Are all halls effect joysticks the same?

[Lilwolf]

Anyone know exactly what the output from a halls effect joystick is?  I know that they use magnets to create the analog values... but does it end up with a variable resistence similar to a potentiameter  (but all 4 directions being seperate)?

I'm considering hacking the analog halls effect joystick in a dreamcast controller... I would like to keep it analog if possible.

whats the chance that I could replace an arcade halls effect joystick directly into a hacked dreamcast controller?

[Dave_K.]

whats the chance that I could replace an arcade halls effect joystick directly into a hacked dreamcast controller?

Should work, if you have a analog or hall effect arcade joystick.  Problem being these are prohibitively expensive controllers...but if you got one, go for it.

[u_rebelscum]

Anyone know exactly what the output from a halls effect joystick is?  I know that they use magnets to create the analog values... but does it end up with a variable resistence similar to a potentiameter  (but all 4 directions being seperate)?

I'm considering hacking the analog halls effect joystick in a dreamcast controller... I would like to keep it analog if possible.

whats the chance that I could replace an arcade halls effect joystick directly into a hacked dreamcast controller?

The Hall Effect output is a variable resistence just like a POT, and the two can be interchanged (but for joysticks, only if the next sentence is true).

AFAIK, the left/right pair resistances are wired in series (same with the up/down pair), so the controller has two axes outputs just like a POTs analog stick.  And IIRC some of the third party DC controllers used POTs instead of hall effect for the analog stick, which would support this.

I don't know what the resistance range of the DC controller hall effect is, but a multi-meter can help you.  If you can find pots with the same resistances, simple.

You could hook an arcade halls effect joystick to a hacked DC controller if the resistance ranges are the same, and if both have the left/right & up/down paired.  (Hmm, I guess if both not paired would be okay, too .)

[Lilwolf]

I thought the difference was that there are 4 values coming away from a halls effect joystick (some range away from each... )

So going from far left -> far right would look like this
(left) left 2k right 0, (middle) left 0 right 0, (right) left 0, right 2k

Instead of pots.

(left) 0, (middle) 1k, (right) 2k,.....

Am I wrong with this?  

Last... the hall effect sensors... can you replace them easily?  So I could get a halls effect joystick and try to find the right sensor to make it work with the dreamcast?

Thanks

[u_rebelscum]

I thought the difference was that there are 4 values coming away from a halls effect joystick (some range away from each... )

So going from far left -> far right would look like this
(left) left 2k right 0, (middle) left 0 right 0, (right) left 0, right 2k

Instead of pots.

(left) 0, (middle) 1k, (right) 2k,.....

Am I wrong with this?  

Last... the hall effect sensors... can you replace them easily?  So I could get a halls effect joystick and try to find the right sensor to make it work with the dreamcast?

Thanks

This site says the Atari I, Robot Hall Effect joystick had two outputs, one for each axis.  
CH Products hall effect joysticks seem to have one signal per axis (0 - 5V, 2.5V center).
This pdf from a hall effect joystick manufacturer shows one analog output per axis (with the option of digitial outputs, too).  (Note that each of the diagrams is for only one axis and has "outputs X 2" below.)


I'm not as sure on the DC controller hall effect output, since (according to howitworks.com) there are four sensors, but I was under the impression that the four sensors worked in pairs, much like a single axis, dual pedal, POT based pedal set.  

To follow your example:
DC controller
(left) [left 0 + right 0] = 0k, (middle) [left 1k + right 0] = 1k, (right) [left 1k + right 1k] = 2k.
1 axis, 2 pedal (or 1 axis, 2 trigger button)
(gas) [gas 0 + brake 0] = 0k, (middle) [gas 1k + brake 0] = 1k, (brake) [gas 1k + brake 1k] = 2k.

With the DC/computer only seeing the total.

However, if the DC controller sends each part rather that the total (much like two, 2 axis, 2 pedal pedal sets), then you run into what you discribe, and would have some problems hooking the original arcade hall effect sticks since they only sent 2 signals instead of four.

OTOH, the DC controllers only have 6 output pins to the DC, so the pcb in the controller must be doing somthing.  And if you use the controller's analog triggers, you will notice at least they are in serial, like a 1 axis, 2 pedal.  Only real way is to test it, I guess.


From the quick skims I did, looks like hall effect sensors run the price gamut, much like POTs.  But, if you can find the hall effect sensors with the correct range, and if the DC controller serials the left/right (& up/down) (which means the sensors you'd want would be two times the range of the DC's), then you could hook up easily.

[jerryjanis]

I sacrificed a DC controller today...  Here's what I found out:

First, some pictures:


1) This is the controller I hacked.



2) DC Controller Doodad turned to the left  (Hall effect, or just a pot?)



3) Turned right



3) Trigger, conveniently labelled.  It's a 5kohm doodad.  I own a PC gameport joypad with an analog thumbstick that uses something that is this same style.  There are 3 input/output lines, and I get the 0..5kohm range when I take a multimeter reading.



4) Inside of trigger.  It's a pot as far as I'm concerned (but I was never much of one for exact wording).  The trigger slides up and down the metal contact line that you can see - it makes contact with two lines, the middle line and the far right line.



