0

In datasheet at FT2232H chip:

The FT2232H has the following advanced functions:

  • ...

  • Single channel synchronous FIFO mode for transfers upto 40 Mbytes/Sec.

  • ...

I configure device with the recommended settings, according to FTDI application note AN_130 and other.

When in the program I call FT_SetBaudRate() function, maximum value of baud rate is 8_000_000, which provides the max speed 8 Mbytes/sec (parallel interface). Attempts to set the upper baud rate are throw error FT_INVALID_BAUD_RATE.

So, how can be reached the maximum promised speed?

Aave
  • 103
  • 3
  • You've not specified what host side software you are interacting with this through, it is entirely possible the limitation is being applied there, and not in the hardware. As to why it might exist, that could be anything from not differentiating between models to no one needed it to the software authors having realized the hardware couldn't do that (or at least not well) in practice. If the software you are using is open source you should read it. In either case, you should read the data sheet or programmers manual and see how the desired baud would be encoded as a command. – Chris Stratton Jul 11 '19 at 14:46
  • @ChrisStratton I develop my own software, based on the FTDI's free to use libraries and follow manuals directly. – Aave Jul 11 '19 at 15:50
  • You seem to be missing the point: you should examine the FTDI libraries and see if the limitation is being imposed there. And you should examine the FTDI documentation for *directly* interacting with the chip, and see what actual USB commands would need to be sent. You might also compare various open source libraries to FTDI's offering. – Chris Stratton Jul 11 '19 at 17:45
  • @ChrisStratton ok, you're right. I just was sure that someone did it already. – Aave Jul 11 '19 at 18:03
  • See the comment to this answer: https://electronics.stackexchange.com/a/4338/2246 – Chris Stratton Jul 11 '19 at 18:09
  • how sure are you that one baud isnt actual 8bit in this mode? – Christian B. Jul 12 '19 at 09:04
  • And maybe check out the example(s) from libFTDI for some input regarding max performance: http://developer.intra2net.com/git/?p=libftdi;a=blob_plain;f=examples/stream_test.c;hb=d5c1622a2ff0c722c0dc59533748489b45774e55 – Christian B. Jul 12 '19 at 09:19
  • Ahh actually I think there exists a missconcept: the CLKOUT is fixed at 60 MHz if I am not mistaken. So you cannot actually change the clock speed by setting a baudrate which means the baud rate setting is actually meaningless in this mode. – Christian B. Jul 12 '19 at 09:27
  • @ChristianB. According to FTDI's documentation, in sync FIFO mode speed is changable, although clock is 60 MHz constantly. – Aave Jul 13 '19 at 10:18
  • @Aave can you give me a link + page as I cannot find that. Furthermore did you look into the linked example(s)? – Christian B. Jul 13 '19 at 13:31
  • @ChristianB. http://www.ftdichip.com/Support/Documents/AppNotes/AN_130_FT2232H_Used_In_FT245%20Synchronous%20FIFO%20Mode.pdf – Aave Jul 14 '19 at 13:41
  • @ChristianB. In this document there are few examples presented too. – Aave Jul 14 '19 at 13:42
  • @Aave exactly and there is no word of "changeable baud rate" or even one line in the examples for "setting baud rate". Only timing relevant setting is the latency setting which is only relevant for the USB layer. The effective speed is limited by the clock but by leaving out read and write opportunities one can "slow" down. But the clock is fixed. So why do you think that the baud rate is relevant at all? Or did I miss an important piece of info (that is why I asked not only for the doc but for a page/line as well)? – Christian B. Jul 15 '19 at 17:57
  • @ChristianB. I saw 60 mhz CLKOUT on the oscilloscope's screen, yes. But if speed is unchengable, why do datasheet tells about 40 mbytes per second? Is it just meaning that with technical losses in the line data transmission average speed is about 40 mb/sec? – Aave Jul 16 '19 at 04:27
  • @ChristianB. I've understood my missconcept. Battle for the speed is should be at the upper logical layer. Thank you very much! – Aave Jul 16 '19 at 04:37

0 Answers0