Using highspeed spi on RT595-evk

[kkuryllo]

I am trying to make use of the high-speed SPI on the RT595-evk board.
The documentation for the board (https://www.nxp.com/docs/en/data-sheet/IMXRT500EC.pdf -1.7.5 High-Speed SPI interface (Flexcomm interface 14)) states that it supports use of the SPI up to 50 Mbps.
However, I find that I can not even use it at 5 Mbps, because the drive rate of the pins is not sufficient to modulate between high and low voltages at this transfer rate. This issues occurs with both the PMOD connector (J36) and the HS_SPI connector (JP26).

I have attached an image of the logic analyzer output. In channel 3 the clock voltage changes too slowly. It looks like there could be some capacitance on the bus, however I have not attached any hardware to the board or reconfigure it.

I have similar issues when I try the provided SPI driver examples in the SDK at this baudrate.

How should I configure the board to make use of the High Speed SPI?
I have jumper J32 set already.

Additionally, in the manual it mentions a Jumper -J10- that does not appear to exist:
"Install J10 when using the SPI interface at connector J36 and/or FC0 UART
at J16. This disables the Link2 SPI and UART (VCOM/bridge) connections."
Where can I find this Jumper? It does not appear to be labeled on the board.

[Omar_Anguiano · Accepted solution]

Hello
Hope you are well.

I suggest you use the "spi_polling_b2b_transfer" so we can be sure that the SPI runs in HS mode. For this, we need to make sure that the board jumpers are in the default state and connect the jumper J32.
The HS SPI pins in this example are in the J36 connector, it is important to make sure that the macro USE_HS_SPI is set to 1 and the variable userConfig.baudRate_Bps is set to the desired baud rate.
I will also try to replicate this issue on my side.

If you have more questions do not hesitate to ask me.
Best regards,
Omar

[kkuryllo]

Hi Omar.

I have managed to resolve the issue on my end. It turned out that there was no issues with the signals after all. The logic analyzer that I was using was aliasing the output signal, making the signal traces appear more gradual than they really are. I was able to acquire a oscilloscope to test the lines, and with that was able to verify the signal output is as it should be.

Thank you for the quick response!