Serial performance - Prop <-> FT232R .... happy but...
Stephen Moraco
Posts: 303
Hi All,
·· I'm just starting week two of a dedicated "work on project full time" for two weeks·effort.
·· I haven't done this for a while and am definately enjoying myself while working
·· in prop assembly/spin.
·· The default FTDI ComPort drivers on Windows (in my case Windows XP) define comm
·· buad rates of [noparse][[/noparse]110·to·921,600].· The "fun" thing about these drivers is that when you
·· change this baud rate in, say, Hyperterm the FTDI chip is reconfigured to that
·· new baud rate.· [noparse][[/noparse]Was anybody expecting this? -ed.]··
·· Anyway,·since we do have control over the FTDI·baud rate·this·is also the means
·· to now get better serial rates thru this chip.· The FTDI chip also supports 1.5M baud,
·· 2.0M baud, and 3.0M baud.· The means to get windows drivers to ask for the 3 new
·· rates is described in http://www.ftdichip.com/Documents/AppNotes/AN232B-05_BaudRates.pdf
·· basically we decide to replace the configuration bits for three of the pre-defined baud
·· rates, uninstall and reinstall the drivers and you have the new rates.
·· Now you have the backgound, let's get to my findings/question...
·· I've just completed boosting the Serial Receive/Transmit for my application (PropCAN)
·· to a relaible 2Mbits/sec but I think this is the limit.· The FT232R can go to 3Mbit and
·· my transmit side can do bursts of 3Mbit per second· but the receive side effort of
·· receving bits, assembling into bytes gathered into longs, storing the longs into
·· Cog RAM seems to be just past the limit...
·· I just wanted to see if I'm being brain-dead here or this is really the limit.
·· I've reached my project design goal of >1.5Mbps but after staring at my
·· code for two days (amongst doing other things of course) I'm just not
·· seeing it.
·· Anyway... just curious and I'm learning in a more focused manner now
·· so learning what I'm missing here will help my next development to
·· finish up this project.
·· If you're interested in seeing how I'm doing this it's rather simple and
···easily reusable.· You can follow my thinking and progress at
·· http://propcandev.blogspot.com· After I get the code measured and
·· in final shape I'll post the transmit/receive and autobaud objects for
·· for ya'll to use.·· ;-)
·· Certainly, if you have successes in this area please reply to this post.
Regards,
Stephen, KZ0Q
--
http://propcan.moraco.us
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Post Edited (Stephen Moraco) : 3/31/2008 6:01:58 AM GMT
·· I'm just starting week two of a dedicated "work on project full time" for two weeks·effort.
·· I haven't done this for a while and am definately enjoying myself while working
·· in prop assembly/spin.
·· The default FTDI ComPort drivers on Windows (in my case Windows XP) define comm
·· buad rates of [noparse][[/noparse]110·to·921,600].· The "fun" thing about these drivers is that when you
·· change this baud rate in, say, Hyperterm the FTDI chip is reconfigured to that
·· new baud rate.· [noparse][[/noparse]Was anybody expecting this? -ed.]··
·· Anyway,·since we do have control over the FTDI·baud rate·this·is also the means
·· to now get better serial rates thru this chip.· The FTDI chip also supports 1.5M baud,
·· 2.0M baud, and 3.0M baud.· The means to get windows drivers to ask for the 3 new
·· rates is described in http://www.ftdichip.com/Documents/AppNotes/AN232B-05_BaudRates.pdf
·· basically we decide to replace the configuration bits for three of the pre-defined baud
·· rates, uninstall and reinstall the drivers and you have the new rates.
·· Now you have the backgound, let's get to my findings/question...
·· I've just completed boosting the Serial Receive/Transmit for my application (PropCAN)
·· to a relaible 2Mbits/sec but I think this is the limit.· The FT232R can go to 3Mbit and
·· my transmit side can do bursts of 3Mbit per second· but the receive side effort of
·· receving bits, assembling into bytes gathered into longs, storing the longs into
·· Cog RAM seems to be just past the limit...
·· I just wanted to see if I'm being brain-dead here or this is really the limit.
·· I've reached my project design goal of >1.5Mbps but after staring at my
·· code for two days (amongst doing other things of course) I'm just not
·· seeing it.
·· Anyway... just curious and I'm learning in a more focused manner now
·· so learning what I'm missing here will help my next development to
·· finish up this project.
·· If you're interested in seeing how I'm doing this it's rather simple and
···easily reusable.· You can follow my thinking and progress at
·· http://propcandev.blogspot.com· After I get the code measured and
·· in final shape I'll post the transmit/receive and autobaud objects for
·· for ya'll to use.·· ;-)
·· Certainly, if you have successes in this area please reply to this post.
Regards,
Stephen, KZ0Q
--
http://propcan.moraco.us
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Post Edited (Stephen Moraco) : 3/31/2008 6:01:58 AM GMT
Comments
there is a possability by synchronizing two or more cogs.
here is one thread about this
http://forums.parallax.com/showthread.php?p=697888
i think receiving serial data is more complicated than just sampling an IO-PIN.
But maybe this is a possability
best regards
Stefan