Smartwatch part 1

I recently received the STM32F469 Discovery Board! It features a MiPi DSI Display with capacitive touch an SRAM for buffering Frames, a micro sd card slot, audio and so on... Because it's the same processor that I want to use in my new design and because there is really good support with exampels, I plan to complete almost all Software work on this platform before starting to buildup my own hardware. Just to make sure I dont't miss something important.

I've already successfully deployed TochGFX by Draupner Graphics, which blew my socks off! They seem to take their motto 'outstanding graphics and smooth animations with minimal resource and power consumption' really seriosly! You should check it out if you plan a project with a  GUI.

Lenovo G70-70 updates

Finally I can say for sure, that you can expand your storage capacity by adding an odd board to your G70-70! I searched quite a while for a good price to performance SSD and finally bought a really fast (500+mb/s) one made by Transcend. It was due to Amazons prime day also really cheap.

All that's left to say is that there is unfortunately a small issue with the covering of the odd slot. The original odd dummy has a unconventional round edge which is near to impossible to find on the web and I'm still searching for the best solution. 

Beelink XII repair

Untortunatly due to the exam phase at my University I really have no time to work on my projects. But this weekend I decided to try to repair a broken TV box. The box is an Allwinner H3 based single chip device (Beelink XII), running Android 4.4.2.. 

One day the box wasn't powering up at all and because it was in a working condition on last shutdown I thought about a software issue. I started with trying to flash the latest stock firmware but no success. In addition I got no response by the Software on connect of the device. So the device was unreachable bricked. 

After some reasearch I found the linux-sunxi website which offers really good reference information about almost all Allwinner SOCs. And as always If you're trying to fix or reverse engineer something, spending some time on a good research for informations is always well spend time. 

With the new knowledge about these devices I now soldered some wires to the serial interface pads on the boxes mainboard. And after powering up the device I got some output about the boot process.

So far so good So at least the processor wasn't dead. Also boot0 and uBoot seemed to be in a working condition. This also means that the nand flash was at least not entirely broken. Good news for the further process. 

I now tried to build a bootable SDCard for further digging into the boot issue from the nand. But after serveral hours of image building compiling and copying files I stopped the repair session for that night (It was already 4 o'clock in the morning). 

Later that day I thought about searching for possible issues of the box in the serial output. And there was indeed a really strange line which I missed out before (last line in first screenshot). Dram value=1. I tought either that is just an index or it's the actual size of the RAM. If it's the actual value of the RAM there was something entirely wrong happening. 

And because the box was already broken I fired up my reflow soldering station and started to preheat the Pcb. After that I reflowed the RAM modules and let everything cool down to room temperature.

Now the final test, will it work at all after that?

YES it does! Now the output of the console showed up the correct 1Gib capacity and also the blue light came back on and the device was booting properly. 

So just that tiny little line in the serial output pointed me into the right direction and saved another device from being tossed out into the garbage. I hope I can motivate you to repair some gear yourself and trust me the feeling when you got it back up is just outstanding. In addition you always lern a lot about the devices and how they work. 

PS: After several checks I can say that also the overall speed has also increased by at least 15%. Therefore a production fault can be a possible reason why the box stopped working.