Worklog Wii U R&D Thread (WURD)

zal16

.
Joined
Apr 4, 2024
Messages
15
Likes
1
I know the vWii don't works without the disc drive, but about the GC/Wii injects? They works without he (and de_fuse patch)?
It's possible to do a undervolt, specially on more demanding lines like 1v and 1.15v?
What is the power consupmition and current(A) in vWii mode? Is much less than last revison of original Wii hardware?
 

Lazr

.
Joined
Oct 3, 2023
Messages
49
Likes
132
I know the vWii don't works without the disc drive, but about the GC/Wii injects? They works without he (and de_fuse patch)?
A missing disc drive is fine for GC/Wii injections. They don't even know if the disc drive is even missing, as its entirely emulated.
 

YveltalGriffin

First Wii U Trimmer
.
.
Joined
Jun 7, 2016
Messages
366
Likes
1,566
Location
South Florida
Portables
5
Undervolting doesn't seem possible, at least with wires. Whenever I try to decrease the core rail voltages the system just hangs.

However, after brainstorming with Benge at MGC, I've decided to design a modified 1v/1.15v reg board that will sit on the motherboard next to the SOC and solder directly to the core rail copper. This will eliminate all voltage droop and should improve stability. There's a chance some mild undervolting will be possible with this new reg board.

I haven't measured vWii mode power, simply because it doesn't matter for portablizing until we get vWii disc drive patches. Keep in mind that in vWii mode, the console has a bunch of extra hardware blocks running to upres the video to 1080p and pipe it to the 1080p HDMI transmitter. Plus, all the Wii U hardware that's "turned off" in vWii mode still consumes some static current while idle. So it's virtually impossible for the system to beat an undervolted Hollywood-2 Wii.
 

zal16

.
Joined
Apr 4, 2024
Messages
15
Likes
1
Why not use stock regulators to undervolting? Are they that much worse?
 
Last edited:

Viilmo

.
Joined
Jun 23, 2021
Messages
216
Likes
108
Location
Germany
Portables
1
Why not use stock regulators to underclock? Are they that much worse?
Yes, you would have to trim them out first and they are physically bigger and much more inefficent
 

YveltalGriffin

First Wii U Trimmer
.
.
Joined
Jun 7, 2016
Messages
366
Likes
1,566
Location
South Florida
Portables
5
Right. They're huge, they need at least 12V, the 1v one is two-phase (meaning it has two giant inductors), and we don't have full documentation for the controller ICs.

The 12v input they require is the main issue. Going from 3.7V→1v is a lot more efficient than 12v→1v. And managing 4S batteries in a portable is trickier, since you need balancing. Custom regs solve all of these problems
 

Viilmo

.
Joined
Jun 23, 2021
Messages
216
Likes
108
Location
Germany
Portables
1
Right. They're huge, they need at least 12V, the 1v one is two-phase (meaning it has two giant inductors), and we don't have full documentation for the controller ICs.

The 12v input they require is the main issue. Going from 3.7V→1v is a lot more efficient than 12v→1v. And managing 4S batteries in a portable is trickier, since you need balancing. Custom regs solve all of these problems
They actually need 12V? I thought it was less so I didn't write it down.
12 fucking volts and all because drive go spin
 

zal16

.
Joined
Apr 4, 2024
Messages
15
Likes
1
I have no interest in trimming the board. I think about removing the useless disk drive and filling it with batteries. so I can play on the gamepad and keep the console wherever I am, without depending on wall sockets. It is also in my interest to make it more efficient, so I think I will choose this model that ends at -50, it will also be necessary to replace the regulators, regardless of whether they are spacious or not, just targeting efficiency. Using SD cards/flash drives and SSDs can help with low consumption too.
 
Joined
Dec 30, 2021
Messages
56
Likes
10
incredible project I have a Wii U, I want to use it without the disk drive, from what I understand I should use a picodrive plate, is there any guide to teach the process?
 

Lazr

.
Joined
Oct 3, 2023
Messages
49
Likes
132
incredible project I have a Wii U, I want to use it without the disk drive, from what I understand I should use a picodrive plate, is there any guide to teach the process?
You can just set the SEEPROM device to None in minute. Don't need a pico to do that, you can load minute with a browser exploit
 
Joined
Dec 30, 2021
Messages
56
Likes
10
You can just set the SEEPROM device to None in minute. Don't need a pico to do that, you can load minute with a browser exploit
my disk drive is defective, error 160-1400, I bought it like this
 

Lazr

.
Joined
Oct 3, 2023
Messages
49
Likes
132
The LOLWUN is real!!

Over the past few weeks I gained an interest in trimming the earlier Wii U board revisions. The WUP-50 will always have the superior trim, but with them being so rare compared to the WUP-01/WUP-30 (and to some extent, the WUP-40), it makes more sense to trim one of the common rev's than to risk breaking a WUP-50. I can't say I'm not a little bit bummed about how much bigger the LOLWUN is to the LOLWUT, but the fact it can be trimmed at all is super exciting!

