Today was my first experience with signal integrity. As Artur said, our day was devoted to soldering. After making the poles, we decided to make a short test. We used the same program we made to test the leds strips. On Artur’s hemisphere, everything went fine. All the commands were working well. With mine, it was more a random color generator. At first I thought it was a short circuit. After cutting and resoldering a few wires, I always had the same problem.
After a few experiments with a logical analyzer, I came to this conclusion. My strip was working only if I monitored the clock signal. Could you imagine that ?!?
Thanks to Alexis, who probed the board everywhere, the mystery was elucidated. Even if the SPI frequency was not high (10MHz), the wires brought an inductive effect which distorted the signal. This effect was compensated by the capacitive effect of the logic analyzer.
For our final system, we will use ribbon cables.