Re: TransWarp IIgs GAL's - Upgrade now possible

["Henry" <apl2research@comcast.net>]

Hello Eric.

Thanks for the data sheet link.  Saved me some time there.

As far as replacing the PLCC FPGA with something faster, although do-able I 
don't think many people would attempt such an upgrade since the IC is 
surface mounted.  Shame too that it is.

I know my cards are 50 MHz speed FPGA.  Probably all that is needed to run 
at 8 MHz reliably.  I do have all the tools for the surface mount operation, 
but finding just one or two of the FPGA might prove to be a head ache.  Most 
of my vendors prefer an order of 50 IC's minimum.  Boo, hiss!

Thanks again for the info.


Henry S. Courbis
www.GSE-Reactive.com
Apple II Series Legacy Hardware - Come take a look at what we have to offer!

"Eric Smith" <eric@brouhaha.com> wrote in message 
qhirs1jxlk.fsf@ruckus.brouhaha.com">news:qhirs1jxlk.fsf@ruckus.brouhaha.com...
> Henry wrote:
>> I'm also not too sure about the FPGA, what
>> it does or doesn't do.  I need the data sheet on that as well.  Now the 
>> fun
>> part - start looking for it.  Anyone?  I'm sure an email to Xilinx will
>> solve that, but maybe it's already on-line somewhere?
>
> http://www.datasheet4u.com/html/X/C/2/XC2000_Xilinx.pdf.html
>
>> Using an FPGA, I don't see why breaking
>> the 14 MHz barrier would be that big of a deal.
>
> In the XC2000 family, the speed grades were specified by the "toggle
> rate" in MHz of the flip-flops in the FPGA.  Standard speed grades for
> the XC2064 were -70, -100, and -130.  That sounds really fast, but in
> actuality the real performance you get out of the FPGA depends very much
> on the number of levels of combinatorial logic and the routing delays.
> So the "70 MHz FPGA" may be marginal for running a TWGS at 14 MHz or
> more.
>
> Note also that switching to a -130 part does not in actuality mean that
> the FPGA will operate 85% faster than the -50 part.  The ratings are
> what Xilinx guaranteed, but the -50 part probably has much more
> headroom (overclocking capability) than the -130 part.
>
> However, since in practice many people have reported success running
> their TWGS at 14 MHz or more using the original FPGA, it is reasonably
> plausible that anyone having FPGA-related trouble might be able to solve
> it by switching to the -100 or -130 FPGA.
>
> The XC2064 has been out of product for quite a while, but parts brokers
> still have them.  I have no idea what kind of prices they'd charge.
>
> Eric