T4.1 SPI bus hard crash - potential bug in SPI library?

[jensa]

@mjs513 The SPI mode difference is just since you're looking at https://github.com/adafruit/Adafruit_MAX31865 rather than https://github.com/adafruit/Adafruit-MAX31855-library ;-)
(The MAX31865 is for PT100 sensors, whereas MAX31855 is for Thermocouples)

[mjs513]

@mjs513 The SPI mode difference is just since you're looking at https://github.com/adafruit/Adafruit_MAX31865 rather than https://github.com/adafruit/Adafruit-MAX31855-library ;-)
(The MAX31865 is for PT100 sensors, whereas MAX31855 is for Thermocouples)

Yep you are right was looking at wrong lib for some reason - should have mine tomorrow by tomorrow night

[jensa]

I've been stuck on other projects for some days now, but my coworker found that changing clock and reducing speed was a good solution to keep things stable without removing transactions.

It really annoys me that I cannot reproduce the bug on a breadboard, so I'll purchase some more breakout boards and IC's to do more testing next week.

[KurtE]

I've been stuck on other projects for some days now, but my coworker found that changing clock and reducing speed was a good solution to keep things stable without removing transactions.

It really annoys me that I cannot reproduce the bug on a breadboard, so I'll purchase some more breakout boards and IC's to do more testing next week.

Question that I have had with this, is, are you using the exact same setup for breadboard versus your board? Or you using someone else's breakout board or your own with same
CAPS, Resistors, ??? connected to them as with your board?

[jensa]

Question that I have had with this, is, are you using the exact same setup for breadboard versus your board? Or you using someone else's breakout board or your own with same
CAPS, Resistors, ??? connected to them as with your board?

The board is designed by me based on Paul's schematic. You can see the layout, schematics and BOM on that github repository. I now have 11 MAX31855 running successfully on breadboards with an original T4.1. They will also run without any issues when connected to the board based on the T4.1 Kicad files mentioned. It's only on the larger and more complex PCB that the error occurs. The board has a T4.1 that controls these 11 MAX31855's, 3 ADS124S08 and 1 W25Q80DVSNIG via SPI. Apart from SPI, it also has 3 sensors via I2C.

The board is actually more complex than this, since there is also an embedded Raspberry Pi Compute Module 4 (HDMI, Ethernet, USB hub and lots of other things) driving the UI of the machine. This part is separated from the T4.1 on the PCB and has it's own PSU so it should not affect the T4.1 at all. The T4.1 is only connected via the USB hub on the Pi, so nothing on the Pi should cause the problem. I've messaged @guzu to have him test on his old board (as mentioned before in this thread).

My main problem that is that I cannot reproduce it with an original T4.1 and that the customer does not want me to spend more time on the issue since we have a workaround. I would rather find the root cause of the problem, but if I cannot reproduce it, it might not be an issue that is relevant to the Teensy community at large.

J