A major step

By Pierre Buffo July 20, 2018 Discover Guido project

Dear CHIC community,

Yesterday was quite an important day for us, aside from having visited Tencent we also had to finish our prototype.

Clearly having a full working prototype wasn’t realistic. Choices were made, we wanted to have at least a system that could work on battery, have a working (and non crashing) firmware as well as a led matrix fully completed.

While Mehmed and Gabriel went at the x.factory pcb factory to assemble boards I had to deal with memory issues in freeRTOS. Joanne had to mold some silicon, Romain and Vignesh had to do business card and updates on the user manual.

The PCB that had just been received two days ago had never been reviewed by anyone else that Mehmed which made it. Wich means that there was a high risk of a fail somewhere. Nevertheless a board was built. It is only after that that M.Beuchat made an inspection on the schematics and PCB, but had nothing to say against them. A good point.

During that same amount of time I managed to find the cause of our endless hard faults issues that were caused by bad stack size allocation.

At this point I took the freshly assembled PCB and tried to program it. And it worked, from the first time ! We thought that we just broke the primary rule of a PCB maker “the first version will never work”, but even is the system seemed to work properly a hidden mistake suddenly make surface. We use a BGA chip type microcontroller, which means that pads are located under the chip. Because of the high density, some chips pads were directly putted above vias. Vias that we forgot to ask to be filled. And so after tests we realize that almost all of these pins above vias were faulty. We then choosed to re-purpose some of these pads that aren’t above vias to drive our led Matrix. Some wires after the trick was done and worked.

In less than 24 hours we went from nothing to a first non complete but functional prototype.