I am new to SD SPI, using ELM_CHAN FatFs with fsl_sdspi, interfacing with a MicroSD card. Can anyone help figure out why I'm getting kStatus_SDSPI_ResponseError.
Traceback:
K8x_FatFS_demo JLink Debug [GDB SEGGER Interface Debugging]
K8x_FatFS_demo.axf
Thread #1 57005 (Suspended : Breakpoint)
SDSPI_SendCommand() at fsl_sdspi.c:373 0x7d8
SDSPI_GoIdle() at fsl_sdspi.c:425 0x8d0
SDSPI_Init() at fsl_sdspi.c:1,091 0x1674
sdspi_disk_initialize() at fsl_sdspi_disk.c:171 0x198e
disk_initialize() at diskio.c:127 0x5110
main() at FatFSdemo.c:50 0x528
arm-none-eabi-gdb (8.3.0.20190703)
Code Excerpts:
/* fatfs_0.11a/src/fsl_sdpi_disk/fsl_sdspi_disk.c */
:
:
status_t spi_exchange(uint8_t *in, uint8_t *out, uint32_t size)
{
dspi_transfer_t masterTransfer;
masterTransfer.txData = in;
masterTransfer.rxData = out;
masterTransfer.dataSize = size;
masterTransfer.configFlags = (kDSPI_MasterCtar0 | DSPI_MASTER_PCS_TRANSFER | kDSPI_MasterPcsContinuous);
return DSPI_MasterTransferBlocking((SPI_Type *)BOARD_SDSPI_SPI_BASE, &masterTransfer);
}
:
:
void sdspi_host_init(void)
{
/* Initializes timer to make host's callback function "timer_get_current_milliseconds" workable. */
timer_init();
/* Saves host state and callback. */
g_host.busBaudRate = DSPI_BUS_BAUDRATE;
g_host.setFrequency = spi_set_frequency;
g_host.exchange = spi_exchange;
g_host.getCurrentMilliseconds = timer_get_current_milliseconds;
/* Saves card state. */
g_card.host = &g_host;
}
:
:/* sdmmc_2.1.1/src/fsl_sdspi.c */
:
:
static status_t SDSPI_SendCommand(sdspi_host_t *host, sdspi_command_t *command, uint32_t timeout)
:
:
/* Wait for the response coming, the left most bit which is transfered first in first response byte is 0 */
for (i = 0U; i < 9U; i++)
{
if (kStatus_Success != host->exchange(&timingByte, &response, 1U))
{
return kStatus_SDSPI_ExchangeFailed;
}
/* Check if response 0 coming. */
if (!(response & 0x80U))
{
break;
}
}
if (response & 0x80U) /* Max index byte is high means response comming. */
{
return kStatus_SDSPI_ResponseError; /* <<< Breakpoint Set Here <<< */
}
:
:Inspecting variable: response has value 0xFF. MSB remained 1 for the 9 exchange cycles, so clearly it never hit the break.
Thanks,
Denis
Thanks for that Sabina.
It looks like it may be a hardware design fault. I immediately noticed that the circuit design in the document that you referenced is quite different. It has PULL-DOWN resistors on DAT1 and DAT2 signals...
...while manufacturer's specify pull-ups, like this one from Kingston:
Mine has PULL-UP resistors on DAT1 and DAT2 only, and no pull-ups on any on the SPI signals at all, and so relies on the K8I's internal pad pull-ups instead. I also notice from your reference document that the pull-ups/downs are quite strong – 10KΩ, while my pull-ups are much weaker – 100KΩ for the external pull-ups, and I think the K8x internal pull-ups are somewhere in-between – 47KΩ.
This is what happens electrically;
With and SD card inserted, if I apply the K8x pad pull-ups, with no VDD present, then the card keeps all SPI signals low. When VDD is applied, the card releases CS, CK and MISO, and they immediately go high, but MOSI appears to remain driven low by the SD card.
Thanks,
Denis
Looks like it was a hardware issue - the internal pull-ups may not be strong enough. I soldered some 10K pull-ups on:
- CMD/DI - SPI_MOSI
- DAT0/DO - SPI_MISO
- DAT3/CS - SPI_CS.
- DAT1 & DAT2
Now, all the pins, including MISO, are at the correct levels.
Okay, I have this (uses SPI2, like the custom board):
