Closer To The Final Replacement Of Stick

[DaveC]

Dissected it...
Enough for me of it... I will now focus solely on making replacement. This joystick is weird with its lubricant, not equal pads, diagonal bias and who knows what else! 

​

Wow you completely hacked up the stick. Did you see the original pics that I did in the other thread? It already had a complete disection, and showed the construction and diagonal bias.

This one:

I didn't open the stick on the mainboard, I just ordered a replacement stick assembly from Alps and dissected that without touching the GP2X itself. You could have saved yourself some time doing that.

 

[Radek]

Looks like it is an insulator but why?!

Actually I had to provide voltage for both sides seperately because of it. Bizarre?

Enough for me of it... I will now focus solely on making replacement. This joystick is weird with its lubricant, not equal pads, diagonal bias and who knows what else!

​

I had forgotten about that myself, I don't understand why they did that, but there must be a reason to complicate the design like that.

Go here, ALPS web catalogue and enter RKJXL into the search box and you'll be able to find a PDF which will show you the exact details of the connections.

Was there. Looked at that PDF and still didn't get any idea...

I would have linked directly, but it uses session-based URLs, which will have timed out by them time you're reading this.

My last experiment on the joystick didn't turn out a success, so I'm going to give it one more try with the second one I got, but since my thoughts the other evening, I'm leaning towards a joystick replacement now, rather than a d-pad.

I'm considering both but imho the soldering on the GP2x's mainboard should be avoided. I couldn't get any decent joystick replacement and it is actually far easier for me to assembly my own dpad than joystick. So given the parts I could get I'm doing dpad now.

I think you're on the right track with your ideas, but rather than duplicate the effort, I think I'll concentrate on a microswitched alternative, and between us we should be able to come up with some useful options.

There is no duplication just more wide experimentation. There are still lots of work to do and different people might prefer different options.

I just got N-Gage classic pads and they are... perfect fit for GP2x!

And a bit closer:

Not only size is just right but looking at the height of that pad it should fit without removing bottom part of GP2x's stick! That means (if it will show correct) that it could act like a socket for replacement dpad and I don't have to tell you how mechanically it is making everything simpler.

So my plan is:
- drill gp2x's case
- make appropriate pcb for that dpad with microswitches
- use polymorph plastic to cast all needed supporting parts
- put ready unit into "socket" (the bottom part of stock stick)
- connect wires to my logic adapter
- enjoy!!!!
- try to combine converter PCB with dpad PCB into one (to reduce installation troubles)
- save the word...

BTW, what camera are you using to take your pictures? They're very clear, especially compared to some of the rubbish macro shots my cheap camera takes.

​

It's Olympus "C-740 UltraZoom". All pictures were made using flash lamp then downscaled using Gimp's bicubic filter to appropriate sizes. No special settings just C-740 automatic presets. The camera is only 3.2 megapixels and isn't produced now probably. Wasn't expensive either.

 

[Radek]

Dissected it...
Enough for me of it... I will now focus solely on making replacement. This joystick is weird with its lubricant, not equal pads, diagonal bias and who knows what else!

​

Wow you completely hacked up the stick. Did you see the original pics that I did in the other thread? It already had a complete disection, and showed the construction and diagonal bias.

Yes - I was completely aware of your work with the stick. But because my goal is to replace it fully I hadn't any hesitation to hack it or even destroy it. After all I can still use my GP2x even without stick.

And I can even replace it with new but I see no reason why go back to it.

Then I was interested in making some additional electric measurements. So I had reasons to hack it myself.

 

[bacteria]

Radek - excellent work. We are all awaiting with baited breath!

You will need to produce "idiot easy" kits for people to upgrade their units, with good instructions.

Just a thought - if you can connect directly to the joystick pins, why not just wire the suitable connections directly to a D-pad or joystick, and therefore not need the extra electronics? Surely, if we connect "up" on a joystick we have to "up" joints on the GP2x's joystick, that will work?? Please explain why this won't work (incase I try it and something goes wrong)!

 

[DijiTao]

Radek - excellent work. We are all awaiting with baited breath!

You will need to produce "idiot easy" kits for people to upgrade their units, with good instructions.

Just a thought - if you can connect directly to the joystick pins, why not just wire the suitable connections directly to a D-pad or joystick, and therefore not need the extra electronics? Surely, if we connect "up" on a joystick we have to "up" joints on the GP2x's joystick, that will work?? Please explain why this won't work (incase I try it and something goes wrong)!

