Worklog Wii U Mini Worklog

[Lazr]

Guess it's my turn for one of these.

I've been planning on doing something like this ever since I saw Yveltal's Preliminary trim and how similar it looked to the Wii Mini, and with the Wii Micro being a thing, I thought something similar could be done with the Wii U.

I cut out a Wii U board specifically for this:

These are pretty old pictures, but you probably get the idea. (feel free to roast the puny 1.5v and 3v3 wires)
I really want to have the console use the stock heatsink, fan, and plastic shroud. Mounting the heatsink should be possible with the other 3 screw holes being free from a cafeflex. The fan and plastic shroud are mounted to the plastic things on the sides (not a clue what they are called).

I did also want to use the stock 1v/1.15v regs, but after destroying the 1.15v reg (I am still not sure how) and killing another WUP-01 trying to get PGOOD stuff working, I've decided to just make a reg board that takes 5v in. The RPi5 PSU should be more than enough to power the WUP-50, 25W of raw power! It will have 1v, 1.15v, 1.5v, and 3v3. I don't need to worry about 1.25v and 1.1v, since on the WUP-50, those just need 3v3 and they are right next to the DRH, completely untouched.
Speaking of stock regs, the 5v reg just works when its provided 15v. I'm not sure if it can be given any lower, since I never tested it. (You can see it on the left of the Wii U board in the first pic)

This is the current state of that WUP-50 board:

It looks a bit ugly right now, but eventually it will be a working Wii U running on custom regs, where the 1v and 1.15v is soldered directly to their respective rails, with 1.5v and 3v3 also getting power from the reg board.

 

[Mr Game & Cube]

 

[Lazr]

After the thread creation, I wired up the bistro just to test the board (since I haven't actually powered it on in awhile). I also wired up the SMC/RTC and a pico (because I love the logging with the pico):

But the system doesn't power on..

I am 99% sure I have the SMC and RTC soldered correctly and I checked each voltage multiple times, so I'm unsure what the issue is. Might have killed the board entirely, I think I borked the always-on 3v3 rail and the original RTC before I trimmed, so maybe that also broke the SoC. The RTC im using here was is from the WUP-40 board, so this RTC should be functioning completely fine. It actually turning on will yield results like those shown on the de_Fuse release page, or it just boots the boot1 image from the SLCCMPT (I don't remember if I flashed it or not).

Because this system might be dead, I bought another WUP-50 from Japan. It might be a little while to arrive, but that gives me time to think about the 1v and 1.15v regs and to figure out if this board is actually dead or not. If it isn't, I have another WUP-50! I also have an idea for a flex that solders to the SD vias, the BT traces, and the back USB vias. Maybe the SATA traces could be on that flex too, to use a mSATA SSD for the MLC instead of an SD or eMMC. We'll see, how much room is at the front of the console for that.

 

[Lazr]

I started (and finished) the reg board:

I'm calling it "minireg" because..well, its for the Wii U Mini.
Since I am still a big noob at KiCAD, I based the design off Yveltals bistro, because it already had whats needed and it wasn't very difficult to adapt for this. The design is open-source, but I haven't tested it yet, so I'd recommend waiting a little while before trying to use it.
It's about $20 with 5 PCBs and an electropolished stencil.

Next on the list is the SD card flex. I am thinking of putting a MicroSD slot on it. It will also feature the BT module and break out the vias to some testpads for the back USB ports. I'll get on that soon!

 

[Lazr]

I forgot to mention this in last nights post, but the RPi5 power supply is handling the WUP-30 like a champ. It was running at 15v, but now I'm confident that it can power the WUP-50 with 5v. It might work with the previous board revs, but honestly supporting them with this makes no sense to me, because the WUP-50 is much smaller.

 

[WizzardoftheDisc]

Next on the list is the SD card flex. I am thinking of putting a MicroSD slot on it. It will also feature the BT module and break out the vias to some testpads for the back USB ports. I'll get on that soon!

I'm very interested in the SD card flex. I love seeing flex cables that add stuff and this one sounds like it'll have a bunch of fun stuff!

 

[Lazr]

I actually started designing the flex not long after ordering the boards and BOM for the minireg. I'm pretty sure i've put everything I wanted on here, minus some TPs for the SMC if someone wanted to use those.

It will be a while before i order these though. Spent sooo much money last night just on the minireg..

