SV: [sdiy] Mopho modulation DAC configuration, update rate, etc

[Dave Manley]

karl dalen wrote:
>> That's not so bad. Gives him a theoretical top end of about
>> 500Hz for modulation frequencies. But the LFOs on his stuff
>> go only go up to Middle C at 261Hz, so he's keeping it all
>> on the safe side.
>> He might mean that the modulations within the dsPIC voice
>> processor (e.g. the ones that never leave the digital
>> domain) are calculated at a higher rate than the DAC runs
>> at.
>>     
>
> yes but he have said that in context of the *slow*
> analog side, which of course are not true.
>
> However he might have reffed to CV slew circuit which
> will put up a Bw limit on update rates.
>
> It's kind of peculiar, he uses 16 bit DAC whit rather
> poor INL and then slews of the steps by a RC net??
>
> The Pic are fast enough to generate modulation without
> skipping bits.
>  
>   
I scoped out one of the modulation paths last night,  LFO -> VCA 
controlling the digital noise source.  The max frequency of the LFO is 
250Hz, and the supported waveforms: tri, saw, reverse saw, pulse, don't 
look much like their ideal forms at the max rate.  As the frequency 
increases, the amplitude of waveforms decrease, and they look 
increasingly very non-ideal, and "stepped".  The voltage at the VCA 
control input is very poorly smoothed.  I'll try to get some scope 
captures and post them.

I didn't measure the min frequency.  The waveforms of course look much 
better at the lower frequencies, where they find their main use.

One annoying find: there's no way to completely shut off both VCOs.  The 
VCOs have a mix pot that lets you turn either off, or get a mix of the 
two, but you can't simply turn them both off.  The sub-oscillators do 
have their own independent level controls, as does the noise.  You can 
set the VCOs to their pulse wave, and set their duty cycles to 0 or 99%, 
but this wasn't reliable - it only shuts off the outputs completely at 
some frequencies.

-Dave