[sdiy] ADuC702x (Re: programmable synth module)
Eric Brombaugh
ebrombaugh at earthlink.net
Thu Feb 16 19:33:22 CET 2006
Depends on what language you're using. For low-level ASM and C, you
could go out an buy commercial devtools for ARM, but if you use Linux or
MacOS X then ARM gcc is free. Under WinXP you can get ARM gcc running
under CygWin. Of course the free tools don't have an IDE with
source-level debugging on the embedded processor, and JTAG hookups are
harder to come by. If you want professional tools, expect professional
prices ($1k & up).
If you're using Forth (second link in genie's msg) then the development
environment runs right on the embedded processor - your only devtool is
a terminal app.
Eric
Gorka Garcia wrote:
> How expensive are the development tools for one of those? JTAG emulators and development enviroment (IDE with compiler, debugger...)?
>
>
> --- genie at netsynth.org wrote:
>
> From: genie at netsynth.org
> Date: Thu, 16 Feb 2006 00:54:33 -0800
> To: xyzzy at sysabend.org
> Cc: synth-diy at dropmix.xs4all.nl
> Subject: [sdiy] ADuC702x (Re: programmable synth module)
>
> Hi,
>
> How about using a modern 40 MIPS ARM7 chip from Analog Devices - ADuC7026.
> It's a true system on chip with inner OSC, 12-Ch 12-Bit ADC + Quad 12-Bit
> R/2R DAC and external bus for extra memory or peripherals. Price is
> @US$16.66/Digi-Key (or, you might get some samples from AD).
>
> http://www.analog.com/en/prod/0,,762_0_ADUC7026,00.html
>
> And, Dr. Ting told that his STC eForth V2 for ADuC7020 runs on ADuC7026.
>
> http://offete.com/FORTHstamp.html
>
> It should be a nice combination for 'programmable synth node' though
> ADuC702x chip runs slower than Philips LPC2100 series if farmware stored in
> some wait-state flash rom.
>
> Cheers,
>
> genie
> http://netsynth.org/
>
> Tom Arnold writes:
>
> -snip-
>
>> Personally what I'd like is a nice fast Forth based solution. To this end
>> I've played with a couple Atmel based Forth solutions, but they have been
>> short of my rough goal of 2MIPS. I reached the 2MIPS goal by figuring out
>> roughly how many Forth instructions I'd need to process per second to handle
>> 2 44.1khz sampled inputs and 2 44.1khz outputs along with a few instructions
>> to mutilate the data. This was arbitrarily chosen with the idea that it
>> would roughly translate also to 8x10khz -> 8x10khz.
>>
>> It'll be hard to hit these speeds with an interpreted language and a
>> cheap/small CPU. Note I say Difficult, not impossible.
>>
>> Oh, and if someone does cook up a solution, make sure its compatible with
>> licenses other then GPL, or at least make sure you really understand the GPL
>> before trying to use it...
>>
> -snip-
>
>
>
> _____________________________________________________________
> www.deMadrid.com , el correo gratuito de Madrid. 25 Mb de espacio.
>
>
More information about the Synth-diy
mailing list