On 07/24/2012 07:08 AM, Kent A. Reed wrote: > Any thoughts? From what I've seen looking into USB 2.0 device programming, the transfers are too quick for most (maybe all) microcontrollers to keep up with in software, so you are generally tied to DMA transfers on the microcontroller side. FTDI has some chips that eliminate the DMA issue for the programmer, but they may do DMA internally.
There is a mode called 'isosynchronus' where a peripheral negotiates a certain amount of transfer for every time USB transfers data. This mode can be used for more reliable data streams. I suspect that a high bandwidth peripheral that has a USB port all to itself will not be as problematic as one that contends with a mouse and keyboard for the port. I also suspect that the USB connection has a certain period between communication events. A base period that gets lots of real time errors from contention with USB*/may/* have much less problems with both faster and slower base periods. Perhaps when documenting what works, and what doesn't, it would make sense to keep track of: 1: USB 1.1 or USB 2.0 2: What is the device? 3: Does it have the port to itself? 4: Is a hub being used? A hub may complicate the timing. 5: The appropriate part of the log files. ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
