Closer To The Final Replacement Of Stick


There are SMD versions of µC what I'm using but you would have a different PCB for that. I'd have to design new board or even resort to double sided PCB/metalized vias.

For such low scale mod it's not viable cost wise I'm afraid. And the size saving would be around 1/3 for whole unit (because there must be also place for connectors).
I understand that, I would be making my mod by removing the stick entirely and then soldering something in instead.

I just wanted to know how difficult it would be to program the micro-controller if it was a Surface Mount Device.

I would do all my own PCB work on my own.

I want to go about it in a different way (or at least think about doing it).

Such as single-sided board with the SMD controller and SMD Resistors and the button contacts (maybe even actuall clicky buttons :)), I would personally rather solder something like this in and completely save all the back room for other mods like USB host sockets or an internal WiFi ala Shuttle XPC circuit board.

I would investigate the size of the GC pad or NGage QD pad and then try to copy that precisely, then mount the SMD devices out to the side on top (assuming room in the faceplate of course), I would wire it with single core wire and make the wires go in whatever direction I wanted.
I think it might work, obviously it would be much more involved than your idea, but I just can't stand not messing with hardware :).

I just realized something, there must be room for this because there is already a buffer pad on the QD that is really unnecessary, the SMD can't be all that much higher off the board, we just need the room in the faceplate to put the controller and any necessary resistors, I am positive it can be done.

The only thing different about my idea is that the board can be single-sided, but it would need a plug. Narrow near the GP2X with locating pins, and a bit wider on the new PCB side. Just need to solder in the wires, plug in the plastic bit, then route the wires to the correct location, place the pcb (it could even clip to the new PCB), getting the wires in the proper solder holes, then solder it up.

If the plastic bit was to be molded in bulk in China or something it could be designed with little "clips" or grooves to hold the wires in the right spot for the GP2X joystick pinout, then the new PCB could be epoxied and soldered beforehand to the plug and the wires, just an in-and-out replacement.

I am going for the ultimate replaceable mod here, pull-out/drop-in and solder. I do see your point about the double-sided PCB, I wonder what a custom plastic bit would cost if the batch was big enough.

Less parts ;).
 
Last edited by a moderator:
There are SMD versions of µC what I'm using but you would have a different PCB for that. I'd have to design new board or even resort to double sided PCB/metalized vias.

For such low scale mod it's not viable cost wise I'm afraid. And the size saving would be around 1/3 for whole unit (because there must be also place for connectors).
I understand that, I would be making my mod by removing the stick entirely and then soldering something in instead.

I just wanted to know how difficult it would be to program the micro-controller if it was a Surface Mount Device.

With DIP I can use a socket and program a chip in it.
With SMD it'd be harder but I can do it. There is the ISP option too but...

I would do all my own PCB work on my own.

...you will have to chose a proper chip first. Then design a PCB around it.
Only then any programming will have sense as to get advantage of µC pins flexibility.

I want to go about it in a different way (or at least think about doing it).

Another ideas are always a good thing.

Such as single-sided board with the SMD controller and SMD Resistors and the button contacts (maybe even actuall clicky buttons :)), I would personally rather solder something like this in and completely save all the back room for other mods like USB host sockets or an internal WiFi ala Shuttle XPC circuit board.

I would investigate the size of the GC pad or NGage QD pad and then try to copy that precisely, then mount the SMD devices out to the side on top (assuming room in the faceplate of course), I would wire it with single core wire and make the wires go in whatever direction I wanted.
I think it might work, obviously it would be much more involved than your idea, but I just can't stand not messing with hardware :).

I just realized something, there must be room for this because there is already a buffer pad on the QD that is really unnecessary, the SMD can't be all that much higher off the board, we just need the room in the faceplate to put the controller and any necessary resistors, I am positive it can be done.

