nrf52 spi reads extra zero

[Matts]

Hi.

I'm using nrf52 as slave for another device. I'm reading test string from the nrf buffer and I can see extra zero (rest of the content is then offset by one -> corruption). However I cannot see this extra clock in the bus when I check it with scope.

The data should be:

0x6a (0110 1010)
0x65 (0110 0101)
0x73 (0111 0011)

however in the buffer it's

0x35 (0011 0101)
0x32 (0011 0010)
0xb9 (1011 1001)

As you can see there's extra 0 and offset. In below scope picture there's only one zero.

nrf52 setup as follows:

• SPIS0 enabled and used with default config
• CS used
• SPIS clock mode is 0 (default, so active high, rising edge)

I set both buffers at the same time like this: APP_ERROR_CHECK(nrf_drv_spis_buffers_set(&spis, (uint8_t *)m_tx_buf, strlen(m_tx_buf), m_rx_buf, SPI_RX_BUF_SIZE));

buffer should be in heap: static char m_tx_buf[] = "nrf_test\r\n"; /**< TX buffer. */

We can see the same phenomenon on two different master devices from different manufacturers.

EDIT: corrected stuff removed.

[Matts]

And to elaborate more, The first zero should be there. The extra zero in the output (not visible in the bus) is the one I'm worried. More urgent is the missing reply even tough there's a clock available from the master.