 

[WizzardoftheDisc]

I actually started designing the flex not long after ordering the boards and BOM for the minireg. I'm pretty sure i've put everything I wanted on here, minus some TPs for the SMC if someone wanted to use those.
View attachment 34583
View attachment 34584
It will be a while before i order these though. Spent sooo much money last night just on the minireg..

I can't wait to see how these turn out! I'm gonna be watching this project like crazy for updates, I have a feeling the final product is going to be amazing!

 

[Lazr]

It will be a while before i order these though. Spent sooo much money last night just on the minireg..

I lied, they were quite cheap. About $25 in total (the boards, a stencil, and shipping).

I also had to redesign the SD slot slightly, I didn't realize that the one I used was actually a hinge-style, so I switched to the SD slot used on the cafeflex.

 

[Lazr]

The Japanese Wii Us have arrived. Both are WUP-50s!

As you can probably already tell by the image, I already tore one apart and am going to be using that one for the Wii U Mini. The other board I showcased here I started sanding down instead since it was most likely dead when I tried something dumb with one of the 3v3 rails. Oh well, doesn't matter too much.

I finally got an ABLIC bypass working, which is cool as I can test with (almost) all of the stock regulators as I go along with the process. (obvs not the standby 3v3 rail)

I have the 1v PGOOD pullup elsewhere, since its inside my trim line and its the only one that really matters here.

I didn't have the proper size resistor to force ON the 1.5v rail, but the vias were far enough apart where I was able to make the 10kOhm resistors I already have work. You also need to cut the TP149 trace (can be seen on the ABLIC bypass).

I think at this point I can just trim, but I'd like to get all of the wireless module relocations working beforehand because it's good practice to do so and its probably much easier when you haven't trimmed the board.

Spoiler: Extra funny thing

So the 12v rail on this console was causing it to randomly shut off. I'm going to include what I did to remove the 12v rail entirely:
1. Remove the inductor
2. Connect TP132(3v3/12v PGOOD) to TP122 (3v3)

Easy enough. This of course removes the ability to use the disc drive forever (until you properly fix that rail), but I kinda don't care.

 

[Lazr]

Be prepared for a picture-heavy post, today has been a long day for me and got lots of work done, and I come with great news!
Relocated the 5Ghz module (the one used for the Gamepad). It took me a few tries but after changing my routing and adding the resistors, it worked immediately.

Next I did the Bluetooth module. I do have modules with the big pads on the back, but I just used the module that came with the console. They do have pads as well, just they are slightly smaller and might be a little bit more annoying for a beginner. At least you don't have to solder to the B2B for everything except connecting 3v3 to that SoC pin.

Next I tested the regulators I got for cheap off AliExpress. I am using them for 3v3 and 1.5v since they are the most annoying rails to deal with. (I don't have a good picture of the 1.5v reg being used)

I actually never got the minireg boards made. I'm pretty sure I sent a message to the JLC engineers but they have yet to reply.
It's alright though, since the stock 1v and 1.15v (and even 5v) regs can be used, just need to pull the 1v PGOOD high.
Eventually I did extend the wire to the 2.4Ghz WiFi module, as it won't get power from anywhere else (the location that gives it power is cut off).

Then I actually trimmed the console, and relocated the SMC and RTC first. Cleanest relocation of those guys I have ever done.
You can see the place I used for the 1.5v regulator and yeah the daughter board has some masking tape under it to prevent a short.

Next on the list was the NANDs. I wanted to challenge myself here while also avoiding the screwhole for the heatsink/heatsink clamp, and it turned out really well! Trace soldering is not that bad once you get used to it.

I need to redo the eMMC wires because the console doesn't recognize it yet, but it's not something that's super annoying to do.
There was a small hiccup with the NAND relocation, because I needed to move it to a different board because I messed up the one I trimmed from the original board (got too close to the resistor pads -_- rookie mistake I guess). Anyway, the problem was that not all of the pins were connected. I was really glad to see the error code da on the Pico after I wired it up to the console, since it meant that the console was still alive and well, but it couldn't init the NAND. Easy fix.

And finally, the system booted up!....Almost...
I never reflashed the eMMC, so I still need to rebuild that, but to do that I need it to actually be connected to the system.
I also need the SD card to rebuild, but SD card is easy to deal with:

It's really ugly, but it's only temporary until the bt-rex flex arrives, or if I find something else that'd work.