​

Ultra-Simplified Answer:
To handle the diagonals. On a d-pad (4 switch) up-left is what you get when you push up and left, on the gp2x abomination (6 switch) up-left is it’s own switch and not the combination of up and left.

 

[Radek]

Radek - excellent work. We are all awaiting with baited breath!

You will need to produce "idiot easy" kits for people to upgrade their units, with good instructions.

That's my very intention.
(and I have a method what will not need any soldering to interface with GP2x's mainboard)

Just a thought - if you can connect directly to the joystick pins, why not just wire the suitable connections directly to a D-pad or joystick, and therefore not need the extra electronics? Surely, if we connect "up" on a joystick we have to "up" joints on the GP2x's joystick, that will work?? Please explain why this won't work (incase I try it and something goes wrong)!

​

Because of software compatibility - there are 8 switches in the GP2x's stick. Not 4 like in typical d-pad or digital joystick but 8. Applications are expecting to get 8 discrete signals (on/off states) not 4 so by connecting just normad d-pad you will lost half of possible directions (usually diagonals).

However using combis like up+right 4-switch based controller can give you diagonals but software will not recognize them as such. Here comes my logic circuit - it translate signals so programs are getting values what they are expecting.

So if all software would be recoded to expect 4-switch dpad/joystick then any translation wouldn't be needed. But it's very unrealistic proposition...

BTW - how do you like GP2x like this?

 

[Khatoblepas]

I think it looks very neat and usable. I suppose it would work like a real D-Pad now ^^. With awesome controls. Though, the silver does kinda clash with the rest of the unit. Black with orange arrows would look really sleek.

But this is majorly cool. Have you tested any games with it yet?

 

[scorpio]

Was there. Looked at that PDF and still didn't get any idea...

For some reason they've made the top part of the casing in two parts, as in your photo. When the unit is assembled, the bottom half is the voltage source for the up, up-left, left, and down-left directions, and the top half is the voltage source for the up-right, right, down-right and down directions. The tricky thing is, when you are crossing between down-left and down, both the top and bottom halves are now shorted together! The same happens when the stick is pushed between up and up-right. It makes no sense to me.

I'm considering both but imho the soldering on the GP2x's mainboard should be avoided. I couldn't get any decent joystick replacement and it is actually far easier for me to assembly my own dpad than joystick. So given the parts I could get I'm doing dpad now.

From having seen that the legs of the joystick assembly have been left protruding out the bottom, and you can wire-wrap connections onto it, I can see your point. It would certainly be the safest option.

I just got N-Gage classic pads and they are... perfect fit for GP2x!

I'm quite impressed with how well it fits! Obviously, the colour is going to be a big point for a lot of people, but I'm sure that can be sorted out as a separate issue.

It's Olympus "C-740 UltraZoom".

Sounds like it is more suited to the close-up shots then. Look forward to seeing more photos.

I tried taking some pictures of my one today, to try and document the internals a bit better, as there seems to be nothing on the wiki yet with good pictures of the circuitry, and I think we could all do with something to look at and draw on for ideas. Needless to say, they didn't turn out particularly well.

 

[Series-8]

I don't think the silver black clashes so bad...

Also how does the Ngage Dpad itself actually handle?

Looks kinda large and strangely shaped for a dpad, but it will shure beat a stick senseless.

It actually looks kinda similar to a stick assembly I've seen earlier... (a self-contained one...which was discounted as it ironically only contained 4 switches...abit moot now)
:blink:

 

[Radek]

I don't think the silver black clashes so bad...

I liked it that way too!

Also how does the Ngage Dpad itself actually handle?

Actually sorry for teasing you... I haven't yet a PCB with microswitches for it. It was just drilling GP2x's front panel and I succeeded at it. Wasn't that hard but drilled a small bit too much... However such mistakes are expected at this stage. I'm happy the case didn't crack.

I have to say it's doable for everyone if someone isn't too brutal with a chisel.

Looks kinda large and strangely shaped for a dpad, but it will shure beat a stick senseless.

It has nice in touch and feels solid. Should allow precise control and is inexpensive and easy to get. It's very important imho to use parts what aren't exotic.

It actually looks kinda similar to a stick assembly I've seen earlier... (a self-contained one...which was discounted as it ironically only contained 4 switches...abit moot now)
:blink:

​

Ugh... The ones what I could get aren't so "self-contained" - I will have to make supporting PCB for them. Will not be hard but I need some parts to move this project further. In week I should have fully working unit and I'm still loking for other canditates (the one from N-Gage QD is my prime target).
(thought the seller told me that they have switches (had probaly something different in a mind!?)...)

