SPI returning 0xFFF... (custom pcb) (Solved)

[Salim]

Hi,

@AndyA

The power supply is the correct voltage, all the VCC and GND pins are correctly connected and that the supply can source sufficient current.

-Yes, I thoroughly checked power delivery. Everything seems fine (voltage and current wise).

@AndyA

The SPI bus is wired correctly
The reset line is wired correctly (from memory you should be able to leave it unconnected, it should idle high but pulse low on startup)

-I also checked my wiring a thousand times.

@AndyA

Don’t trust the firmware that the SPI / control pins are all doing what you expect, check them with a scope

-Unfortunately, I don’t have a scope on hand at the moment. However, I did look at the SPI lines with a logic analyzer, and I saw nothing abnormal (clock looked fine, the packets were sent correctly to the DWM, the MISO line transitioned from low to high, etc.) (I wish I could take a screenshot, but I don’t have the logic analyzer with me at the moment).

Finally,

@AndyA

If you hold nCS low, hold MOSI low and toggle clock 40 times you should get 0xFF followed by the device ID on MISO.

I’m currently using the HAL libraries for SPI, so I don’t have precise control over timing. I did try this in my own way by using a buffer with 0x00s and a very long receiving buffer, but it doesn’t seem to work. I might try to implement this from scratch later.

Edit: With a delay of 1ms between nCS pull-down and the HAL_Transmit function, there seems to be a different behavior. Other transmission events where I didn’t use a delay only yielded 0xFF in the buffer.