The only thing different about my idea is that the board can be single-sided, but it would need a plug. Narrow near the GP2X with locating pins, and a bit wider on the new PCB side. Just need to solder in the wires, plug in the plastic bit, then route the wires to the correct location, place the pcb (it could even clip to the new PCB), getting the wires in the proper solder holes, then solder it up.

If the plastic bit was to be molded in bulk in China or something it could be designed with little "clips" or grooves to hold the wires in the right spot for the GP2X joystick pinout, then the new PCB could be epoxied and soldered beforehand to the plug and the wires, just an in-and-out replacement.

I am going for the ultimate replaceable mod here, pull-out/drop-in and solder. I do see your point about the double-sided PCB, I wonder what a custom plastic bit would cost if the batch was big enough.

Close to nothing per unit I guess. But as you said the batch would have to be big enough. How big is enough - I don't know (and where).

Less parts ;).

As I understand you are going to integrate dpad with converter?

It's possible but it will need more mechanical work. Size constrains wil be also more harsh and changing a dpad might (probably) result in need of complete redesign including µC.

I was thinking about such idea too but one of my main goals was to make an universal converter first, then think about a controller. And be able to use many kinds of them including small joysticks too.

So... try it. For the µC side of things I can help.
 
Last edited by a moderator:
So....can we buy them yet, or at least pre-order.

The µC converter should be available very soon in EvilDragon's shop.
(he has some units already)

I'm planning some extra things using it as well. You will be able to write your own code for it if you would have such desire too.
(it using µC from Microchip - the PIC16F627A)

It has two free tristate digital I/O pins with RX/TX hardware USART.

The PCB board supports ISP programming.

It can run from 2V to 5.5V. The current is far less than 1mA (actually with current firmware it eats less than 200µA). It has built in 4MHz quartz and 1mips of processing power.

The QD dpad should be availabe soon...
(I could sell it already but I'd like to improve one thing)

I don't get pre-orders.
 
Last edited by a moderator:

Intersting... I did a search, though, and that is $65 from the first store I've found. I'd rather buy one of the $20 ones off ebay myself, and crack the case open to do my own internal thing, if I were going about it like that. With me, money is the primary concern, really, I understand yours may be portability and reliability, but mine, is cost.
 
Last edited by a moderator:
Would these arrive as a 'kit' of sorts? Or will you just get a PCB and a D-Pad and have to do a bunch of intermediate steps yourself? It'd be nice if this was as easy as possible for the mechanically-challenged like myself.
 
Would these arrive as a 'kit' of sorts? Or will you just get a PCB and a D-Pad and have to do a bunch of intermediate steps yourself? It'd be nice if this was as easy as possible for the mechanically-challenged like myself.

The converter and dpad will be connected already in a "kit".

Mounting converter is trivial and it can not be easier.

The same can not be said about a dpad unfortunately.

There will be need of drilling front panel. It will need some skills. Then the dpad element has to be mounted inside. Again some skill will be needed.

I will try to make that step as easy as possible. But face it - it's a hardware modification after all.

Ideal would be just a decent 4-switch based stick. Then no drill would be needed but I couldn't find suitable one. It such case the existing stock stick could be just cut and replacement put simply in its place.

So... if you aren't sure that you can do it then don't.

I'm sure that dpad could be done better. Some persons are experimenting with other ideas as well. I wish it would be easier. Do you know in what state is my GP2x after all those experiments on it?

Don't ask. :rolleyes:
 
Last edited by a moderator:
What is necessary besides driling out the faceplate and sanding down the stick base a bit? Anything else? If that's all, I should be able to handle it without too much hassle.
 
Or will there be pre-drilled case cover for this purpose selling from shops too? I suppose such parts won't be difficult to order from GPH...
 
