Closer To The Final Replacement Of Stick


grahf posted on Jul 7 2006 at 07:50 AM said:
Its a fairly simple install. The board should just snap right on to the stock stick pins. Mine was a little different since i removed the stock one completely, and had to solder new pins to mount it.

I am sorry guys, but only one picture came out on my camera. It was late, and i didnt have it set right.
board%20installed.jpg


This board works very well. The microcontroller Radek is using is much quicker than needed, so the diagonals register instantly when you press them. The center push function works good as well. On my GC pad i had to trim it a little bit to allow me to press all 4 directions at once to trigger the center push, but it will work ok on the QD pad that Radek is including.

Any question let me know.
I love the look of the fully soldered pins :).

I want to try a mini-stick with four button membrane microswitches. I have seen this in the Hat switch of a PC joystick (CH Flightstick Pro, had two given to me by different people, try a second hand store) and it seems robust and no "dead-zone".

I likely won't get around to it this year though. And your GC pad mod is brilliant, just need to find one with a broken analog stick for cheap :D.
 
Last edited by a moderator:
nubie posted on Jul 9 2006 at 11:23 PM said:
I love the look of the fully soldered pins :).

I want to try a mini-stick with four button membrane microswitches. I have seen this in the Hat switch of a PC joystick (CH Flightstick Pro, had two given to me by different people, try a second hand store) and it seems robust and no "dead-zone".

I likely won't get around to it this year though. And your GC pad mod is brilliant, just need to find one with a broken analog stick for cheap :D.


Is there enough room inside the Gp2x to perhaps have the original stick still in place but ALSO install the logic chip and a 8-way hat switch/dpad? I don't supposed there would be much point sense everybody says no game uses the 16-directions anyways.. It'd just be nice to have both. BTW, this is off topic but perhaps something like this could be done to the Zodiac? mines just collecting dust as we speak... :p

EDIT: I just looked at my GP2x and I see there would be no way to have both unless maybe the volume button was relocated and even then I guess the left speaker would be in the way.. looks like the 16-way stick has to get axed regardless.
 
Last edited by a moderator:
Radek,
Any updates on the MKII circuit?
Is there any reason I could not use the existing circuit in my MKII with a little rewiring?

DaveC,
How your project is is coming along?

Sorry to be impatient, but it’s kills me to use this insufferable joystick now that were so close to a replacement.
 
Well actually things are being worked on.

I didn't want to post anything as I don't want to give any false hope. The devil is always in the details. It always sounds easy at first until you actually try to do it.

I received Radek's kit along with a few qd pads. I laid out everything mechanical in a CAD program now . I have concerns with the copper oxidizing on the contact pads which will cause the operation to become unreliable after awhile. I was thinking of having the contacts gold plated but as you can imagine that adds alot of cost. Also the contact points on the metalized rubber seem to vary and are convex. This convex shape causes the rubber to make contact in a small spot making the point where it bridges the metal contact critical. Then when you do a diagonal this point moves slightly making the contact even less unless you press down harder to "mush" the metalized rubber pads to make contact. If the pad works intermittant or if you have to use more force to make it work sometimes it will be useless. I don't want to trade a bad stick for a worse d-pad.

I am at the point of doing a feasability test with different things. This means it is still R&D at the moment, no production. I have ordered $50 worth of different discrete switches the other day. When I get those I will test for the best feel and action. *IF* I find any of those acceptable I will go forward with the switch design that will probably only use the QD pad as an actuator for the switches. I won't bother with anything that is not high quality. It just has to be right. I will know more next week when the switches arrive what direction is best.

Then after all of this I have to be able to make this cheap enough so people won't bitch that it is too expensive. Radek will have to supply his chip too and program it so that adds a bit of cost too.

The good news is that if I do find a solution it will be high quality and will be a GOOD d-pad not just a d-pad. There are bad ones. I thought the one on the GBA-SP was bad, the one on the giz awful, the original DS poor (the one on the DS-lite rocks though), and the PSP bad. At this point I am hopeful but I can't promise success just yet.
 
DaveC, you well and truly rock. Thanks in advance; looking forward to your results :)
 
Instead of using a PCB to emulate the diagonals couldn't a software patch either be autoloaded or built into the next firmware to register for example d + l and dl. This woulkd save the complex wiring and I shouldn't imagine that hard to complete
 
That was discussed-- I suggested it actually, and I think 2 other people did. It'd work fine as a kernel module for say, C/SDL apps but for things that access MMSP2 registers directly (not many but still worth consideration) it wouldn't work.

As for 'dead' ... depends if Radek comes back. But suffice to say, I tried his kit, and it worked very poorly, it was basically unusable. Only a couple directions registered reliably, and you needed to MASH down on the pad. Getting it all to fit was a nightmare, the attachment with the polymorph or whatever it's called was a mess and dangerous if you didn't do it EXACTLY right. It was extremely difficult to install and in the end, never worked. There was no documentation provided and it wasn't even mentioned there were wires needed-- I had to blunder my way through a 20 page thread to figure out how to wire it up from some barely-adequate photos, and even wired up correctly, the IC he provided fried itself when connected. All it did was generate a LOT of heat. Using Bacteria's wiring method and the same equipment, the PCB was much too thick and the carbon pads rarely made contact. Resistance measurement showed thousands of ohms resistance across the carbon pads-- no wonder they worked poorly. Out of 2 included d-pads (Nokia QD), only one halfway worked-- the other had resistance SO high it was completely useless for all but one direction. Soldering to the PCB was a nightmare, it's more plastic than silicon/copper, and there isn't enough surface area. Pin attachment was a big hassle, and for it to be stable I had to solder everything together. I don't know why the kit was called solderless, I had to solder over 20 points, and I had to get my own wire, none was included.

The Polymorph is a great toy to play with though ... at any rate, I appreciate Radek sending it to me, but the kit's definitely not ready for prime time. I'm basically of the opinion now solutions should be provided from the factory or the reseller, because this isn't a feasible solution for the average consumer-- I couldn't get the thing working properly despite knowing the electrical aspects of it like the back of my hand, and a mechanical guru friend of mine trying to get it to FIT. I can imagine the difficulty an average user would have.
 
Never tried outcaST but I can tell that PoP runs fine in the genesis emulator of your choice with the GP2X' joystick and buttons
 
Back
Top