[sdiy] FPAA for analog synthesis?

[cheater cheater]

I assume the arrays in the CEMs are still just the parts that were
necessary for the chip?
The difference between the complexity required for that and a
general-purpose solution is obvious.

D.

On Wed, Mar 3, 2010 at 01:05, Eric Brombaugh <ebrombaugh1 at cox.net> wrote:
> On 03/02/2010 04:54 PM, Dave Manley wrote:
>>
>> All the original SSM/CEM chips are based on an analog array - it isn't a
>> *programmable* array - you need to make a mask, and connect elements in
>> the array with metal. That said, the number of unique elements in the
>> array is small.
>>
>> If you'd like to see similar technology, including libraries, look
>> around here:
>>
>> http://www.arraydesign.com/
>>
>> Ok, so what? If there was a big enough market for these parts I'd hazard
>> to guess it shouldn't really be all that difficult to design an analog
>> array with technology similar or better than the ancient CEM arrays that
>> has a modern programmable interconnect. After all, these old devices
>> used only a single mask layer, so the routing possibilities aren't
>> endless. The question is who is going to pay to develop such a
>> technology given the state of ADC/DAC/DSP technology?
>
> Great point.
>
> I would imagine that a big part of the problem in going from a
> mask-customizable analog array to a field-programmable one is that the
> routing resources (switches, buffers, etc) behave much differently than pure
> metal runs.
>
> When Doug Curtis sat down to design a fantastic CEM circuit, he was able to
> make a few simplifying assumptions about the impedance of the interconnect.
> The task gets orders of magnitude harder when there are pass transistors &
> long-line buffers in the mix.
>
> Eric
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>