One thing occured to me Radek: in your earlier posts you mentioned that the centre push action would happen if all 4 directions are pressed at the same time. Sounds like a good idea, as a centre push button is needed for most games, but having played and disassembled some D-pads (for research), I discovered that the middle of the D-pad is raised, so that it pivots the D-pad itself to ensure that one direction is only pressed at a time. This is especially true of the Nintendo pads. Without this pivot, it would be hard not to press all 4 directions at the same time all the time, and compremises the usefulness of the D-pad itself. I had to overcome this issue myself when I made my various D-pads (I am still using the one I made about 6-7 weeks ago incidentally), which is why I am aware of the problem.

The D-pad you are using must either therefore logically suffer from the same problem, or have some method of stopping this (if the pads you use only rely on the rubber mould (separating the contacts from the board), it would be too easy to press multiple directions at the same time by mistake). It is a fine line between having the ability to press all 4 directions at the same time intentionally, and doing it by mistake; and pressing two or three directions at the same time instead of one direction.

I would suggest that either you have (intentionally or otherwise) solved this problem, or may be unaware of it. If you have solved it, it would be interesting to hear how; if you are unaware of this issue, I suggest you test your D-pad by playing a suitable game (Vektar, or Marble Madness for example) and see how well it behaves to your expectations.
 
I`m sorry, I just can`t read throughthe whole 16 pages/230 posts here, but I am very interested in your dpad solution... my question is:

Radek posted on May 5 2006 at 05:58 AM said:
...
26th.jpg

...

With this logic controller, there will be no "pad" button? It would be quite bad to miss the extra button... (I`m sorry again if this has been asked/answered already)
 
Last edited by a moderator:
What is necessary besides driling out the faceplate and sanding down the stick base a bit? Anything else? If that's all, I should be able to handle it without too much hassle.

The sanding down of stick base might be actually not necessary. I got the new batch of PCB what are much thinner but I can not test it as I lowered stick base in my GP2x already.

What else? If the drilling would be done with high degree of precision then it'd be all.

I will send some samples to some person first to test.

Or will there be pre-drilled case cover for this purpose selling from shops too? I suppose such parts won't be difficult to order from GPH...

Well... I can not buy anything from GPH but some other persons can. Eventually they might offer premodded GP2xes.

One thing occured to me Radek: in your earlier posts you mentioned that the centre push action would happen if all 4 directions are pressed at the same time. Sounds like a good idea, as a centre push button is needed for most games, but having played and disassembled some D-pads (for research), I discovered that the middle of the D-pad is raised, so that it pivots the D-pad itself to ensure that one direction is only pressed at a time. This is especially true of the Nintendo pads. Without this pivot, it would be hard not to press all 4 directions at the same time all the time, and compremises the usefulness of the D-pad itself. I had to overcome this issue myself when I made my various D-pads (I am still using the one I made about 6-7 weeks ago incidentally), which is why I am aware of the problem.

The D-pad you are using must either therefore logically suffer from the same problem, or have some method of stopping this (if the pads you use only rely on the rubber mould (separating the contacts from the board), it would be too easy to press multiple directions at the same time by mistake). It is a fine line between having the ability to press all 4 directions at the same time intentionally, and doing it by mistake; and pressing two or three directions at the same time instead of one direction.

I would suggest that either you have (intentionally or otherwise) solved this problem, or may be unaware of it. If you have solved it, it would be interesting to hear how; if you are unaware of this issue, I suggest you test your D-pad by playing a suitable game (Vektar, or Marble Madness for example) and see how well it behaves to your expectations.

The implementation of center push can be changed as the chip can be repgrogrammed and it has two free digital I/O pins. So center push could be done a different way.

I was examining the QD dpad and it has built a rubber spring on its center too. So even without raised middle it works adequately. It's realieble to hit ordinals, somewhat more difficult the diagonals.

I was experimenting with raising the middle as well. I didn't like it that way but it's just my opinion. I will send some samples to test first of course. There is always place for an improvement - I will not be selling anything if it's not decent enough.

There is always possibility to add some filtering/debouncing for events in a microcontroller too.

So... good that you made your point. My implementation of dpad isn't perfect but it's simple. I can buy QD elements rather easily and I can make PCBes too. I can only use what I can get.

