Worklog Toshiiba

[vikMKW]

Background
Wii laptops were popular among early modders and beginners, like myself, but all attempts I've perused were incomplete, dated, or were not portable (i.e. tethered to wall power).

I don’t stand a chance against this year's unique ideas and skilled modders, but I hope to take the title of best Wii Laptop. In tribute to the Toshiba laptop I found in the trash chute that donated its LCD, I went with the obligatory, cleverly placed double “i” in the name.

Goal
Build a battery powered console with a large display for multiplayer Wii & Gamecube games (mostly Mario Party, Melee, & Project+).

Means of transporting consoles are fairly discrete; it's either a pocket or a backpack. As long as I can fit the build in a backpack, I am within my loose definition of a portable. I see this being played on long car rides, game nights with limited TVs, and similar situations.

Primary Features
Features that were conceived prior to R&D. Laying out the features serves to keep me focused on the governing ideas to avoid veering off track or becoming distracted.

• Uninterrupted charge-and-play.
• 3S2P Li-Ion cell configuration.
• Bidirectional active cell balancing.
• Sleep when closed.
• Graphic fuel gauge & power monitor.
• Standalone monitor mode (for RPi/Chromecast or secondary monitor).

Secondary Features.
Features that manifest in the R&D stage, mostly due to other parts or design preferences.

• Aluminum sheet metal frame.
• Wii native VGA.
• HDMI in & DVI in ports (for RPi/Chromecast or secondary monitor).
• USB power port (for RPi/Chromecast).

High Level Diagram
At the cost of being redundant, laying out a block diagram should keep me focused because it’s too easy to diverge from the plan and end up with extra, but incomplete features.

I just finished defining and the project, so I can get to R&D now. This also means I don't have any progress pics just yet. I'll leave you guys with my ridiculous 17" LCD.

 

[bxepfhiz1c390]

How and where did you find a driver board for that display?

 

[vikMKW]

How and where did you find a driver board for that display?

Search for "LCD driver" on ebay and it should be the first result. Contact the seller and ask if it's compatible with your screen's serial number.

 

[vikMKW]

The system booted after the U10 relocation, so I trimmed the Wii today. Took no more than 10 mins thanks to power tools. It's too easy to remove material though.

Big Boy Tools:


USB ports and A/V connector weren't trimmed yet. USB might be relocated. My current setup demands that the A/V connector remains attached until the VGA mod is complete.

Rule of thumb is if you think you're done sanding, sand some more. Can I get a second opinion about whether I should sand with high grit (~1000) sandpaper? The belt sander had a coarse belt (probably 100 grit), but the edges look clean.

 

[cheese]

T O O L S

I'd sand with a higher grit sandpaper, but you won't need to do too much.

 

[Jonny]

Those angles are Crisp! Nice trim!

 

[Lucascito_03]

Nice! I can't wait to see it!

 

[vikMKW]

Remembered I had one of those clip on macro lenses for my phone, so I took a look at the traces to make sure there were no shorts. It's neat that all 4 layers are visible. All edges look as follows straight off the coarse belt sander. I suspect high grit sanding won't be necessary.

The resistances are as follows:

Excel sheet source forwarded by Cheese.

 

[cheese]

https://bitbuilt.net/forums/index.php?threads/wii-trim-resistances.2602/

 

[vikMKW]

https://bitbuilt.net/forums/index.php?threads/wii-trim-resistances.2602/

Shoot, I missed that. Should've looked harder. Thank you for the guidance.

 

[vikMKW]

It's been hard for me to find enough time to make a dent in the build, but I've been working on it. I've used a blend of EasyEDA and Eagle in the past, but decided to commit to KiCAD. Instead of winging it to begin immediately, I opted to take my time and learn it by following instructional videos and reading. It was a good investment.

The power distribution board was top priority because the Wii can't be turned on without it. I have a connector for a switch that will pull the INHIBIT pins on the PTH regs responsible for the Wii voltages to GND. This way the LCD and USB 5V can remain on without the wii for monitor mode.

I uploaded JST-XH Right Angle connectors and PTH08080W footprints I made to Github. Despite the Osh Park images, the order was placed with JLC PCB as they had better prices. This is one of two boards that I'll purchase from a manufacturer. Unless I run into complications, the remaining boards will be produced at home.

The dual USB ports were replaced with a single port. For serviceability, future upgrades, and troubleshooting, I'm striving to build boards as modules, with connectors. Having ample space to work with helps. I've got some more wires to solder, but I'm excited with how well it's turning out.

Single sided copper clad (100mm x 70mm) was etched with ferric chloride. Paper templates are my favorite means of quick testing. I used the acetone toner transfer method. Despite the silkscreen not transferring perfectly, I'm very pleased with it. The exposed pins on the right angle connectors allowed me to solder them to the top side, which meant double sided copper clad wasn't necessary! The holes aligned marvelously and the dimensions were spot on.


The next board to be etched is an A/V board that will contain a DB15 (VGA) connector and a headphone jack. It will go into the empty space between the USB and the Wii's AV connector.

A board for the CG ports will be etched as well. It's a matter of making a footprint in KiCAD. Here's a drawing made in Fusion 360 to confirm my measurements.


The battery holder is only partially built up. It will be 2P3S, but 3S1P is enough for testing. To avoid massive headaches, brand new cells will be purchased to match voltages and capacities. Ignore the salvaged ones in the pic.

At some point I'll need to tackle the charging board too, which will be challenging for me. I'm reading up on passive and active ICs.

If anyone can steer me in the right direction, I will be grateful. The charge board will charge and have a powerpath IC for charge & play. At some point I'll need to get the CAD started. I want to keep the 3D printed parts to a minimum because I don't need them melting in a hot car. The frame will be bent aluminum sheets. Expect more board for the next update.

 

[vikMKW]

Small update today. Some time ago, I foolishly discarded the hinges for the screen when I tore apart the laptop. The replacement hinges came in, and are confirmed to be the correct fit. There are protruding screw posts that will need to be removed with a grinder. The power distribution board is now with my local courier, so I expect I'll be able to populate it this week.

The footprint for the GC ports is complete, and will be uploaded to my Github later tonight. I went ahead and laid out the board.

Etching was quick, but I had to resort to transferring with an iron. Acetone and IPA aren't working like they use to. I have a strong suspicion that it's something to do with the acetone I have, because I'm using the same 3:8 ratio with the same copper clad. 0.5mm permanent marker is a great tool to fill in voids that didn't transfer.

There were just over 40 holes and 8 slots to drill out, but it was worth it. Being able to bust out a board within a morning is valuable for prototyping and reaching deadlines .

No transfer silkscreen this time. Acetone transfer just wasn't having it. Having a few board now (and a paper cutout) inspired me to try to lay out what I've got.

Parts were placed on the back of the monitor to reference as a footprint. A good deal of parts are still not built up, so the extra space is promising to have. The thickest component is the 18650 with its holder. The speakers are just as tall, so I may opt to go with smaller ones.

There's something nostalgic about seeing single sided FR4 boards with through hole components and jumpers. I hope to finally be able to power up my wii for the next update!

 

[vikMKW]

The power distribution board came in. Built it up and confirmed all the output voltages are correct.

Only now did I realize I didn't include a 5V pin on the breakout board mounted to the wii, oops. I was so focused on the fact that the wii does not need 5V that I completely omitted it from the board. There was an incredibly convenient hole pattern near the USB that could fit a JST connector, so the workaround is surprisingly clean.

Inevitably, I've hit my first road block: there's no video output. There's some sag in each line, which I was concerned about, but I can always swap out the feedback resistors. Obtaining 1% tolerance resistors is not an issue. Before starting the trim, I made sure the U10 relocation worked and that the wii booted into PM w/o the wifi module.

I recorded a video showing the line voltages and what I see on the monitor. I'm not familiar with what voltages the Wii2HDMI requires, so there's a chance that could be the culprit. What should I check to troubleshoot?

Edit:
The problem persists, but at least my monitor registers some form of signal with this AV2HDMI converter. Not a solution, but may provide insight.

 

[cheese]

Have you tried holding reset to ground to see if it's a USB issue?

Completely unrelated but, NICE JOB using the PTHs in the intended way, this is probably the first project I've seen on a modding forum that actually uses them as TI wanted

 

[vikMKW]

Thank you! I made sure to read the datasheet first. I know I'd want people to read a document like that if I made one. To be fair, I'm working with ample room, so I don't need to cut corners (literally too) as much. People here are far more competent than I am, so I have no doubt that someone can do it right with enough space.

Jumped reset to GND, but still no video.

Jeff has been helping me on the side. The rails sag a little under load, so we opted to change the resistance to raise the voltage.

Still no video out after changing the resistors. The CPU and GPU are still gradually heating and don't immediately become hot, which keeps me hopeful.

 

[vikMKW]

Stayed a bit later after work today to examine the composite video signal.

I brought an unmoded, known good wii as a "gold" sample.

The scope wasn't configured well, but I think I got the info I needed.

I'm not familiar with NTSC, but it's evident that there's a front porch, back porch, & color burst. The difference is in the picture information. Since the signal is being initiated with a sync pulse, I want to say the AVE is doing it's job, but is not receiving valid data to output. Correct me if I'm wrong please! At this point, I need to make hypotheses to progress. I'm willing to probe around! Thoughts?

 

[cheese]

That's a black screen, you did softmod this wii before trimming right? If not, you'll need to wire bluetooth and wifi to get it to boot into the wii menu, and then install homebrew from there.

 

[vikMKW]

That's a black screen, you did softmod this wii before trimming right? If not, you'll need to wire bluetooth and wifi to get it to boot into the wii menu, and then install homebrew from there.

Yeah, it's got HBC and PM. I made sure to not proceed with the trim unless it successfully booted into PM without the wifi module and the U10 relocation was correct. BT is currently wired up.

I've got a spare 6 layer board I can use as a donor. Not to promote the shotgun troubleshooting approach, but some components I'll be able to swap.

 

[Reinierr1]

First of all, this looks very neat! I'm interested to see what it looks like when it's finished.

I do have a question though. I am in the process of starting a project and am wondering what your main reason is for a 3S2P battery setup. Does your screen require more than 5V, or is it just more manageable with 11.2V battery output? I am kind of in between going for 2S3P giving more juice as I have a screen that works on 5V.
Thanks

 

[vikMKW]

The LCD controller I use requires 12V in. The lowest voltage the LCD controller was able to run off was 6.0V before there wasn't enough power to keep the screen on, but as I dropped the voltage, the screen dimmed. Although I can theoretically use any 2S configuration, it'll require more current and the screen will be dim. This is why I'm sticking to 3S, as the operating voltage range (~9.5V to 12.6V) does not have as drastic of a performance drop.