Joystick 45 Degree Rotation ?


Hi all,

So ... turning the controls 45 degrees and changing the bias could be done in one go on old units through a software/firmware updates?.
Is that what they did for MK2?,
or did they physically turn the piece? leaving the software untouched?.

Cheers,
Diego,
 
Radek (or DaveC if you're reading this), how hard is it to open the stick and put it back together without damage? It looks like adjusting the length of the contacts wouldn't be too hard once it was opened, but I think I remember hearing that putting it back together tended to mess the stick up.
 
FluffyPanda posted on May 4 2006 at 11:55 AM said:
Radek (or DaveC if you're reading this), how hard is it to open the stick and put it back together without damage? It looks like adjusting the length of the contacts wouldn't be too hard once it was opened, but I think I remember hearing that putting it back together tended to mess the stick up.

Opening it needs cutting off of two pins. After that it's easy however I couldn't get it to work as well as before. Inside is some kind of lubricant and because of its loss it was so?
(stick needed much stronger force to register a movement)

So I hadn't luck with it. But I was planning removing its top anyway.
 
Last edited by a moderator:
reallynotnick posted on May 3 2006 at 09:38 AM said:
So the new models have a stick with the correct ratio of up to up left?

Damn......

First the nand, then the cap, and now the joystick.

I know we are talking about the stick ... but and what about the earphone jack? My gp2x broke again , and the other 3 units I sent to friends also had the same problem. All my friends here says that "the gp2x is indeed very cool , but unfortunately very poor quality . Things like that earphone jack falling inside in its very first use should never happen"

Are they wrong? :)

I mean ... I love my gp32 , and it never broke , I like my gp2x , but it was too expensive to experience so many problems like the stick , the nand , the cap , the usb connection , the firmware updating problems , the autobrick feature ... and so on.

I won't be fixing my unit soon because I am very angry at this moment.
 
Last edited by a moderator:
Rombic posted on May 4 2006 at 01:13 PM said:
reallynotnick posted on May 3 2006 at 09:38 AM said:
So the new models have a stick with the correct ratio of up to up left?

Damn......

First the nand, then the cap, and now the joystick.

I know we are talking about the stick ... but and what about the earphone jack? My gp2x broke again , and the other 3 units I sent to friends also had the same problem. All my friends here says that "the gp2x is indeed very cool , but unfortunately very poor quality . Things like that earphone jack falling inside in its very first use should never happen"

Are they wrong? :)

The headphone jack is placed too deep inside so to connect it plug must be inserted strongly. I fixed that removing the side plastic so I got also advantage of connecting wires for some other purposes as well. :)

Rombic posted on May 4 2006 at 01:13 PM said:
I mean ... I love my gp32 , and it never broke , I like my gp2x , but it was too expensive to experience so many problems like the stick , the nand , the cap , the usb connection , the firmware updating problems , the autobrick feature ... and so on.

I won't be fixing my unit soon because I am very angry at this moment.

It would be great if GPH would be selling GP2x's mainboards too. I could use such ones for small but powerfull appliances.
 
Last edited by a moderator:
Like Radek already suggested, I think the best route would be a daughter-board.
You can make one easily from diy-pcb kits that are readily available. You make a small board with its own pins, and mount the joystick to it. The joystick legs would have to be shortened so as not to protrude through the daughter board all the way. This should be fairly simple, if someone wanted manufacture some boards. I would do one myself, if i hadnt already done a d-pad mod.

You would have to do some measurements for sure, but here is a quick mockup to give the idea:

daughterboard.gif
 
It's not an elegant solution, but what's to stop you from simply snipping the legs off, mounting the joystick to the board with an insulating material (sticky floor tiles would work well ;)) and then running wires from the joystick to the contacts on the PCB?

Actually, looking at your diagram, this could be what you are suggesting...
 
I know there was some talk about 'painting' a conductive material to extend the primary direction contacts and perhaps paint an insulator over the corners of the diagonal contacts. Was that deemed unfeasible? seems less risky to me...
 
Radek posted on May 4 2006 at 01:54 AM said:
Knoxximus posted on May 4 2006 at 09:32 AM said:
guilleml posted on May 3 2006 at 01:06 AM said:
why sticks work better rotating them?

Yeah, can somebody shed some light on the logic here?

Sure - just look at its internals:

It's clear that diagonals contact pads are much bigger than the ordinal ones. By rotating it by 45 degress the situation will be reversed. Usually you want easier ords than diags so handling should improve.
Thanks for that excellent picture!

