Hey guys, just posting to talk a bit about the DPad situation on the 8bitdo. It turns out that the Super NT is coming out soon and, in preparation, I decided that something had to be done with the 8bitdo dpad.
I own an SF30 and I first noticed DPad issues when playing Megaman 5 on my AVS but at the time chalked it down to me not being used to using the SNES pad. I would constantly get accidental diagonal inputs which would cause me to slide when I intended to jump (a very annoying experience!). It turns out that I wasn't alone with diagonal issues on the pad and it became a huge talking point around the announcement from Analogue. Last night, I decided to try a few rumors and quick fixes floating around the internet.
Before anything, lets talk about good ways to test input. There are a few homebrew roms and utilities out there designed to let you test out button configurations on your platform of choice. If you happen to have access to some arcade games like the street fighter series, you should be able to access the input test screen of your choice. I imagine most arcade games have this type of thing too -- just pick any of those methods that let you clearly see the inputs you are putting in. Another good test that I did involved link to the past, in which I tried to walk and look for cases where link would veer off course slightly from intended (which happened more often with an unmodded controller than I had originally thought.)
Also, open your controller at your own risk. If you don't want to be responsible for your own device, simply do not open it!
Anyway, some of you may have heard the electrical tape rumor floating around the internet and may have been skeptical regarding how much of an effect it would have. I certainly was! I'm here to tell you that it works, and it seems that the reason the dpad has so many accidental diagonal presses is the result of the dpads' center pole element being slightly too short. This may also be why there's so much division on this issue within the community, it could actually be a factory issue and only affect a handful of us. I don't have enough 8bitdo pads to verify that, though.
I've seen some images on the internet that went crazy with tape all over the controller, but honestly I wasn't happy with either the way that looked or how accurate it is in terms of testing. All you have to do is add a single 0.5cm square of electrical tape at the center between the dpad's conductive pads. Obviously avoid covering up any of the conductive pads. The type of tape you use is important because every tape has a different thickness, keep that in mind for later as well. This will prop the dpad up higher which means it requires more action for the metalic pads to connect with the PCB's conductive pads. This will also noticibly make the dpad more "clicky" as a result, which isn't exactly like a SNES controller but may be to some peoples preference anyway. If you launch your input-testing method of choice, you will notice that it is easier to get your left - right - up - down directions without getting any diagonals. Success, sort of...
The issue for me was that this made diagonals pretty difficult to do. This may be different from controller to controller, but it seemed to require very acurate diagonals. Now this may be ideal for some players, and it certainly isn't *impossible*, just a little more tough than I would have liked.
Since I knew what the issue was (the height of the dpad above the pcb) and the rough thickness of the material I used (it was around 0.7mil electrical tape), I decided to test it with a material that was thinner and also adhesive. I tried using a small piece of scotch tape which had roughly half the thickness to test it again. I wouldn't worry about the (very very very small) conductive elements of scotch tape too much but keep it in mind before using this method.
I found that accidental diagonals were still a problem with that method, but it seemed to tame the beast a little bit better.
Anyway, long story short, there is a non-permenant solution to the problem that helps identify how exactly the controller differs from the standard SNES. It may also explain why replacing the pad with the SNES controller dpad also works for some people (I haven't verified this myself yet, but I will try it out this weekend.)
There's definitely a more permanent solution to this problem, and I'll look into it more personally, but I thought I would chime in with the results of my own experimentation based off of the internet rumors. It is surprising how much of a difference a single piece of electical tape can make!