, to verify that it's not going to fall on you.

Removing and fitting the bearings and cross-piece to the yokes requires a decent bench and vice, a large hammer or other thumping device, and a lot of force. I had never replaced a uni-joint before doing this joystick, and I'm glad it didn't have to go back on the car, as it copped a bit of a hammering. This is much harder than getting it out from under a car, so you may want to pay a mechanic to replace the joint, which shouldn't cost much anyway.

FlyingColander
09-02-2009, 03:11 AM
Julian,

Thanks for the advice on finding a good uni-joint for my project. As it happens, I walked into the neighborhood motorcycle mechanics and they got me something new for just $20. For others also looking, Ebay has great deals too.

the Flying Colander

Fehler
09-05-2009, 03:03 AM
When I first built my stick using the Julian-Gimble as I like to call it... (U-Joint) the only thing I discovered to be difficult about the whole affair was drilling holes in the U-joint caps as they are hardened. I ended up popping a small hole in them with a plasma cutter and then drilling from there... made it really easy.

about a year (or more) later and the stick still works flawlessly. I was able to coax a KG-13 grip from a friend that makes them to use as my grip and I absolutely love it.

I have used both BU0826 and BU0836A's for the stick, and both are equally good. I discovered that my first generation BU0836 would not work with x64 bit operating systems, so now my pit is powered by two BU0836A's running all the trim axis and a separate button for every function in IL2 with room to expand.

I have started drafting up my quadrant for the next release which is being advertised as supporting multiple throttle and pitch inputs (yay!!!)

julian265
09-06-2009, 02:00 AM
^ great!

How long ago did you buy those 836A's Fehler? I'd love to get one or two more, but Leo isn't responsive to emails.

FlyingColander
09-09-2009, 01:11 AM
Hey Julian and other builders,

FYI:

Allegro is discontinuing production of the A1321EUA-T sensor. The replacement is the A1321 LUA-T. Essentially this is the same package only smaller. Really small. Only 2x4mm!

Fehler
09-12-2009, 01:57 AM
Originally posted by julian265:
^ great!

How long ago did you buy those 836A's Fehler? I'd love to get one or two more, but Leo isn't responsive to emails.

I was fortunate to find a new BU0836A over at a racing forum. A guy bought a bunch of stuff to make a racing pit and never finished the project. Then his schooling/work took him in another direction in life. Again, I was very fortunate.

I was talking to a good friend of mine from Germany who, in turn, is good friends with Leo. Leo was quite ill for some time, but he should be back up and running very soon. (I am so glad that wasnt too serious, or at least if it was, he pulled through!) Leo has always been a great guy to deal with. A very honest fellow with a passion for his work.

I noticed that Tarmac Aces and others that produce similar sim equipment using the Bodnar controller are not even taking orders right now because they equipment is based off of Leo's controllers.

It is quite amazing that Leo has really cornered the market with his little units. They are very simple, compact, and yet very robust. The first Hall sensor I ever hooked up I crossed the signal and ground wires. Within seconds I could small that burning electrical smell (You know, the one that spells instant doom)

The sensor got so hot so fast that I still have the burn scars on my fingers where I reached down and touched it (a year ago!)

But the Bodnar controller still works! Simply amazing.

M_Gunz
09-12-2009, 06:30 AM
How did you make out with the LED? Did you use a resistor or a turn-pot?

FlyingColander
09-17-2009, 12:50 PM
Need a little advice from the experienced among us. I have my uni-joint. It comes from a Yamaha and seems the perfect size for a joystick. It fits in the palm of your hand.

I am concerned, however, because it has limited throw. The front and back movement is quite good but the left and right movement is limited. As you can see from the picture the upper yoke does not clear the shoulder of the lower one. Can this be calibrated away or must I get a new joint?

http://i1008.photobucket.com/albums/af203/Photobasin/th_Yamahauni1.jpg (http://s1008.photobucket.com/albums/af203/Photobasin/?action=view&current=Yamahauni1.jpg)

Thanks,

Flying Colander

Sokol__1
09-17-2009, 10:46 PM
You can use a Dremel bit and remove some material inside upper U part, to allow more movement.I make this in one Volkswagen Wheel U-Joint.

But these motorcycle U-join seen have a ~30 degrees of movement in each side, is not adequate?

I like these "spider" with bolts, make fix in some base more ease.

Sokol1.