Here is what the top and bottom of the system look like as I am writing this:

And here is a picture of the console on in the minute_minute menu:

It's 3AM as I am writing this, but I can only imagine what will happen later today(?) to make this thing completely functional. Exciting stuff!

 

[Lazr]

I got the system to boot today!

I gave up on trying to use an eMMC, so I am just using a redirected MLC. Works quite well (even with my less-than-ideal SD card wiring).
The partition size for the MLC is just the standard 32GB. It's a 64GB SD card and I don't want/need any crazy sizes to store games.

I forgot to include what I was doing with the 2.4GHz WiFi module in yesterdays post, but that's fine, I can post about it here.
Since the connector footprint is well in the trim line, we can re-use it.

There is a pin that needs to be pulled up for the WiFi module to be functional, it's really easy to do.

I'm not worried about how much the module extends the front, but I need room for the front panel anyway.
It's a blessing for the previous board revisions, since those will be supported just because of the extended front.

EDIT: I would recommend removing the can from the WiFi module OR shield it very well from the board. You don't want to accidentally short 3v3 to GND!!

I also re-added the Pico. I wasn't wanting to (as I wanted to challenge myself. All of my trims have had a Pico wired up) but I was stumped as to why the system wasn't booting so I wired it up anyway.

Turns out that not all of the NAND pins were connected. Enough were to load boot1, but aside from that, the system would not boot into CafeOS. After fixing that, we're booting!
I did need to set up the redMLC again (format to WFS, install system titles..), but that's just boring software stuff.

Finally, I also moved the 1.5v regulator off the 3v3 regulator, to 'torture' 5v instead. With it pulling power from 3v3, the 3v3 reg would get extremely hot and shut off, which is not good. The stock 5v regulator is actually quite powerful. I don't know any specifics, but I do know that it can power a whole range of USB devices, all plugged in at the same time, so it should be able to provide enough amps for the 1.5v reg.
The downside to moving it is it messed with my perfectionist wiring, since I needed to lift up the SMC/RTC daughterboard, but it could be worse.

Now I desperately need to learn Fusion, or any other 3D modeling software, since we are sooo close to a finished product (and my first ever project working with a trimmed console)!...or I can just go play Mario Kart.. /s

 

[YveltalGriffin]

This stuff is really challenging, but you make it look trivial! Awesome work.

If you want to be extra safe, you can also tie Wi-Fi pin 9 to 3.3v, the way Cafeflex does. Just to ensure it's in the correct state.

 

[Lazr]

Thanks haha, it seems like I always make things look trivial.. lol.
Eventually I would want to pull that pin high, but I dont have good access to it atm

 

[Lazr]

Quick update on the hardware side of things:
I finally wired up USB. The top port is the 'USB0' port (for vWii stuff I guess?) and the bottom port is one of the front USB ports, just in case I need to boot into the recovery_menu for whatever reason (the UDPIH exploit only works on the front USB ports).

For the front port, you have to solder to the trace pairs. It's not too bad, but I did also solder to the traces on the other side too

I didn't twist the wires like I should have, but it works, so I am just going to leave it..

For the back ports, I used the vias on the bottom of the console (not pictured) and I had to remove a resistor pair before it started working, but not the end of the world.
EDIT: They were not resistors, but common mode chokes.

And finally, to get power to the ports, I did the classic "ignore any protections and jumper across 5v"

I think this is the last thing I wanted to do, since now the console is completely functional, minus two USB ports, but two is just fine.

 

[Lazr]

OK it's been way too long since I last wrote here, so here are the updates:

1. bt-rex was mostly a failure, but I was at least able to use the SD card portion, with slight (lol) modifications.
- When I imported the scans into KiCAD, the scale was way out of wack, so the BT module footprint was too small and the traces weren't even close to being right. I also forgot that the back ports had accessible vias on the bottom when designing, so it was pretty much useless from the start.
- I somehow swapped around DAT2 and CD. I had to cut that part of the flex off and then run a wire from DAT2 to the trace on the flex.

2. minireg was also kind of a failure.
- The 1v and 3v3 rails are busted, so I need to order new parts from mouser for that, but it's not entirely necessary so I just haven't.
- The 1.15v IC gets incredibly hot when powered on. I'm not sure why. Maybe I read the datasheet wrong when seeing if I could use 5v IN? It does output the correct voltage though, so its a mystery to me for now.

