Approximately how long does it take the G30 to turn on and start running code when deployed and not debugging? I checked the STM data sheet but there is at least 5 startup times listed and I am not sure which one applies to my situation.
I am having some issues with the G30 being the slowest IC to power up compared to the display as well as some coprocessors. The issues only arise when cycling the power and not when I just restart the code with the debugger. I think the G30 powering up late might be doing some weird stuff to the COM buffer of the PIC connected to it. I need to design a sequencer so that the G30 is the first IC to power up. I am 90% sure this will solve the last of my bugs.
You could get a relatively precise measurement by deploying a program to toggle an output then measure that and the power pin on a scope to get the delay before toggle.
Ok thank you, my theory is correct - I added a 100ms RC time delay to the reset pin of the co-processor starting up before the G30 and it has all been running flawlessly. I had previously tried ~20 ms and it didn’t solve the issue.
Issues like you are seeing are generally down to the rise time on the power supply. Big capacitors on your board can cause this too. Your fix with the timing on the reset is the correct way to sort this out. I had the same issue with a G120 based design. A change to the timing fixed the issue permanently.
Yes I have a 220 uF cap on each board because the motor and driver can both draw ~7A (definitely reducing the caps in the final version, that was just what I had access to and was cheap). The motor was sagging the logic lines just running a few hundred mA…probably just from my poor running of traces \ not big enough ground plane for the motor driver. That was my first PCB ever so I didn’t for-see such issues. My second and third came out much better.
I still need to sequence the reset on the display…having occasional intermittency on power-up with that component as well. The G30 is significantly slower booting up.