Fehler
09-18-2009, 01:47 PM
Originally posted by M_Gunz:
How did you make out with the LED? Did you use a resistor or a turn-pot?

Sorry about not answering your questions very fast there, Max...

Actually, the guy at Radio Shack hooked me up with the proper resistors for each color I am using in my pit. So for the DPDT switches, I use LED's and I also have one for my flap control which is on a linear pot. So, at a glance, I can see if I have flaps up or not with a visual clue, and the further I lower the flaps, the brighter the LED gets. Sweet, huh?

Fehler
09-18-2009, 01:57 PM
Originally posted by FlyingColander:
Need a little advice from the experienced among us. I have my uni-joint. It comes from a Yamaha and seems the perfect size for a joystick. It fits in the palm of your hand.

I am concerned, however, because it has limited throw. The front and back movement is quite good but the left and right movement is limited. As you can see from the picture the upper yoke does not clear the shoulder of the lower one. Can this be calibrated away or must I get a new joint?

http://i1008.photobucket.com/albums/af203/Photobasin/th_Yamahauni1.jpg (http://s1008.photobucket.com/albums/af203/Photobasin/?action=view&current=Yamahauni1.jpg)

Thanks,

Flying Colander

And to add, a great deal of it will depend on how long you pan on making your stick. The longer the stick, the less "throw" you will need. From 0-100 front to back mine doesnt move more than 3 inches. But because of the length of the stick, my arm moves about 12-14 inches. So, dont be too concerned about the amount of clearance unless you are considering building a really short throw stick.

Fehler
09-18-2009, 02:08 PM
Originally posted by FlyingColander:
Need a little advice from the experienced among us. I have my uni-joint. It comes from a Yamaha and seems the perfect size for a joystick. It fits in the palm of your hand.

I am concerned, however, because it has limited throw. The front and back movement is quite good but the left and right movement is limited. As you can see from the picture the upper yoke does not clear the shoulder of the lower one. Can this be calibrated away or must I get a new joint?

http://i1008.photobucket.com/albums/af203/Photobasin/th_Yamahauni1.jpg (http://s1008.photobucket.com/albums/af203/Photobasin/?action=view&current=Yamahauni1.jpg)

Thanks,

Flying Colander

One more thing....

The Hall effect sensor is really only good for about 90 degrees of motion. The reason for this is that when it picks up the magnetic field and turns more than 45 degrees (each way which equals 90 degrees) the field readings get a little iffy. But through that range of motion, 45 degrees away from center in both directions, the signal is nearly perfectly linear which is what you really want. That's why it is never recommended to use non-linear tapered pots in any build because the signal strength degrades rapidly at both ends of the pot.

Therefore, if that U-joint you pictured travels at or more than 45 degrees from center, then it will actually travel too much. That's why, in this picture, I had to build a stop ring. That, and it looks good... haha!

In the picture, you will notice that there is more throw to the rear than the front. This was by design too. Not many times do you toss your plane into a nose down configuration, but most of the time you pull back on the stick. So, with the configuration utility provided by Leo diview.exe, I have set mine to be more precise while pulling on the stick than pushing on it. http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

That's why, in a turn fight, I can really attack the edge of departure during a long engagement. Remember, all things equal, the longer the throw, the more precise. So a guy using even a good joystick cannot be as precise because of my longer throw. And that is a huge advantage in a prolonged dogfight...

http://webpages.charter.net/cuda70/KG13-2.jpg

M_Gunz
09-18-2009, 06:23 PM
Originally posted by Fehler:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by M_Gunz:
How did you make out with the LED? Did you use a resistor or a turn-pot?

Sorry about not answering your questions very fast there, Max...

Actually, the guy at Radio Shack hooked me up with the proper resistors for each color I am using in my pit. So for the DPDT switches, I use LED's and I also have one for my flap control which is on a linear pot. So, at a glance, I can see if I have flaps up or not with a visual clue, and the further I lower the flaps, the brighter the LED gets. Sweet, huh? </div></BLOCKQUOTE>

Yes, very sweet! Reason for using turn pots is that new LEDs lose about 10% brightness in the beginning so can be trimmed up.
You have Radio Shack in Germany? At least, I have this idea that you are in Germany.

FlyingColander
09-19-2009, 02:15 AM
Fehler and Sokol,

Thank you both for your replies.

I do plan on making a long stick. I was dreading having to Dremel-away on a uni-joint. I guess I can always do it if I don't have enough throw.