3. I have done almost nothing to the shell for this.
- I made a rectangle in OnShape, but that's about it. I've barely tried learning OnShape, which is a problem as I was going to use this project to learn before I start an actual portable..
- Just been burnt out I guess. The project isn't dead, but motivation to do anything with it is really low, been working on other stuff too, like Wii unbricking guides.

 

[Lazr]

Weird update, and no I still have yet to do the shell..

A few days ago at the grand ol' 3AM I had a random thought. I don't remember what the thought was, but it made me realize that I didn't actually know what turned on the 15v PFET. The ABLIC bypass always forced it on to get 15v across the board to the regulators, so couldn't really tell if it was the ABLIC or the RTC that turns it on. After sleep and remembering that I wanted to test it, I grabbed one of the Wii U boards I had torn down and did the ABLIC bypass there, but with two major differences:
1. Bring in my own 3v3 power supply for the standby rail. That will power the SMC and RTC.
2. Leave the PFET gate alone. If it was forced on, then I would just be wasting my time.
Anyway, after plugging in the console and giving the standby rail power, the red LED was turned on! After pressing the power button, the console did too! This is great news, as I can now have:
- Standby mode enabled (with Quick Start as an added bonus)
- The damn thing doesn't immediately turn on, and instead the PWR button has to be pressed.
Basically the same functionality as a stock console, which should be the goal with most of these projects.
The one downside is that there's another daughterboard that needs to be cut out. Luckily it's only needed for A5X boards, as the WUP-50 has the PFET gate by the power plug.
Here's a picture of the hackyness while I was testing:

Also I've "adapted" (changed a single resistor and removed the 1v/1.15v rails) the PSU-Plus board to be used here.
I already know most of the rails just straight up work on the Wii U. I even overvolted the 1.5v rail with 1.8v by using the base PSU-Plus. Generally it is not a good idea to do that, but I am an idiot. :^)

This was done before testing what enabled the PFET, so I still need to add the standby rail to it, but it should just be a basic copy-paste of the existing 3v3.
I did want to add an optional 1.25v rail to it for WUP-01 boards, but not sure if I even want to bother with it? We'll see I guess (WUP-01 boards have the 1.25v regulator right next to the 5v regulator. Every other board revision has it by the DRH).

Last little thingy ill show is a flex to relocate the SPI flash to the DRH. The flash sticks up incredibly high, so securing the 5GHz module with glue doesn't last very long.
(Thanks Yveltal for the vertical suggestion!)

It'll sit right next to the DRH, and the intention is that its compatible with all boards. The 3v3 via is designed for the WUP-50, but I added a pad and a line to cut at for the previous revisions. The via should technically work, but it wont be very secure since its not entirely on this copper area:

On my latest LOLWUT, the 5GHz module is very secure with the flash moved to the top.
Unfortunately its not entirely wireless, as there is no easy to reach via for the CS pin and in my testing it is required for the DRH to boot.

I do want to finish this, but I am incapable of just sitting down and learning 3D modeling. Doesn't help that there aren't any good 3D modeling programs for Linux.

 

[Luca]

Doesn't help that there aren't any good 3D modeling programs for Linux.

GitHub - cryinkfly/Autodesk-Fusion-360-for-Linux: This is a project, where I give you a way to use Autodesk Fusion 360 on Linux!

This is a project, where I give you a way to use Autodesk Fusion 360 on Linux! - cryinkfly/Autodesk-Fusion-360-for-Linux

github.com

The install is really jank but in my experience it works

Put your designs to work ... everywhere ... Ondsel | Ondsel

Put your designs to work...

ondsel.com

A little less annoying than FreeCAD but still annoying

OpenSCAD

openscad.org

And of course openscad for the hardcore

 

[Nold]

Awesome work so far, makes me want to tackle the wii-u myself O.O...

Doesn't help that there aren't any good 3D modeling programs for Linux.

I feel ya, best thing i can offer is this: https://github.com/cryinkfly/Autodesk-Fusion-360-for-Linux

I had it running for quite a while until Autodesk broke it, idk if it's back to working for me but it's worth a shot.

 

[Lazr]

I appreciate these suggestions, but for FreeCAD I have only heard bad things about it.
Never heard of OpenSCAD. Maybe its worth checking out..?
As for Fusion, I've tried using that script for it, but I kept running out of space during the install. I should be able to put it on my HDD, which has lots more space.

I was going to use Onshape, but I suppose I can try installing Fusion again.