Preferably I should cannibalize something for quick prototype but I haven't anything worth sacrificing. Of course I'd rather not destroy a new, possibly expensive gadget just for that.

In next week I will go shipping for various parts (microswitches, cmos chips, pinheads, connectors, cables and perhaps few other thing too). I should also get some other helpfull things... The project is 70-80 percent done and the worst is behind.

I want it done as fast possible because it's already taking too much of my time! (not that I'm not enjoying it but I had to freeze other project). But enought of for now - tiime for bed!

 

[DaveC]

I want it done as fast possible because it's already taking too much of my time! (not that I'm not enjoying it but I had to freeze other project). But enought of for now - tiime for bed!

​

I know how it is. Stuff like this is ALOT of work. I work on different hardware projects time to time (not all related to gaming) . It is ALWAYS more work and takes more time than you think in your mind originally. I had some ideas for improving the stick too but never started them (other than the cap). I guess I don't hate the current stick enough to motivate me being the lazy fuck that I am.

I see that you only have just begun really. Wait until you start trying to mount the pad assembly in the case requiring lots of futzing around

 

[second exodous]

Wow, this is impressive. Maybe this will prompt GPH to do a slight revision on future batches of the GP2X? Somewhere on these forms I heard someone using the GP2X upside down and using the buttons for a d-pad and was comfortable. Maybe GPH should just do that, I don't think it would be that hard to change the PCB to buttons on each side and eliminate the stick, might even make it cheaper to produce.

Thanx,
Stan

 

[airwolf]

Hey Radek, thanks for all you are doing.

Something I just noticed was something lately that I have not been using because of DrMDx.....it is the use of my in-perfect-condition Sega NOMAD.

That system has the perfect pad. It feels like it is 8-way. The pad pivots inside a square plastic inner housing that rests inside another black circular fixed sphere on the Nomad motherboard.

I would like to take pics, but there is 1 reverse allen screw that is keeping me from taking a much better peak. I think if this part could be found it might fit inside of the Gp2X and give us all the dream machine we keep wanting this unit to be. As soon as I take this baby completely apart (these things are very rare), I will post pics.

Can you find out more on this part? Maybe it is still being made or we could find some of them? Thanks.

 

[Radek]

I want it done as fast possible because it's already taking too much of my time! (not that I'm not enjoying it but I had to freeze other project). But enought of for now - tiime for bed!

​

I know how it is. Stuff like this is ALOT of work. I work on different hardware projects time to time (not all related to gaming) . It is ALWAYS more work and takes more time than you think in your mind originally. I had some ideas for improving the stick too but never started them (other than the cap). I guess I don't hate the current stick enough to motivate me being the lazy fuck that I am.

I see that you only have just begun really. Wait until you start trying to mount the pad assembly in the case requiring lots of futzing around

​

Mechanically yes but I have good idea to make it easier than usual. However there are always small unexpected bits what shows and need fixing...
(and they can take most of the time)

So... I agree...
(but I'm still trying to get another controller, perhaps next time it will be more complete?)

But once I will get more parts (should be soon) then things will move again.

Wow, this is impressive. Maybe this will prompt GPH to do a slight revision on future batches of the GP2X? Somewhere on these forms I heard someone using the GP2X upside down and using the buttons for a d-pad and was comfortable. Maybe GPH should just do that, I don't think it would be that hard to change the PCB to buttons on each side and eliminate the stick, might even make it cheaper to produce.

​

No. They should change the stick to another but with 4-switches. Eventually add logic circuit to retain software compatibility (with SMD it'd be really small).

Hey Radek, thanks for all you are doing.

Something I just noticed was something lately that I have not been using because of DrMDx.....it is the use of my in-perfect-condition Sega NOMAD.

That system has the perfect pad. It feels like it is 8-way. The pad pivots inside a square plastic inner housing that rests inside another black circular fixed sphere on the Nomad motherboard.

I would like to take pics, but there is 1 reverse allen screw that is keeping me from taking a much better peak. I think if this part could be found it might fit inside of the Gp2X and give us all the dream machine we keep wanting this unit to be. As soon as I take this baby completely apart (these things are very rare), I will post pics.

Can you find out more on this part? Maybe it is still being made or we could find some of them? Thanks.

​

I will try... (I'm already)

It looks nice indeed.

 

[Radek]

Another update...

