[sdiy] AS3397 what is it?

Benjamin Tremblay btremblay at me.com
Sun Jan 19 01:09:29 CET 2025


Thanks to all. 
In comparison to a 3394 the 3397 seems to offer more pitch stability, need just as many multiplexers for your DAC, and require more coding to do things like pitch bend. 

So I need to get comfortable with making PCBs with DACs, multiplexers, and a SOIC-28 ic soldered on. 

It could be awesome. 

Ben

> On Jan 18, 2025, at 6:39 PM, Benjamin Tremblay via Synth-diy <synth-diy at synth-diy.org> wrote:
> 
> Yes honestly I don’t want to recreate the six trak tuning sequence.
> Benjamin Tremblay
> 
>> On Jan 18, 2025, at 3:45 PM, David Moylan via Synth-diy <synth-diy at synth-diy.org> wrote:
>> 
>> True for the 3396, but the 3394 is an all analog VCO with poor enough tracking that CPU based auto-tuning is recommended.
>> 
>>> On Jan 18, 2025, at 2:18 PM, Gordonjcp <gordonjcp at gjcp.net> wrote:
>>> 
>>>> On Fri, Jan 17, 2025 at 11:16:20PM -0500, Benjamin Tremblay via Synth-diy wrote:
>>>> 
>>>> So the Juno is an analog wave-shaping circuit processing control signals coming off some mid-80s industrial microcontroller. I love that sound.
>>> 
>>> Pretty much. Taking it back to basics, the Juno oscillator is just a saw-core VCO hard-synced to a digital pulse, and without the comparator that would make it reset itself. You must provide a control voltage from a DAC to set the ramp current which sets how fast the integrator charges, and a pulse to short the integrator capacitor and reset the saw which comes from a digital timer.
>>> 
>>> In the Juno 106, this comes from a pair of lookup tables where it interpolates between the semitone you're on and the next semitone (above or below depending on which output you're generating) and then loads a ratio into an 8253 programmable divider chip and a value into a DAC where it is latched by a sample-and-hold. This is calculated by the voice board processor every 4.3 milliseconds, based on things like the note played, pitch bend, LFO depth, and so on. Interestingly the "footage" switches on the panel do not play a part in this - they simply switch in a couple of flipflops for the clock and a couple of resistors for the control voltage!
>>> 
>>> The internal circuit of the Juno 106's thick film dual-DCO module MC5534 is floating about on the 'net but it's basically the same as a Juno 6 or 60, just with the polarity flipped.
>>> 
>>> Anyway the CEM3394 and CEM3396 do exactly the same job - you get a microprocessor to generate a bunch of control voltages and program some dividers to generate clock pulses, and the rest is just analogue.
>>> 
>>> --
>>> Gordonjcp
>>> 
>>> ________________________________________________________
>>> This is the Synth-diy mailing list
>>> Submit email to: Synth-diy at synth-diy.org
>>> View archive at: https://synth-diy.org/pipermail/synth-diy/
>>> Check your settings at: https://synth-diy.org/mailman/listinfo/synth-diy
>>> Selling or trading? Use marketplace at synth-diy.org
>> 
>> 
>> ________________________________________________________
>> This is the Synth-diy mailing list
>> Submit email to: Synth-diy at synth-diy.org
>> View archive at: https://synth-diy.org/pipermail/synth-diy/
>> Check your settings at: https://synth-diy.org/mailman/listinfo/synth-diy
>> Selling or trading? Use marketplace at synth-diy.org
> 
> ________________________________________________________
> This is the Synth-diy mailing list
> Submit email to: Synth-diy at synth-diy.org
> View archive at: https://synth-diy.org/pipermail/synth-diy/
> Check your settings at: https://synth-diy.org/mailman/listinfo/synth-diy
> Selling or trading? Use marketplace at synth-diy.org




More information about the Synth-diy mailing list