Rainer Buchty wrote:
>
> > The Xilinx XC9500 CPLD series might be a better choice. It's available
> > in smaller configurations, it can handle 3.3 and 5V I/O, it's flash
> > based so you don't need the external EEPROM, but it's still in-circuit
> > programmable.
>
> Personally, I'd say that's a matter of price.
>
Yes, definitely.
>
> But going for a REC clone we
> already needs 10 registers of 8bit which comes to 80 flipflops, going to
> 16MB will take another 8 for bank selection. So we already have used 88
> flipflops (macrocells) before it comes to functionality... So we probably
> won't go for a XC95108/PC84 but at least a XC95144/TQ100 which costs
> $15.95 at min buy 25 (www.avnetmarshall.com) and slowest speed.
>
Hmmm... right. I didn't realize that there were so many registers.
Yeah, maybe the XCS10 is a better choice. If the 61 I/O's of the 84-pin
package is enough it should be about the same price as the XC95144/TQ100
(with 72 I/O's), but offer a lot more logic power.
Isn't the Spartan limited to 1.5 or 2.5V supply though?
Marko Mäkelä wrote:
>
> > Working *together* on *free* VHDL models of the various custom chips (and
> > the CPU) is something I'd highly appreciate.
>
> Me too. Do you know if the tools are freely available? I like the Atmel
> AVR family of microcontrollers, because everything can be done with free
> software. I use the avra assembler and the uisp programmer.
>
There is some free VHDL tools available
[http://www.opencollector.org/collector.php3]. In my experience though,
they offer far from what you get with commercial tools.
I have some experience with PeakFPGA from Accolade
(http://www.peakvhdl.com/) which I found to be very powerful, although
maybe a bit pricey for this audience at $2500.
--
Christer Palm
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail 2.1.1.