Let me introduce, my trimmed WUP-40:
top-40.jpg
bottom-40.jpg


It's only slightly bigger than the LOLWUT, with the dimensions being around 123x95mm.
40v50.jpg

That's not too bad with it having a much larger SoC, along with the RAM being further out.
I haven't relocated any of the wireless modules yet, but those can always be done later. They don't halt booting into the Wii U Menu, it just makes doing anything impossible.

I have talked with Yveltal a little about the LOLWUN, it is going to stay an "experimental" trim for a little while. I'm not sure if it will ever be added to the guide.

Here is a video of me restoring a vWii NAND dump on the trim:

Along with this post, I want to share the pinouts I drew up while working on the trim:
(None of these are as fancy as the ones Yveltal made for the trim guide, but this isn't a guide, so its fine, probably)
WiiU_NANDEMMCPICO.png

I drew up this pinout months ago (right after the LOLWUT guide was released). It is meant to match up with the pinouts shown for the NAND/eMMC/WiFi/de_Fuse on the LOLWUT guide. The bare numbers mean the ONFI NAND, e means eMMC, d means DEBUG (for de_Fuse), and w means WiFi. Ignore the "TP2" via. That does not go to TP2, and its not useful in trimming.
smc.png

SMC wiring. I had used the WUP-01 board scans to find the I2C pins, nRST I managed to find by poking TP144 and the SoC pins with a multimeter.
sd.jpg

The nice part about having everything so spread out is that you get to keep some of the SD resistors by the SoC. Since I didnt label them on the picture, I will do so here:
Green - CD Dark blue - DAT1 Lighter Purple - DAT0 Orange - CLK Darker Purple - DAT3 Light Blue - CMD White - WP
The green-ish trace is DAT2, that is the only resistor for the SD card I trimmed off.
sdslot.png

The pinout here is meant to match up with the pins above. I'm sorry I didn't label them properly!
frontpanel.jpg

This is for the front panel, soldering the pins there makes the wiring cleaner IMO.

The SD slot pins and the front panel pins are identical between all revisions.
I hope this post gives someone the inspiration to trim one of these Wii U boards! Yveltal has said in the past that the bistro should be fine enough to power the A5X. For cooling, I would recommend sticking with the stock heatsink since you can actually properly mount it here (if you don't have the WiFi module that is).
bottom_heatsink.jpg


and maybe the stock fan (or whatever fan you want I guess). You want the fan to be blowing air into the heatsink.
 
Last edited:

zal16

.
Joined
Apr 4, 2024
Messages
15
Likes
1
@YveltalGriffin

I don't want to trim this console. I only want to reduce overall power usage, increasing efficiency.
Can you mark the all stock regulators are should be replaced with custon better ones?
Can you mark too the points to input the custom regulated voltages?

I am very grateful
 

Lazr

.
Joined
Oct 3, 2023
Messages
49
Likes
132
Here with another update! This should make trimming much better for everyone.

rsz_wiibt.jpg

The Wii Bluetooth module works (almost) completely fine on the Wii U!

For those who don't know, every Wii BT module has those testpads you see I soldered the wires to.
This is not the case on the Wii Mini/Wii U (but the focus here is Wii U). You have to get lucky, since not all modules have the testpads (and soldering directly to the B2B footprint is really annoying).

The only problem with the Wii module on Wii U is I guess the firmware on it is returning something the Wii U doesn't expect?
1716356150782.png

The way I got it working was simply to patch out the kernel from panicking:
C:
// Disable panic in kernel
ASM_PATCH_K(0x08129ce0, "bx lr\n");
Once I added this to my patches, it works completely fine! Even the Wii U Pro Controller pairs!
ItWorks!.jpg

It's pretty stable too. I played NSMBWii with a WiiMote and NSMBU with a Pro Controller for quite some time and I experienced no problems.

To try and combat the issue of it crashing, I have dumped the Wii Us BT module firmware, and @VoxelTek dumped the Wii BT firmware, but there is a massive difference between the modules though, so it's not a very easy "flash the Wii U FW to the Wii FW". The sizes are different, the bus the flash uses is different.. I'm surprised this even worked!
Since portables are most likely going to have isfshax anyway, I don't think requiring the kernel panic patch is much of an issue.

Something else I would like to share:
threeofthem.jpg

The top is the Wii module, middle is the Wii U module without the pads, and the bottom is the Wii U module with the pads. Quite interesting how different they are.

Big thanks to @LambBrainz for sending me some Wii modules for testing!
 
Last edited:
Joined
May 4, 2024
Messages
15
Likes
4
Alright boys, No promises. No guarantees. But lets see what we can do.

