[sdiy] Novation peak NCOs

[Veronica Merryfield]

Since an FPGA can run fast and you usually get a lot of logic blocks but limited IO pins (cost) it makes sense to use the glut of logic and speed to over come the lack of pins - 8 DACS using 8 pins saves the costs of pins and external devices.

Another ring benefit is that the internals of the circuits are hard to reverse engineer and if they used an FPGA with internal flash/rom and the security bit set, impossible, and of course, the circuit is now in software and easily changed.

I suspect this was mostly driven by cost for feature equations.

> On Apr 23, 2017, at 12:17 PM, Mrs Paula Anne Maddox <paula at synth.net> wrote:
> 
> There are very few 24KHz DACs available, plus with pwm you can kind of get a dac for the cost of a simple r/c network. There's nothing to stop you using a 16 bit DAC, but you'd need FPGA pins to drive it, so why not skip it completely?
> 
> ALso , as it's 8 voice poly, you'd need 8 dacs, as opposed to 8 mins on the FPGA.
> 
> So I suspect it's a cost thing to be honest, but personally as an engineer I see it as my job to remove excess parts where not needed, if a board doesn't need a part, why go through the hassle of adding it?
> 
> Paula
> 
> 
> 
> On 23 April 2017, at 16:06, Andy Drucker <andy.drucker at gmail.com> wrote:
> 
> 
> I'm trying to understand, in the NCO architecture being discussed (tentatively as we don't have exact details from Novation), is the choice to use a 1-bit DAC forced on us by other aspects of the design such as the very high (96MHz) sample rate?  
> 
> If not, is it a cost issue, and is it likely to be perceptibly worse (to some ears) than alternatives?
> 
> On Sun, Apr 23, 2017 at 2:28 AM, Mrs Paula Anne Maddox <paula at synth.net> wrote:
> I had lots of customers who said 002 sounded more analogue than 008, which was odd, but that's people's ears for you.
> 
> 
> On 23 April 2017, at 05:31, David G Dixon <dixon at mail.ubc.ca> wrote:
> 
>> Question on the Novation NCO thing: In the video I saw they
>> are going out of their way to imply without saying directly
>> so that an NCO is somehow more "Analog" than other digital
>> oscillators - is there anything to this?
> 
> It's because the sampling rate is 96MHz, or about 2000 times faster than
> your typical digital oscillator, so there is no digital character to the
> output.  At least, that's what I got out of it.  What I know about digital
> oscillators could be explained in about 96 microseconds.
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy