GBA-2 Go

After an almost year break due to covid *cough cough* yeah covid.

Nah, to be honest. Ordered the boards, but it had some design flaws. Noticed them before building the boards luckily, so redid the design and simplified it a little. Removed some functions and feature but also some minor changes;

• no more 2 types of LCD support, 
• no more original LCD support(unless mod board is installed, see test points)
• no power switch adapter board
• changed all part to be SMT assembly compatible with my current fab house

new deadline is to have a working version of the board before the end of the summer break.

wish me luck

(list will be updated with features that will remain in this version)

Well I kind off forgot this existed. Not that anyone is reading but hmm..

Anyway, I've finished the boards. Yeah should've told you, I swapped the AtTiny for a Greenpak. Why you ask? -Just because I can, and I wanted to play with them for a while now. 

So yeah the 2 PCBs have been ordered;

The Greenpak Adapter board, although this euhm yeah went kind of wrong. So I've reordered them

The GBA2 itself;

So yeah all that remains is waiting for the PCBs and getting the parts. then hopefully I've got myself a working GBA2

"Huh, he hasn't updated his project in a while. He must have dropped it"

"What happened? Did he just quite? Was it too hard for him?"

and so on..

Just kidding, anyway. 

I kind of went quiet because my school exam continued. It's taking up way more time than I expected. Just 3 more weeks than I can continue with the GBA2.

I promise I'll finish the GBA2 before the summer holiday starts. Atleast, that's my deadline.

It has been a while, the holiday came to an end so I had to work on my exam again. 

After giving a whole speech on keeping it as cheap as possible in my last log I noticed there was a rather pricey part on the BOM. The ADXL335 would cost €5,50 a piece, combining that with the Gyroscope I was planning on using (L3GD20HTR) which would cost another €3,37. After doing some research I found a replacement for the ADXL, the LIS2DE12TR, which would only cost €1,34. Then I would save ≈€4. 

While doing some research for my exam I came across a very interesting GitHub page Nintendo Switch Reverse Engineering, for my exam I was only interested in the way data was being sent from JoyCon > Switch and back. But for the GBA2 I was interested in what Gyroscope or Accelerometer they used, the LSM6DS3TR. Since it would only cost me €2,82 for the Gyroscope and Accelerometer, it would save me a whopping €6,05 a unit. So aside from being cheaper, saving me space, already being used so having a PoC(Proof of Concept), there's also already a library for it. Which would save me a lot of time. 

I know I know, it's only a minor update. But I should be able to finish the first version in the next couple of days

All that remains to be done is route the power path

Even though 90% of the schematic was complete there was still 1 part which wasn't complete. I've been pushing it forward ever since I started on the GBA2. Today was the day; I put on some good music, got myself a cup of coffee and got to work. And...

well it didn't really go as planned, I think I wrote dozens of notes and possibly even more schematics. My Spotify playlist was already nearing the end of its 2nd round. Before I decided to quit the idea of a DDS/PWM driven Charge pump. It would just use too much space.

So aside from

• DDS/PWM driven Charge Pump

There were the other options:

1. Just picking a random Step Up and hooking a simple charge pump inverter to it
2. a SIMO ic (Single Inductor Multiple Ouputs)
3. Make my own switching regulator and hooking a simple charge pump inverter to it

After reading the upper part of the log, you're probably wondering. 'What the heck was he trying to achieve?' good question, it's very simple. The 2 voltages specifically used by the original GBA LCDs. Hah! well if money grew on trees I wouldn't even have added support for it. 

You probably went on a re-read, if you're actually interested at least or didn't quiet get the whole story.

So I had 3 options left, looking back at it. Well to be honest the 3rd one was never gonna work. Let's be honest.. So the 1st and 2nd were the ones I was deciding between in the end. I drew both schematic and PCB design.

option 1

and option 2

I liked both options, both used almost the same amount of space and both used parts that were easy to get my hands on. But well.. option 2 costed a bit more. The unit price of option 1 was €0,29 for the SU and €0,40 for the other parts. While option 2 the IC alone already costed €2. 

So I decided to go with option 1 for the prototype, and if something were to go wrong with option 1; unstable voltage, noise, not being able to even reach the voltage etc. etc. Then I'll swap it with option 2 for the next revision

Decided to swap the pot-meter for a multi-direction slide switch

This would allow for an easier control of stuff that need to be changed in steps; volume, back light, frequency and more.

I though about adding a current source amplifier for the 34p (GBA SPs LCD) models to control the back light. But ended up with just a transistor and noted it down for a possible next revision. Freed up some IOs for IPS back light control and headphone button(you know the small ones halfway on the cord).

Aside form these minor updates the design is coming along well. Just need to change almost all the part reference designators:

Yeah.. I euh... no-one is perfect!

As I'm having 2 weeks of holiday I'll be uploading more pics and info about the project. I kind of hijacked someone's discord and have been posting updates there for the past months

Minor setback, I had to redo the PCB design. 

It just didn't quite fit. Also decided to make some changes:

• Analog Joystick
• DDS driven charge pump for 13.6v and -15v
• Swapped the Gyroscope for a cheaper one
• Added support for AA batteries
• Overall component placement