You're right! I put the missing .1uF in the breadboard and it started working with the Nano 50cm away. On the PCB, I added a shunt from pin 19 to the GND in the 0.1uF capacitor, making it a lot near to the 7221 and it started working too. Wow!!! The hours I wasted with this. You guys are great!!!
This problem took > 20 posts before the problem was found and solved. The reason for that is because someone else's schematics were posted in the original post, not your schematics. They did not accurately represent your circuit. This wasn't simply unhelpful, it was positively misleading.
@PerryBebbington Hi Perry, I wanted to suggest a change to the forum guide related to this, but the feedback topic is closed.
I did post a schematic in the first post. I was not from someone else. I created that schematic only to post here since I didn't have one, and it represents what I had in the PCB. I didn't had a better one at the time, but I created one as soon as possible and attached it in the response to kolaha.
And after all, the problem was not on the schematic. I was on the lengthy track in the PCB. The schematic didn't help at all. The PCB drawing did, so what did I do wrong?
Ok, sorry about that, but you mentioned a Mega and Nanos, yet the schematic shows an Uno. That's why I assumed it was a schematic you copied from somewhere.
The schematic showed bypass capacitors, but they were missing in your Breadboard and PCB (although your PCB had a space for the bypass cap, it was not populated).
Many beginners don't understand the importance of bypass caps, and often they find the circuit seems to work fine without them, so they assume they are optional. Until, one day, weird things happen!
There are a lot of switching on the SPI Bus/Port lines during data transaction, which causes ripples to appear on the Vcc-point of the MAX7219 chip. The purpose of the bypass capacitor is to absorb/kill those ripples; otherwise, the random dip states of the Vcc-point might prevent the chip from normal functioning.
