Idea: Gamepad repair guide / tools

[Arisotura]

I've had the idea of putting together a repair guide for WiiU gamepads, to help identify and repair common failure modes.

For example, I've observed that the gamepad's Flash memory going bad seems to be a common failure mode. If your gamepad seems to turn on, blue light comes up, screen maybe flashes, but it immediately turns off, it is likely a bad Flash.

I acquired a set of 4 gamepads, and two spare motherboards: out of the 6 total motherboards, 3 have a bad Flash (the 2 spare motherboards and one of the complete gamepads). 2 of the other gamepads work and one gives no signs of life, still haven't fixed it but I think it's a bad PMIC.

It is also evident that buying spare motherboards from Chinese eBay sellers is a gamble, these boards are not seldom damaged or at the very least have a bad Flash.

-

My idea was to help putting together tools and resources to help repair common failure modes, namely, the bad Flash case.

One would need to have soldering equipment for this sort of fix. Gladly, out of all the gamepad's chips, the Flash is the most human-solderable, but you'll still want to have some experience with SMD rework. Ideally you'll need a hot air station to cleanly remove the old Flash.

One would then need to obtain a compatible replacement memory. Something like a MiniPro (with adequate adapter) could be used to program the new chip.

The tricky part would be obtaining or generating a proper Flash image. This is where I'd want to try something.

I documented the Flash layout here: https://kuribo64.net/wup/doku.php?id=flash_layout

The firmware and language bank can be retrieved from Nintendo's CDN. The missing parts would be the bootloader, diagnostics firmware, remote control data, and startup screen data. The latter might require some imagination, I don't know. The diagnostics firmware is probably unimportant at this point. The bootloader might be replaced by an open-source equivalent if legality is a concern here. The remote control data could probably just be acquired from an existing firmware dump.

I'm thinking of putting together a basic tool to generate a Flash image from the provided data (which the user would have downloaded from Nintendo's CDN). I want to try this on the dead motherboards I have, to see if this is workable.

Would there be demand for this sort of thing?

 

[Tur7L3]

I think your discussion already sets the precedent.. 6 motherboards and only two are perfectly fine. This could become a problem down the road if Gamepad emulation isn't figured out (which thankfully, it is being). Anyhow, I think most people in this forum can agree the Gamepad is unique enough it must be preserved, and it gives much DNA to owning a Wii U at this point. Plus, your other plans suggest flashing from the PC to a Gamepad ideally, so use could become widespread among the interested.

TL;DR: Yes.

 

[Arisotura]

I think your discussion already sets the precedent.. 6 motherboards and only two are perfectly fine.

Little status update on this... actually, these "bad FLASH" boards seem to have mostly fixed themselves? Maybe that's actually related to battery stuff. I was trying to start them with a battery pigtail and it didn't work, but when I tried an actual battery, it did, and now even the pigtail works... weird.

That leaves one truly bad motherboard. That one was indeed a bad PMIC, and swapping that chip fixed it.

Regarding the bootloader, I did make a custom one (as part of my bigger project)...

 

[awesomeee]

I think GitHub has some website thing, perhaps you could make some site there if this idea goes anywhere?

 

[N7Kopper]

Little status update on this... actually, these "bad FLASH" boards seem to have mostly fixed themselves? Maybe that's actually related to battery stuff. I was trying to start them with a battery pigtail and it didn't work, but when I tried an actual battery, it did, and now even the pigtail works... weird.

That leaves one truly bad motherboard. That one was indeed a bad PMIC, and swapping that chip fixed it.

Regarding the bootloader, I did make a custom one (as part of my bigger project)...

Maybe the flash has some kind of self-refresh mechanism that needed time to kick in? Error recovery is hardly a unique concept in computing.

 

[Ysecond]

Great！！
Will the donor's flash data work properly on other PAD？

 

[Arisotura]

There's one catch: you have to make sure the language bank is installed at the correct address compared to the setting in UIC EEPROM.

It's either 0x900000 or 0x11000000. If you don't know what the UIC EEPROM setting is, you could just duplicate the correct bank to both addresses.