Re: [Balloon] Re: Next Balloon

Top Page
Attachments:
Message as email
+ (text/plain)
Delete this message
Reply to this message
Author: Steve Wiseman
Date:  
To: balloon
Subject: Re: [Balloon] Re: Next Balloon
27/07/2004 03:45:16, Jamie Morken <>
wrote:

>Oops, just found it in the design guide:
>"The GPIOs are not fast-switching devices and do not switch any

faster than
>10 MHz."


I think that the SA1100 also had such a warning, and its GPIOs
used to go astonishingly fast. During development, we were able o
get a 3-intruction loop to wiggle them, on a 200MHz clock...
Not tried on a more recent chip, so dunno - easy enough to
check, though
However, what is it that you're contemplating doing that way?
Sounds like the job for a tiny PLD / FPGA / Scenix micro, rather
than wasting the other 99.9% of a PXA on something it'll struggle
to manage.
(Free (money) design tools for FPGAs are extremely usable - Xilinx
Webpack stuff has, up until I needed to start embedding 32-bit
micros in FPGAs, been fine for even some relatively tricky stuff)
If you really must use a micro to do this, I suspect that you'll end
up trawling thought TI's (or other - I have no experience other
than TI) madder DSPs, and even then, I'm not sure you'll get near
200MHz interrupts. FPGAs have to be the way forward.

Steve