I`m sorry, I just can`t read throughthe whole 16 pages/230 posts here, but I am very interested in your dpad solution... my question is:

Radek posted on May 5 2006 at 05:58 AM said:

With this logic controller, there will be no "pad" button? It would be quite bad to miss the extra button... (I`m sorry again if this has been asked/answered already)

What "pad" button? The center push?

The converter is capable outputting the center push. In a fact it can even read the stock's inputs as its I/O has tristate buffers. For now it's implemened by all four switch pushdown of QD dpad.

The whole kit looks like:
p14kj.jpg

p38tg.jpg

p24fv.jpg


Perhaps it'd be good idea to start a new topic about it? I understand that this one is huge and harder and harder to get track of.
 
Last edited by a moderator:
Looking very good :) , I'm really looking forward to the release of this mod.
 
[The whole kit looks like:
p14kj.jpg
This looks nice. I don't need it, I have a fully working alternative (this has been discussed previously, so isn't relevant here). You have obviously put a lot of work into the project and it is nice to see how far you have come with it - it is neat and tidy.

The three concerns I would have are: 1) the centre push button 2) the comment about it being "difficult to press the diagonals" - that would be a problem; and 3) positioning the D-pad correctly inside the GP2x to align to the hole correctly which needs to be cut in the casing - a mistake in cutting the case is not an option!!

One alternative to the centre push issue - how about providing a simple switch (you can buy them from any electrical store) which can be connected to the board, and a hole made for the push-button on the side of the GP2x case itself (the switch is about 10mm long, so would be concealed inside the GP2x with the button sticking out) - this would remove the need to press the 4 directional keys at the same time, and give a neat solution - just a thought - and cheap. Anyone willing to bore a hole out of the GP2x for a D-pad will hardly moan about making a 4mm hole in the side of the casing too!

Interesting about the "spring" in the D-pad - that is obviously how they got around the problem. My solution uses a "spring" in effect too - a self-adhesive foam pad (about 1.5mm thick) - which in essence is a springy material and does the same job - surprisingly, it works reliably and robustly! I had to do a lot of experimenting to get it "feel" right.
 
Last edited by a moderator:
Intersting... I did a search, though, and that is $65 from the first store I've found. I'd rather buy one of the $20 ones off ebay myself, and crack the case open to do my own internal thing, if I were going about it like that. With me, money is the primary concern, really, I understand yours may be portability and reliability, but mine, is cost.

Maybe I meant A la, as in "like the shuttle circuit board", just removing the USB connector completely would be nice.

@Radek, I am sure I mentioned making another topic, and/or changing the first picture that people see when they hit the topic, it is mighty confusing because It has nothing to do with the current state of affairs.

Maybe another topic needs a good name, like: "Radek's D-Pad mod", or something a little less cryptic than the current title.

I am not very creative when it comes to titles though, it might not matter if a Mod can change the titles of Topics :), as long as it always says Radek in it somewhere we can always find it.
 
Last edited by a moderator:
Intersting... I did a search, though, and that is $65 from the first store I've found. I'd rather buy one of the $20 ones off ebay myself, and crack the case open to do my own internal thing, if I were going about it like that. With me, money is the primary concern, really, I understand yours may be portability and reliability, but mine, is cost.

Maybe I meant A la, as in "like the shuttle circuit board", just removing the USB connector completely would be nice.

Sorry, I don't read things too thoroughly sometimes. :rolleyes:

Ok, I get what you were sayng now. Just try to find one that doesn't run too hot for an internal installation, though. The Wifi stick I use in my laptop runs quite hot at times.
 
Last edited by a moderator:
[The whole kit looks like:
p14kj.jpg
This looks nice. I don't need it, I have a fully working alternative (this has been discussed previously, so isn't relevant here). You have obviously put a lot of work into the project and it is nice to see how far you have come with it - it is neat and tidy.