5) The circuitboard.  Ignore the wire I colored yellow - that is me working around some torn-out circuit contact material. The two blue circles mark where the x and y axes are soldered onto the circuit board - each of the 2 hall effect (I wanna call them pots sooo bad) have 3 contacts, just like a pot.  I get a range of 0(left)..5(center)..10kohm(right), making them the equivilent of a standard 10kohm pot.  The four red circles mark where four posts (see the next picture) are soldered from the entire cube that makes the thumbstick.  Each of the four posts is attached to the same piece of conductor which makes up the shield around the whole 1/2 inch^3 thumbstick box.  The 4 posts connect with a very small grid that spans most of the whole circuitboard, and pretty much contacts one side of each button.  I assume that I would call it the ground, but I'm not quite sure what that implies.  After removing the thumbstick box, I soldered 3 wires to where the posts were to connect all of them back together again. 


6) The underside of the thumbstick unit

I could be wrong, but from what I could tell, the controller seemed really sensitive to change - if EVERYTHING wasn't hooked up (including the trigger, the 4 metal posts, and the x and y axes), you can't expect anything to work properly.  It's not quite as easy to work with as the Dual Strike hack in this sense. 

I yanked out the 10kohm pots from the controller, and tried a few different combinations of my own.  I popped one of the crappy Dual Strike 20k pots in, and it did what I would expect.  The effective range was the first half of pot (the first 10k).  Turning the pot to the left resulted in the game's character moving left, 1/4 turn->center, 1/2 turn->right, 3/4 turn->right, full turn -> right.

I just bought a bunch of 5k pots.  They are behave strangely.  (0k left, 4k center, 5k right).  Can you guys tell me anything about that?

Anyway, I wired up some of those to the DC controller.  I don't know why, but they actually worked as far as I could tell what with the strange ohmage of these pots in particular.  I could go right and left, but when I max out the pot in either direction, the character on screen is going 1/2 speed.  I would have expected for the character to move in the following manner: move pot to the left->move character fast left, center pot->slow left, right  pot->stop.  I dunno , it didn't seem to do that.

So my conclusion would be that an arcade joystick with 10kohm pots could work (if there was such a thing).   One possible problem might be that there's no calibration for the DC, so if the center point was a little bit off, it would really suck - you need the joystick to self-center perfectly or you're in trouble.  Darnit, I think that the Nintendo 64 calibrates itself to the center point when you turn it on.  You can confuse the n64 by pushing the joystick to the left on startup.  Old-school PC joysticks used to have mechanical calibrators on the joystick itself.  Something like that would be perfect.

[u_rebelscum]

Thanks for the info!

But, AFA recalibration, the DC docs say it recalibrates on startup.  And if I startup with the analog stick off center, it's f'ed up until I reset the DC, just like you discribe the N64.  Seems to recal to me.

I wonder if the above is why the POTs you tested acted like the did, especially if DC "knows" what voltages the ends will be.


I dug up a book with a little bit on hall effect and it says, if you want to vary the voltage, a hall effect resistor is a non-mechanical POT.  I little late, but... [shrug]

[jerryjanis]

But, AFA recalibration, the DC docs say it recalibrates on startup.  And if I startup with the analog stick off center, it's f'ed up until I reset the DC, just like you discribe the N64.  Seems to recal to me.

So it does...  My wiring is a little flaky so it's hard to be sure of anything.  Hopefully I can tidy it up and play around some more.  But actually I think it's all figured out.

I wonder if the above is why the POTs you tested acted like the did, especially if DC "knows" what voltages the ends will be.

Yes!  That would explain it.

Not sure what this means, but the Happs analog joysticks mention an "Optional pot. amplifier PC board allows variable output from potentiometer" as one of the features.  I wonder if that could help.  I wonder what that is.  I guess it doesn't matter because I doubt it's worth the Happs $250 price tag to anybody.

[Lilwolf]

I don't believe there is any way to amplify a resistor... you can step it up... but not amplify (value*2)...

But you could add a 5k resister to the setup... and it might make it so it will work... but harder to go slow...  But if this is the case... you might as well do the 8way hack.

Is it true that you wont find any arcade type 10k pots anywhere?  It would be sweet to get a few analog joysticks running.

[jerryjanis]

Is it true that you wont find any arcade type 10k pots anywhere?  It would be sweet to get a few analog joysticks running.

I don't know there's gotta be some somewhere, doesn't there?  I looked around and all I could find was monitor knobs and audio knobs - nothing arcadey at all.

[Tailgunner]

I just bought a bunch of 5k pots.  They are behave strangely.  (0k left, 4k center, 5k right).  Can you guys tell me anything about that?

Sounds like they're audio taper pots. You want linear taper pots for a project like this.

[CanStick]

I just bought a bunch of 5k pots.  They are behave strangely.  (0k left, 4k center, 5k right).  Can you guys tell me anything about that?

Yes, you have bought some 5K log taper (aka audio taper) pots, instead of linear taper pots.

Log taper pots are designed to be used in audio circuits, because our hearing behaves more in a logarithmic fashion than a linear fashion...
I hope I didn't get too techno here...  

Hope this helps you a bit...

(added section)
Haha Tailgunner, sounds like we both saw this at the same time!!!  

[Tailgunner]