To display the FileTime correctly, I am considering whether we should divide the result (using 80MHz clock) by 80000 or by 40000.ĭividing it by 40000 gives a result closer to what I record using a normal stopwatch. If (frc2 || nWrite < nRead) break /* error or disk full */įileTime = (ReadCoreTimer()-StartCount)/40000L ĭBPRINTF("\nFile Transfer Time: %ld",FileTime) If (frc1 || nRead = 0) break /* error or eof */įrc2=f_write(&fpdst, &ReadBuffer, nRead, &nWrite ) I also made use of the readcoretimer() function for a more accurate reading.įrc1=f_read(&fpsrc, &ReadBuffer, sizeof(ReadBuffer), &nRead) I tried timing the transfer using a Normal Stopwatch to read the transfer rate. I am using pic32mx with FatFS and I am getting the same transfer rate. Second, How do you measure your transfer Speed? By this you mean kilobits per sec? (400Kbps = 50 KBps).
I would like to clarify first on your unti "Kbps". Res = f_write(&fileptrdst, readbuffer, br, &bw) Res = f_read(&fileptrsrc,readbuffer, sizeof(readbuffer), &br) DBPRINTF("LSEEK returns: %d\r\n", res) PrintString( " OPEN ERROR: DESTINATION \r\n" ) Res = f_open(&fileptrdst, param2, FA_WRITE|FA_CREATE_ALWAYS) ĭBPRINTF("DST open returns: %d\r\n", res) PrintString( "\r\n OPEN ERROR: SOURCE \r\n" ) Res = f_open(&fileptrsrc, param1, FA_READ|FA_OPEN_EXISTING) ĭBPRINTF("SRC open returns: %d\r\n", res) However, I am having a slow throughput on my copy command. I revised the demo to implement Chan's Fatfs ( ). I revised the USB Host-usb logger demo in the sample demos provided by microchip.
It can be used as GUI or command line tool. And there is more: FastCopy is free to use and even available in source code. I am doing a project that uses pic32 as host to a USB drive. FastCopy does everything right Explorer does wrong: It is fast (see my tests below) It supports long paths (longer than MAXPATH, 260 characters) It does not use the Windows file system cache.
0 kommentar(er)
