I am using an ADSP-BF533 in SPI Master Mode and am having trouble when I use DMA to move data. Specifically, no matter what I do, the SPI seems to continuously issue clocks after the DMA has completed. Am I doing something wrong?
When using the DMA modes, the SPI port (as the master) will continuously issue clocks even after the DMA completes, until the SPI port itself is disabled.
Hi Joseph Beauchemin,
Could you please let me know some recomended device(SD-Cards...) which can support this sort of operation(Mode) .
As most of the SPI device does not support such mode . Could you let me know the reason for such sort of design .
Could we be seeing some fixes in future Blackfin . Does Blackfin -Blackfin transfer happen with out any problem in this Mode .
I cannot answer the first question, as it is outside my area of expertise, but the issuance of extra clocks in DMA mode is something that we've been aware of for a long time. I cannot comment as to why it was chosen to be done this way, but Blackfin-to-Blackfin communication would only be affected if the slave remained enabled AND the master did not de-assert the slave select output. If the slave SPI device is enabled, then only the /SPISS input governs whether or not the SPI clocks in data during those extra pulses, otherwise the clock is ignored. Current plans for future SPI implementations include addressing this system-level problem.
Retrieving data ...