Fehler, is your KG-13 from SchreinerSchmid? I have mine coming next week. He does great work. For those of you unfamiliar with his work check out--
http://www.schreinerschmid.com...ey*=*session*id*val* (http://www.schreinerschmid.com/11501.html?*session*id*key*=*session*id*val*)

julian265
09-20-2009, 06:59 AM
My joint only allows +/-22 degrees movement in each axis, which is absolutely fine.

The things that you would need to grind back to extend the movement could probably be done quickest with a coarse file.

Should you need to cut or grind, don't get metal chips in your bearings - not so much because of wear, but you may cause it to feel rough.

Sokol__1
09-20-2009, 01:31 PM
The Hall effect sensor is really only good for about 90 degrees of motion. The reason for this is that when it picks up the magnetic field and turns more than 45 degrees (each way which equals 90 degrees) the field readings get a little iffy. But through that range of motion, 45 degrees away from center in both directions, the signal is nearly perfectly linear which is what you really want.

http://www.mikesflightdeck.com/images/hslin.jpg

That picture - from Mikesflightdeck.com page - show this. The HALL sensor output is a sine curve, so you need to use area between ~70/110 degrees to have a linear response.

Sokol1

julian265
09-20-2009, 06:06 PM
http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif

http://www.jpfiles.com/hardware/linearity.png

FlyingColander
03-27-2010, 06:58 AM
A follow up question, Julian.

In your How To you mention disassembling the universal joint before drilling to avoid getting metal chips into the bearings.

When I inspected my uni-joint, I found it had no ring clips to set free the cups & bearings. Instead, a round plate sealing everything and preventing dis-assembly. (Actually, your Volvo joint looks sealed too.)

How does one drill through without fouling the bearings?

Can the magnet simply be epoxied to the point where the dowel would penetrated into the bearings? Or will the minor rotation of the magnet throw off the sensor? Didn't you use this technique for your rudder pedals with the magnet moving beneath the Hall sensor?

If I am following the design correctly:

On the Base Yoke the magnet moves and the sensor is stationary.

On the Handle Yoke, the sensor moves in an arc around the magnet?

Or, with the dowels inside the spider-cross, do just the magnets move?

Sorry for all the daft questions.

Cheers,

Flying Colander

julian265
03-28-2010, 06:55 AM
Originally posted by FlyingColander:
A follow up question, Julian.

In your How To you mention disassembling the universal joint before drilling to avoid getting metal chips into the bearings.

When I inspected my uni-joint, I found it had no ring clips to set free the cups & bearings. Instead, a round plate sealing everything and preventing dis-assembly. (Actually, your Volvo joint looks sealed too.)

How does one drill through without fouling the bearings? The spring clips were just there as a safety - the bearings are actually press-fitted into the yokes, and are not easy to get out (but not too hard either...) You need to use a fair bit of force to do it, a vice, and possibly a hammer (or a mechanic!). I'll find or write up a method of doing it, and post it when I have. I got the method from the mechanic's manuals for my car - though most DIY maintenance manuals for rear-wheel-drive cars should cover uni-joints too.

Secondly, and importantly - I wouldn't bother trying to drill the bearing cups - they will be hardened. Instead, use a dremel tool or other small grinder to grind through the centre of the cup, just big enough for a dowel to fit through. See page 8 of the pdf - you can grind a small hole with a small bit.


Can the magnet simply be epoxied to the point where the dowel would penetrated into the bearings? Or will the minor rotation of the magnet throw off the sensor? Didn't you use this technique for your rudder pedals with the magnet moving beneath the Hall sensor? I might not be understanding correctly, but it sounds impossible - as the bearing cups are press fitted into the yokes, and hence the base yoke bearings don't move at all, and the handle yoke bearings move in a way which is unsuited to this method of sensing (linear, along a curved path). For both axes, you need to compare the cross-piece's angle with the yoke's angle, and align the sensor and the magnet on this axis of rotation.


If I am following the design correctly:

On the Base Yoke the magnet moves and the sensor is stationary. Yes. The magnet rotates (with the cross-piece), and the sensor is mounted to the base yoke (hence mounting flange). It doesn't matter which of the sensor or magnet you mount to what part though - as long as one is on the cross-piece, and the other on a yoke/bearing cup, you will measure the angle between them.


