[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[avr-gcc-list] [Re: Allocating variables to a fixed, address]
From: |
Robert von Knobloch |
Subject: |
[avr-gcc-list] [Re: Allocating variables to a fixed, address] |
Date: |
Wed, 13 May 2009 09:41:55 +0200 |
User-agent: |
Thunderbird 1.5.0.14 (X11/20060911) |
From: David Brown <address@hidden>
I've often heard of problems with FTDI chips, and especially their
drivers, but I've found them solid and reliable (except for their demo
code for PC software, which is rubbish). The example I gave of sending
multiple megabytes at over 1 MBaud is using FTDI devices.
But no matter what you are using for transferring your serial data, you
should have a system for error checking and retrying in place - you can
never consider such a link as 100% reliable.
Sorry David,
I'll have to disagree with you.
I can and do consider such a link (when done with rs232) completely reliable
and have done so for about 30 years without problems. But when I take a working
system (rs232) and 'insert' the FTDI chip (and USB port), it sometimes breaks.
I have no time, at present, to investigate deeper but will do so when I can. I
am suspicious of the handshaking and possible delays in the chip, but that's
only my guess right now. I use no 'driver' (except the native part in
GNU/Linux). These are FT232R chips, by the way.
Also I regret the way that PC manufacturers are removing the rs232 ports. I
understand they are too complex the average user, but industrially far more
useful than USB, where you a re more or less tied to using a badly specified,
proprietary chip unless you want a complete development project for something
that used to require an hour's work to adapt a trusted routine to a new
platform.
p.s. All my development is done under GNU/Linux.
Regards,
Robert
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [avr-gcc-list] [Re: Allocating variables to a fixed, address],
Robert von Knobloch <=