I did a similar project to this and I would love to know more details on what you did. I took the PCB of the controller, unsoldered the triggers and joysticks and then I connected digital potentiometers to replace them. It worked flawlessly with controlled input.
I had never heard of digital potentiometers until I read your post. Did some Google-ing and am quite intrigued. Obviously I don't have any laying around my workshop... If it turns out that PWM doesn't work and I really need an analog pot situation, I will have to go out and buy a couple. Thanks for the heads up.
That being said, my buttons were nightmares. I used an exacto knife to scrape the trace and solder to that tiny tiny piece with enameled wire. It was terrible and unstable. How are you handling your connection to the buttons?
Agreed. That was the hardest part of the 'wireless' version of the project.
The wired 'matrix' controller (and possibly the wired CG one, too) has test points for every button in question, so that was easy to solder up (although I did have to remove the triggers to get at a couple of them). And, if it hadn't been for the success with the 'wired' version of the project (albeit double the wires), I probably would have given up.
The wireless controllers (all of them based on my research) don't have any easy to get at points for the buttons (except d-pad right -- TP9). The rest of them were indeed the pass-through points, many of which were only accessible from one side and required the removal of the triggers, etc. I had a real hard time with the first controller. I just could not get the solder to stick on a couple of the points, even after I had carefully scraped off what I deemed was more then enough of the 'paint'. I ended up setting that controller aside (it still works normally, for now, but I had to re-wire one of the d-pads directly to the chip -- ugh), and moved on to a second. By then I had determined that I could carefully (and I mean VERY carefully) use a cordless Dremel with a fine metal 'ball' tip. I was able to get a nice copper shine with just a few very light passes (a vice and support for both forearms is highly suggested). Then, I cleaned the points with isopropyl alcohol, applied a little flux, lightly wiped off any excess flux and was able to put down itsy bitsy solder points. Finally, I pre-tinned my [28 AWG?] hook up wire (no idea how I ended up with a spool of it), snipped off the excess wire after the insulation shrank, and attached them to my solder points. I was able to insert the wires about half the depth of the circuit board, effectively making little 'hooks' on the ends. Since those points are undoubtedly fragile, I made sure the wire was already mostly bent along the final path it would have to take before soldering it down, and immediately pinned it at several points, initially with masking tape strips, and finally with a dab of hot glue here and there (usually a couple wires running next to each other at a time). All of the wires were tracked back to one of the rumble pack spaces (permanently removed) and zip-tied into a nice little bundle so that no one wire gets tugged on when I move things around.
If I was anywhere near as confident as that paragraph may sound, I would open up the controller and take some pics for ya... But, I will be the first to admit that I am afraid the simple act of opening the case one too many times could cause one of those flimsy solder points to pull off, possibly even pulling the tiny coper ring (which I may have ground down to almost nothing), would pop off, too, wrecking the controller completely. So sorry.
If I am ever going to prep another controller for this project, it will be a wired CG (I will have to go and buy one though). Since I doubt I will ever run my scripts wirelessly due to battery concerns, I really don't know why I bothered with the wireless controller in the first place. I just happened to have a few of them laying around, so that is what I used. But, I won't do that again.