On the Handle Yoke, the sensor moves in an arc around the magnet? No, for both axes (and the pedals) the sensor is 'above' the magnet, and rotate only. Compare these images by flicking back and forth between them:
http://www.jpfiles.com/hardware/pitch_back.jpg
http://www.jpfiles.com/hardware/pitch_forward.jpg

This is the handle yoke, which does my pitch axis. If I only move the stick along the pitch axis, the magnet is stationary, and the sensor rotates around the axis of the handle yoke's bearings. If I move the stick along the roll axis, both the sensor and magnet move in an arc together (as they're all mounted on the cross-piece), and hence remain aligned.


Sorry for all the daft questions.

Cheers,

Flying Colander No worries, keep asking if something is not clear. I reckon this stick is worth the effort to build! Once you've done the disassembly, grinding and reassembly it gets easier.

julian265
03-28-2010, 07:18 AM
I used these instructions for disassembling the uni joint.
From:
http://www.jpfiles.com/hardware/phone0170.jpg

http://www.jpfiles.com/hardware/phone0172.jpg - "universal joint, replacement" section
http://www.jpfiles.com/hardware/phone0173.jpg
In the second picture in the above link, you're meant to push downward on the prop shaft, when the hammer handle is in place, which will push the cross-piece and bearing cup upward. It might help to increase your leverage by placing some larger or smaller tube over or into the prop shaft.

http://www.jpfiles.com/hardware/phone0174.jpg

Also, use heavy grease to hold the bearing rollers in place when you're reassembling - if they don't stand up, you probably won't get the bearings back together.

If you can't read these pictures I can scan the pages tomorrow.

Be careful and you might avoid scattering the little bearing rollers around like I did... It's no big deal if you do lose them though, you've just got to buy a new uni joint kit, which usually isn't much $, and you're guaranteed not to have any slack.

FlyingColander
03-28-2010, 10:00 AM
Julian265,

Thanks for the time and extra work spelling it out slowly for me. Thought I'd better ask before starting to grind away. I think I'm on the right track now. By the way, I'm not the only one singing your praises. Your good work has led to another great project:

http://simhq.com/forum/ubbthre...rip.html#Post2958931 (http://simhq.com/forum/ubbthreads.php/topics/2958931/Uni_joint_stick_DIY_hand_grip.html#Post2958931)

Cheers,

Flying Colander

FlyingColander
03-28-2010, 10:45 AM
Here is an excellent site detailing the removal cups and bearings from uni-joint yokes. I found it useful as I couldn't visualize many of the parts under discussion.

http://myural.com/driveshaft%2...20joint%20repair.htm (http://myural.com/driveshaft%20&%20universal%20joint%20repair.htm)

Hope this helps someone else too.

Flying Colander

Sokol__1
03-28-2010, 02:12 PM
Since I don't have tools to disassembly U-Joint, I try in this way:

http://img413.imageshack.us/img413/7623/cardsr.jpg

Instead wood pins I use aluminum strip (in red in Sketchup drawing) to support magnets.

Glue these strip's with epoxi putty over + of U-Joint.

Work in the same way.

Sokol1

FlyingColander
03-28-2010, 02:20 PM
Well done, Sokol! Very clever indeed. Keep up the good work.

Flying Colander

julian265
03-28-2010, 05:34 PM
Originally posted by FlyingColander:
Julian265,

Thanks for the time and extra work spelling it out slowly for me. Thought I'd better ask before starting to grind away. I think I'm on the right track now. By the way, I'm not the only one singing your praises. Your good work has led to another great project:

http://simhq.com/forum/ubbthre...rip.html#Post2958931 (http://simhq.com/forum/ubbthreads.php/topics/2958931/Uni_joint_stick_DIY_hand_grip.html#Post2958931)

Cheers,

Flying Colander
http://forums.ubi.com/groupee_common/emoticons/icon_eek.gif That setup is awesome! Well finished too.

julian265
03-28-2010, 05:36 PM
Originally posted by Sokol__1:
Since I don't have tools to disassembly U-Joint, I try in this way:

http://img413.imageshack.us/img413/7623/cardsr.jpg

Instead wood pins I use aluminum strip (in red in Sketchup drawing) to support magnets.

Glue these strip's with epoxi putty over + of U-Joint.

Work in the same way.

Sokol1
Nice idea. Did you have to restrict the stick's movement to avoid squashing the magnet support strips? Or is your uni joint large enough not to have to worry about it?

