Hi all,
I'm running some experiments with the EK-RA6M5, attempting to read data from a USB 3.0 flash drive as quickly as possible. Here is my test setup:
I know that there's some protocol overhead and I'd never get a full 480, but 55 is less than 1/8th of the possible bandwidth!
I've read that FIFO size can play a role here, but can't figure out a way to configure it. It appears to be fixed at 512 in DMA mode, and 2048 in Continuous Transfer Mode (which takes the same or longer amount of time).
Is there some other bottleneck in the RA series that prevents full use of the HS bandwidth? Is there something dumb I'm overlooking? Unfortunately I don't have a USB sniffer so I can't provide any traces there, but any suggestions welcome. Thank you.
Hello Raphael,
Thanks for your request.
Could you please specify , if you are using any of our already provided github examples? Or a custom project?
Thanks and Regards,
IK
Hi IK,
I'm using the USBX_hmsc_ek_ra6m5_ep project, with the use of ux_host_class_storage_media_read instead of usbx_hmsc_operation, and timing using tx_time_get.