nRf51822 SPI Slave not working properly

RE: nRf51822 SPI Slave not working properly

Daniel Tralamazza — Sun, 08 Feb 2015 12:22:35 GMT

I ran into similar issues, the 7100 ns delay is the key here. Make sure you have the proper delay between chip select and clock. Ofc both need the same mode, in my case mode1.

RE: nRf51822 SPI Slave not working properly

Stefan Birnir Sverrisson — Thu, 05 Feb 2015 12:03:45 GMT

Thank you for you feedback Brajesh. That is very interesting. Perhaps the CPOL and/or CPHA configuration is different on nRF51 and MSP430 for the different modes? But, I am glad you found a solution.

RE: nRf51822 SPI Slave not working properly

Brajesh — Tue, 03 Feb 2015 04:20:39 GMT

Thanks Stefan, Iv got it solved by a workaround. I checked the received data by using UART and strangely it 1 but left shifted data of what is actually sent by MSP(SPI Master). I'v checked probably there is some issue with TI's MSP regarding the same. And it worked correctly when i changed SPI Mode of Slave (BLE) from Mode_0 to Mode_1 (while SPI Master is still operating on Mode_0). Ideally both should have same but in may case it worked when i keep two different modes here. Do you have a proper reason for this, why it worked like that ?

Thanks :)

Brajesh

RE: nRf51822 SPI Slave not working properly

Stefan Birnir Sverrisson — Mon, 02 Feb 2015 10:44:36 GMT

What nRF51 hardware/kit are you using and what nRF51 SDK version?
Make sure the settings are the same for CPHA and CPOL for the nRF51 and MSP430.
Try to add a delay on the MSP430 side after setting CSN line low but before clocking out data.
The nRF51 SPI slave needs 7.1 us delay.
What happens if you change the SPI clock speed of the SPI master. Does that change anything? The nRF51 SPI slave can only handle 2Mbps speed.
If you monitor the MOSI line with a logical analyzer or scope, do you see faulty data there or the correct data?