[sdiy] More new Alfa chips - CEM3394

Richie Burnett rburnett at richieburnett.co.uk
Thu Apr 4 20:29:47 CEST 2019

Very true.

And Neil has a good point that not all CVs are unique to each synth voice in a polysynth, so they can be connected to a common analogue CV line. It would make little sense writing the same value multiple times to the same parameter register in each voice chip if everything was handled digitally over something like SPI.


Sent from my Xperia SP on O2

---- Tim Ressel wrote ----

>you can get an 8-channel 8-bit DAC with SPI for about $2. no need to 
>muck about with muxes.
>On 4/4/2019 5:14 AM, Neil Johnson wrote:
>> Hi Rich,
>>> When I saw "Microprocessor controllable..." in the datasheets for these
>>> CEM chips I was initially expecting to see an SPI interface! :-(
>> Indeed not, as they say (AS3394E):
>>> Thus interface to a microprocessor system may be accomplished simply with a single DAC, 4051-type CMOS multiplexer, and 8 hold capacitors.
>> That said, for a polyphonic synth it is actually quite useful to have
>> voltage control of these parameters, in EGs for example, where each
>> parameter is common across all voices so you hang the controls off a
>> voltage bus.  For example this is how the Siel Opera6 does things,
>> with all the SSM EGs driven off common voltages for the A, D, S and R.
>> If you did this with SPI then you would either need some sort of
>> addressing scheme (in that case use I2C), OR you end up with a huge
>> mass of slave select lines!!  And you've now increased the cost of the
>> silicon by putting logic (another voltage rail) and some number of
>> DACs plus reference on the die.
>> Cheers,
>> Neil
>--Tim Ressel
>Circuit Abbey
>timr at circuitabbey.com
>Synth-diy mailing list
>Synth-diy at synth-diy.org

More information about the Synth-diy mailing list