Sokol__1
03-29-2010, 09:05 AM
The U-Joint that I get - from VW Beetle steering column, I think - have free space over +, so I don't need limit movement of magnets.

Sokol1

julian265
03-29-2010, 04:22 PM
Nice! That uni probably doesn't have a hollow cross-piece either, being for steering.

ibeagle
04-02-2010, 01:21 AM
'scuse me if this has been covered before, but if I was to build a row of levers with these hall sensors, how much separation between the magnets is needed?

julian265
04-04-2010, 05:36 PM
I would keep the magnets stationary, and use two sensors for every one magnet (one sensor on either side). That way you use less magnets, which saves space, and the multiple magnetic fields will all be aligned, so you can squash each axis in as close as you'd want - as long as the sensors aren't too sensitive to angle. In other words, the closer the sensor is to the magnet, the more the output voltage will change for a given lever angular movement. If they're too sensitive, they'll read 0% or 100% before the lever is at real 0% or 100% position - which you can't tune out in software. Because of this, you'll have to find the distance from sensor to magnet which gives you the desired sensitivity, and use this distance for all sensors.

JFA2
04-07-2010, 01:52 PM
Hi Julian, just dropping by to say thank you! for the unijoint pdf you made. It was a big help when building my own project. There wasnīt any need for me to build a new joystick but this unijoint concept was just too brilliant idea not to use http://forums.ubi.com/images/smilies/clap.gif

Cheers,

Sokol__1
04-07-2010, 04:11 PM
For us that dont read SimHQ-Pit Builders, JFA2 Uni-Joint stick:

http://i62.photobucket.com/alb...7/JAF--/IMG_0007.jpg (http://i62.photobucket.com/albums/h117/JAF--/IMG_0007.jpg)
http://simhq.com/forum/ubbthre...opics/2958931/1.html (http://simhq.com/forum/ubbthreads.php/topics/2958931/1.html)

Well done!

Sokol1

julian265
04-07-2010, 06:25 PM
You're welcome JFA!

I found out later that someone by the handle of Hands9 (IIRC!) of frugalsworld (RIP) had already made a uni-stick, but it's good to see the pdf and sensor arrangement doing a good job!

JFA2
04-08-2010, 10:01 AM
Sometimes itīs hard to say who was the first with some invention. IMO itīs not so important either. Great thing about builders community is that you get one idea here and another there and finally you make something new of it. Itīs somekind of morphinghttp://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

I really respect guys who are willing to share ideas with other builders. Simple looking things allways requires lots of trial and error and making instructions is somewhat hard to do. For me this unijoint+halleffect sensor setup is as solid idea as baloo pedals (thx for that link Sokol!). Itīs priceless -literally. Thereīs no place where you can buy flying equipment of this quality at least not with reasonably amount of money. I have tried few commercial setups during the years and i donīt look backhttp://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

kepalax
07-12-2011, 06:31 AM
Hi, and sorry for barging in without proper introductions (I'll get to appropriate threads later). I have been reading about the hall-effect sensor method, but haven't yet come to anything about the other casing type (SOT23W). Does anyone have any experience about the orientation of the sensor element in these? I mean, can it be used correctly installed and lined instead of SIP-casing sensor.

Thanks in advance if someone can enlight me on this matter

Mikko

mattinen
07-12-2011, 11:23 AM
That's the biggest stick I've ever seen. There is perhaps something Freudian about it...

julian265
07-13-2011, 12:53 AM
Originally posted by kepalax:
Hi, and sorry for barging in without proper introductions (I'll get to appropriate threads later). I have been reading about the hall-effect sensor method, but haven't yet come to anything about the other casing type (SOT23W). Does anyone have any experience about the orientation of the sensor element in these? I mean, can it be used correctly installed and lined instead of SIP-casing sensor.

Thanks in advance if someone can enlight me on this matter

Mikko

They'll be designed to work identically AFAIK, so I can't see any problems with using other types - it's just that through-holes are easier for people like me to work with, and the bendiness of the pins allows quick adjustment.

kepalax
07-14-2011, 09:52 AM
Originally posted by julian265:
They'll be designed to work identically AFAIK, so I can't see any problems with using other types - it's just that through-holes are easier for people like me to work with, and the bendiness of the pins allows quick adjustment.

thanks, I actually found this picture from some Allegro pdf

http://i53.tinypic.com/r1d6qw.jpg

have to take a bit different approach to the structure, but it might be ok