Posted on May 17, 2011 at 12:21Hello STOne-32, that does help thanks. Since you ask about improvements, thats a tough one - improving near perfection;-)
1)Faster system boot mode comms. What is the shortest realistic time one can expect to be able to program 128k flash via the system boot mode? Some of the delay would be flash erase/write times {[40ms erase + (256 x 40us write)]x 128 blocks of 1k = 6.4s flash delay}, the rest comms: {128 x 1024 x 10bit / 115200} = 11.4s. So a program time of 20s should be achievable. Is there room for improvement on the comms side? I could only use comms at 19200 baud. The auto baud did not work well for me.(I could program 118k in 25s on an STR7 for comparison). Use of the USB interface could be good for faster speed.
2) Uart CRCs in hardware. The SPI can calculate CRC's with hardware but the UARTs can't - it would be nice if the uarts could, or there was an application note that showed how to borrow that capability from an SPI for the use of a UART (it was said to possible at a seminar). I can see why it may be fiddly with start stop bits, parity, etc.
3) UART Auto baud. Modems can adjust to baud of serial port. It would be good if this was a viable option without data loss. The Coldfire 5272 has this facility but it not perfect.
4) UART FIFO - or the ability for UART-DMA to generate interrupt when *ANY* data arrives in the receive buffer rather than *ONLY* when half full or totally full. This would allow us to make full use of DMA in the full knowledge we'd be notified when ANY data arrives without having to poll the buffer and without having to have a buffer of 2 bytes!
Admitedly U(S)ARTs are regarded by many as a bit 'old hat' but the hardware is simple and cheap, and these features would be VERY nice for some of us. We can add value AND reduce costs.
Anyway - back to the question, what are the differences between revision B and Z?
[ This message was edited by: bobz on 16-01-2008 18:43 ]