The Wii U MIGHT be capable of portablization, but there are a lot of caveats. It is going to take a lot of work to get there, assuming it even is possible. Even if possible, the amount of work and complexity to make a Wii U portable will be FAR greater than making a Wii portable, and the results will larger, and generally inferior. If it does become a reality, Wii U portables will not be a common or accessible project for new or even intermediate users.

And with that out of the way...

A few days ago @thedrew posted a picture of his water-damaged Wii U Motherboard.
View attachment 13342
At first glance, this looks like a pretty normal water damaged Wii U PCB. But I noticed something a little odd.
The Main SOC is WAY smaller than any other Wii U SOC. Prior to this discovery, the only 2 known revisions, the 01 and the 30, were extremely similar. Given the commercial failure of the Wii U, we assumed a substantial redesign was extremely unlikely. But, sure enough, this previously undocumented CPU-50 board has a much smaller SOC package. While we don’t know for sure, this could be a chip that received a die feature shrink. If so, this means the chip would use substantially less power, and run much cooler, similar to the changes from the 6 layer Wii to the 4 layer Wii.

The Plan
So where do we go from here? We’ve done this before for the Wii, so we know the basic procedure. Here’s the current plan:
1: Document all revisions of the Wii U
2: Compare the differences between them, and find the revision(s) most suitable for trimming and portablizing
3: Sand and Scan the preferred revision of the board in a manner where every layer is completely visible in the highest detail possible, in a manner where they all line up
4: Create a compendium using these scans, and map out as much of the circuit board as possible.

Once we have all this information, then we can proceed to the more difficult task of removing components, lifting pins, and sniffing data lines with oscilloscopes/logic-analyzers. From there, we can determine what custom software solutions will need to be developed. This is all quite a ways down the road.

Known Revisions:
01: Documented. Launch model. Larger Chip
30: Documented. Larger chip. Very similar to the 01, with only slight tweaks. No drastic changes noticed so far. Larger Chip.
40: Exists. Undocumented
50: Semi-Documented. Smaller chip.

Revision Identification:
You can use the same method of removing the battery tray and checking the code. You can also peek inside the SD card slot and view it silkscreened onto the circuit board. But the easiest way is by checking the label with the QR code on the bottom of the console below the serial and model number stickers. The last 2 digits of that code are the revision identification number.

Scans and layers:
@YveltalGriffin is currently working on scans of the 01. It is believed to be 6 layers.
He will be sanding down a dead 50 once he gets it in the mail from @thedrew.


Challenges:

Regulators:
We have yet to get the Wii U to boot on custom regulators. It has more voltage rails than the Wii does. (1v, 1.15v, 1.25v, 1.5v, 3.3v, 3.3v standby, 5v, 12v). The power consumption of each of these rails has not been measured, and it is not yet known if there is a difference in power consumption between revisions.

Disc Drive:
For now, the biggest foreseeable challenge is the Disc Drive. Games can be loaded with USB through homebrew, but the console will lock up if you attempt to boot it without the Disc Drive present. This will need to be fixed or resolved somehow, preferably through software.

WiFi Module (Internet):
The WiFi module is required to boot, just like the Wii. It will need to be relocated, unless our pizza lord decides to grace us with a NO WiFi IOSU patch. There are vias to solder to under the SOC.

Bluetooth Module:
Required to boot. 2 wires soldered to 2 traces, just like on the Wii. There are no vias near the APU, so it will need to be soldered to the surface of traces.
Gamepad: The gamepad is required for gameplay. I do not know if there is a patch to change this. The gamepad has its own dedicated chip (U7) on the motherboard, and uses the second wifi module to communicate with the console wireless-ly. I suspect the antennas of the console’s wifi module and the gamepad’s wifi module can be wired directly to each other to avoid interference, similar to wii and wii u dev kits. This will need to be tested.

NANDS:
Yes, NANDs, plural. There are 2 nand chips that will need to be relocated one of which is BGA. There are VIAs underneath the Wii U board. A flex relocation board for the 2 chips is the most likely solution, similar to the wii NAND relocation flex PCB sold in the bitbuilt store.

Video:
The Wii U Has the same video options as the Wii, with support for HDMI and 1080p. It has been reported that tthe vWii video quality isn’t as good as the Wii, even through HDMI. As of now VGA (like that used in aurelio’s VGA patch on the Wii) is NOT supported. The video on the gamepad is streamed wirelessly, and the gamepad board directly drives the LCD.


For now, this thread will be a central location to coordinate, consolidate information, and share notes. Hopefully we can make something out of this commercial failure.
Is there a dedicated discord server for this? Im a huge wii u fan and am super curious about how portablization is going.
 

Shank

Moderator
Staff member
.
.
Joined
Jan 31, 2016
Messages
1,324
Likes
2,834
Portables
6
Is there a dedicated discord server for this? Im a huge wii u fan and am super curious about how portablization is going.
Not for the Wii U specifically. It just takes place here on the forums and on the BitBuilt discord server
 
Top