I found this in list archive:
> Queue as much as possible in one request. Read as large as possible
> chunks.
>
> Be sure the FT232H gets recognized as a USB-2.0 device on your PC.
> If you PC setup isn't right and the device is only connected to a
> " Linux Foundation 1.1 root hub" as seen with lsusb, only 1 mS
> frames will be used as opposed 125 us microframes on a 2.0 hub.
I made experiments with USB 1.1 and USB 2.0.
I have 4Mbyte EEPROM chip connected to 4232H chip (using SPI).
I use flashrom program to manage this chip (this program work over libftdi).
I did two sets tests - via USB 2.0 hub and via USB 1.1 hub.
results:
1. reading
I patch flashrom a little bit, now use large blocks (64Kbyte) for reading.
Reading of whole 4Mbytes took 5 seconds on USB 1.1 and 1.7 seconds on
USB 2.0.
So ftdi can use USB 2.0 (and provide transfer with speed over 12Bbit/s).
2. writing
Write can be done only by smal chunks (256 bytes), every chunck write
produce few writing requests and one reading request.
On USB 1.1 flashrom need 65 seconds for writing whole 4Mbytes.
This is exactly 4 USB frames for each chunk, which looks reasonable.
With USB 2.0, I expected major speed boost.
But with USB 2.0 I got same 65 seconds.
What happens?
ftdi chip does not use 125 us microframes?
--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe@xxxxxxxxxxxxxxxxxxxxxxx
| 