And indeed it's not revelant here. I hope more other dpads/replacement will show. And it's nice you like this mod.

The three concerns I would have are: 1) the centre push button 2) the comment about it being "difficult to press the diagonals" - that would be a problem; and 3) positioning the D-pad correctly inside the GP2x to align to the hole correctly which needs to be cut in the casing - a mistake in cutting the case is not an option!!

1) It's a compromise - for me one hole in front panel is enough :)
if someone would like to have a different configuration no problem. I will provide custom firmware for the chip.
2) Not that hard - matter of preference probably. I don't insist it's perfect.
3) A major issue when mounting as drilling has to be done with a precision. I got around it using hot polymorph and hairdryer to make it in that state longer. Then tweaked position of dpad to my liking.

Have an another idea how mount the dpad element. I will be testing it today. Should be more of "plug and play".

But of course - it's mechanical work and it has to be made with some degree of skill. For an example putting too much force when using a file and front panel might cracks, then what? ;)

One alternative to the centre push issue - how about providing a simple switch (you can buy them from any electrical store) which can be connected to the board, and a hole made for the push-button on the side of the GP2x case itself (the switch is about 10mm long, so would be concealed inside the GP2x with the button sticking out) - this would remove the need to press the 4 directional keys at the same time, and give a neat solution - just a thought - and cheap. Anyone willing to bore a hole out of the GP2x for a D-pad will hardly moan about making a 4mm hole in the side of the casing too!

I was thinking about using the existing side buttons for that. They aren't often used so they could be used as center.

Interesting about the "spring" in the D-pad - that is obviously how they got around the problem. My solution uses a "spring" in effect too - a self-adhesive foam pad (about 1.5mm thick) - which in essence is a springy material and does the same job - surprisingly, it works reliably and robustly! I had to do a lot of experimenting to get it "feel" right.

I'd prefer to have microswitch based controler myself. I do like click like behavior but it'd be much more complicated task.

There is one possibility what would be actual almost perfect but it'd need some special parts I couldn't find.

The major advantage of the QD dpad is it has almost everything included.

Intersting... I did a search, though, and that is $65 from the first store I've found. I'd rather buy one of the $20 ones off ebay myself, and crack the case open to do my own internal thing, if I were going about it like that. With me, money is the primary concern, really, I understand yours may be portability and reliability, but mine, is cost.

Maybe I meant A la, as in "like the shuttle circuit board", just removing the USB connector completely would be nice.

@Radek, I am sure I mentioned making another topic, and/or changing the first picture that people see when they hit the topic, it is mighty confusing because It has nothing to do with the current state of affairs.

Maybe another topic needs a good name, like: "Radek's D-Pad mod", or something a little less cryptic than the current title.

I am not very creative when it comes to titles though, it might not matter if a Mod can change the titles of Topics :), as long as it always says Radek in it somewhere we can always find it.

Perhaps I will make a new topic soon.
 
Last edited by a moderator:
This is a last (most probably) update as everything is ready. I will show how to mount dpad now. So... there is some of polymorph plastic involved what must be heated up to around 70 degress in a water:
m10rf.jpg

It's transluent then and elastic:
m26vy.jpg

It's time now to put a blob where the stick's base is:
m34xf.jpg

Then some quickness will be needed because the case must be closed (don't have with all parts) to use that polymorph elastic property:
m66ee.jpg

That time when polymorph is still elastic can be extended using hairdryer. There is some time to tweak a position of dpad but if it didn't get as it should... just remove a plug and repeat the procedure:
m89ql.jpg

It can be done many times as needed. The polymorph isn't conducting electricity too and it's acting with a dpad's PCB as a weak glue.

The wires connecting dpad with converter are placed here:
m92kj.jpg


That's all.

I bought some QD parts and will assemble few kits shortly. Once I will get my sorry LCD running back I will make a new thread showing this mod in a complete form and...
 
Back
Top