Was looking around and found my old keyboard. It was notebook like one and disassembled it and found some rubber/membrame pads switches inside. Immediately tried to fit them to the N-Gage pad:

Fitted very well (it's looking like a spaceships of some sort now...):

Testing with Gp2x's button's contact pads:

Is it working? it is! (images are fuzzy because I turned flash off because lcd reflects it too much)

These rubbers don't conduct electricty themselfes so extra pads are needed for that. (or just glue some conductor inside them). I was using conducting pads from keyboard what I had disasembled. The feel of them is quite good, perhaps too step and clicky (the rubbers are from keyboard after all). But they worked and should give very good grip on directions because of their "clicking like behavior".

If not that scary hole what I have done (actually little too big) in the GP2x's front panel I could have working pad even today. Of course I wasn't planning anything with these rubbers because I just found them accidentally but they look like a good alternative to microswitches (will get them too).

 

[Ravnos]

8======D~~~~~

Good work Radek!

 

[second exodous]

Wow, this is impressive. Maybe this will prompt GPH to do a slight revision on future batches of the GP2X? Somewhere on these forms I heard someone using the GP2X upside down and using the buttons for a d-pad and was comfortable. Maybe GPH should just do that, I don't think it would be that hard to change the PCB to buttons on each side and eliminate the stick, might even make it cheaper to produce.

​

No. They should change the stick to another but with 4-switches. Eventually add logic circuit to retain software compatibility (with SMD it'd be really small).

What I meant was something like this....

I didn't mean buttons necessarily, but the left side could have the exact button configuration as the right. 4 buttons=4 directions, start/select=volume+/volume-. You did exactly what I was thinking about. Or maybe they could do what PS2 or Xbox does, have the dpad up top then right below it have a stick.

As for only 4 directions, isn't two directions being pushed at once for a diagonal just as good as a diagonal contact? How is the PS2 controller set up? I know NES, SNES, Genesis, and GBA just have the 4 directions and you have to push two at once to get a diagonal. Why does everyone think they need diagonal contacts?

Thanx,
Stan

 

[Radek]

Wow, this is impressive. Maybe this will prompt GPH to do a slight revision on future batches of the GP2X? Somewhere on these forms I heard someone using the GP2X upside down and using the buttons for a d-pad and was comfortable. Maybe GPH should just do that, I don't think it would be that hard to change the PCB to buttons on each side and eliminate the stick, might even make it cheaper to produce.

​

No. They should change the stick to another but with 4-switches. Eventually add logic circuit to retain software compatibility (with SMD it'd be really small).

What I meant was something like this....

I didn't mean buttons necessarily, but the left side could have the exact button configuration as the right. 4 buttons=4 directions, start/select=volume+/volume-. You did exactly what I was thinking about. Or maybe they could do what PS2 or Xbox does, have the dpad up top then right below it have a stick.

As for only 4 directions, isn't two directions being pushed at once for a diagonal just as good as a diagonal contact? How is the PS2 controller set up? I know NES, SNES, Genesis, and GBA just have the 4 directions and you have to push two at once to get a diagonal. Why does everyone think they need diagonal contacts?

​

You need more than 4 switches if you want more than 8 possible directions. It's nothing wrong to have say 8 switches like in the gp2x's stick. If they would be all of the same size including the intermediate diagonals then we would have something decent. But stock stick is... weird. It has big diagonal bias, almost nonexistent intermediate diagonals and its mechanism doesn't give you any feeling what direction you are actually pushing.
(for me it feels more like an analog joystick than digital)

With just 4 switches you are forced mechanically to 8 directions (4 ordinals plus 4 diagonals). And most of games are expecting exactly that so it is most reliable (and playable) solution.

There are some arcade games what are using multiswitch joysticks actually...
So there also games what are demanding mouse to be playable or analog joystick.
What is best is very arguable because it depends strongly on type of game.
(but 4-switch based controllers are most common)

Then why GP2x has such weird stick?

It's common and it is cheap. Someone who was designing GP2x when searching for parts just selected it (it looks good on paper for sure) and that was so. There were other things to worry about for GPH so once they decided for it then didn't reconsider later.

 

[Flexinoodle]

Even more great work Radek, Can't wait to see the finished unit

 

[saboteur]

Radek - yet more sterling work - once this is completed I think u should be elevated to 'godhood' status.

Probably a stoopid question, but could u not stick 8 'pads' under the dpad and attach them to the existing GP2X control points and not use your circuit ?.

I'll get me coat.