With the "Mode 2" programming model (to use Dave C's terminology) I don't personally think that the diagonals are too big in the current gp2x. Is there some guideline for how many degrees of angle should ideally correspond to a particular diagonal?
 
Last edited by a moderator:
FluffyPanda posted on May 4 2006 at 05:13 PM said:
It's not an elegant solution, but what's to stop you from simply snipping the legs off, mounting the joystick to the board with an insulating material (sticky floor tiles would work well ;)) and then running wires from the joystick to the contacts on the PCB?

Actually, looking at your diagram, this could be what you are suggesting...

The PCB shim will be the best solution for rotating stick. And once I see the new pins' arrangement then I will do it.
 
Last edited by a moderator:
Radek posted on May 4 2006 at 12:54 AM said:
Knoxximus posted on May 4 2006 at 09:32 AM said:
guilleml posted on May 3 2006 at 01:06 AM said:
why sticks work better rotating them?

Yeah, can somebody shed some light on the logic here?

Sure - just look at its internals:
pads9lo.jpg


It's clear that diagonals contact pads are much bigger than the ordinal ones. By rotating it by 45 degress the situation will be reversed. Usually you want easier ords than diags so handling should improve.

I see says the blind man. B)

Hmmmm.....what if you ordered another GP2X, and slapped the screen from your old one into the new one? That would pretty much be golden, eh?
 
Last edited by a moderator:
Radek posted on May 4 2006 at 11:33 AM said:
FluffyPanda posted on May 4 2006 at 05:13 PM said:
It's not an elegant solution, but what's to stop you from simply snipping the legs off, mounting the joystick to the board with an insulating material (sticky floor tiles would work well ;)) and then running wires from the joystick to the contacts on the PCB?

Actually, looking at your diagram, this could be what you are suggesting...

The PCB shim will be the best solution for rotating stick. And once I see the new pins' arrangement then I will do it.

Awesome.

BTW, I wasn't actually serious about drilling the original PCB.
 
Last edited by a moderator:
jmetal88 posted on May 4 2006 at 11:34 PM said:
Radek posted on May 4 2006 at 11:33 AM said:
FluffyPanda posted on May 4 2006 at 05:13 PM said:
It's not an elegant solution, but what's to stop you from simply snipping the legs off, mounting the joystick to the board with an insulating material (sticky floor tiles would work well ;)) and then running wires from the joystick to the contacts on the PCB?

Actually, looking at your diagram, this could be what you are suggesting...

The PCB shim will be the best solution for rotating stick. And once I see the new pins' arrangement then I will do it.

Awesome.

BTW, I wasn't actually serious about drilling the original PCB.

Please look at my latest latest updates about the stick replacement.

I can now (via some other company) manufacture decent printed board circuits (at decent prices even at low quantity). :)

Even double sided with metalized vias if needed. Nothing special really but after making boards in house using some thermotransfer/enchant that is great improvement.

Even better... (lets have something for future) ;)
 
Last edited by a moderator:
imnotahacker posted on May 7 2006 at 03:36 PM said:
so could there ever be a fix for this via software/firmware?

Software fix for GP2x's joystick?

That was dicussed long time ago with two developments:
- DaveC's biasing schemes
- my software routines for filtering stick's events in time domain ("timegate")

Actually some apps have implemented DaveC's schemes.
 
Last edited by a moderator:
FluffyPanda posted on May 4 2006 at 09:55 AM said:
Radek (or DaveC if you're reading this), how hard is it to open the stick and put it back together without damage? It looks like adjusting the length of the contacts wouldn't be too hard once it was opened, but I think I remember hearing that putting it back together tended to mess the stick up.


It is not hard. You just have to de-solder the stick from the board and separate the top of the stick assembly by lifting the 4 tabs outward until it pops off. You can't do it properly without de-soldering from the board.
 
Last edited by a moderator:
Radek posted on May 8 2006 at 12:57 AM said:
imnotahacker posted on May 7 2006 at 03:36 PM said:
so could there ever be a fix for this via software/firmware?

Software fix for GP2x's joystick?

That was dicussed long time ago with two developments:
- DaveC's biasing schemes
- my software routines for filtering stick's events in time domain ("timegate")

Actually some apps have implemented DaveC's schemes.

Do you mean the Joytest app?
 
Last edited by a moderator:
Back
Top