[sdiy] Divide down question

Tue Aug 25 16:09:57 CEST 2009

What about this:
start out with a normal square divide down, and then when you're e.g.
playing C1, then actually what is output is:
C1+ 1/2 C2 + 1/4 C3 + 1/8 C4

Each of the higher tones would have to go through a latch. That latch
would only let through one pulse of C2, C3, etc, but it would be reset
on C1's rising slope. Maybe this could be a little bit difficult,
maybe not.

This should give you a 'steppy' saw down, shouldn't it?
I think the latches could also add a little bit of delay or,
alternatively, you could add a little solenoid or cap for each of the
partials. Then, when you press the C1 and C2 keys on the keyboard, the
voices could sound a little bit more alive (because of the time
offsets creating phasing)

Actually, couldn't you do a big part of square-based divide down
topology in FPGA?

On a side-note: can you make a solenoid in an FPGA? It should be
possible if FPGAs can have multiple layers of gates, shouldn't it?

Cheers

D.

On Tue, Aug 25, 2009 at 11:28 AM, Tom Wiltshire<tom at electricdruid.net> wrote:
> One thing that you can do to get pulse waves is to start with a standard
> square-wave divide-down scheme, but then run pairs of square waves at
> octaves into AND gates. If you feed a square at frequency f and one at 2f
> into an AND, you get a 25% pulse wave out the far end. You could use 4f and
> have 12.5% if you prefer. Given that you've already got all the octaves you
> need, this seems like the easiest way to get divided-down/up pulse waves.
>
> T.
>
> On 25 Aug 2009, at 09:55, Scott Nordlund wrote:
>
>>
>> It could be done with sawtooth waves too:  use a comparator to make a
>> square wave, invert it, amplify the sawtooth by a factor of two, and mix the
>> two.  Of course the comparator's threshold has to be adjusted properly, or
>> there will be asymmetric pulses and sub-harmonics sneaking in...
>>
>> Pulse waves would be more difficult.  The thing with sawtooth and triangle
>> waves is that you get an instantaneous one-to-one mapping of input amplitude
>> to however many octave-doubled outputs you want.  You obviously can't do
>> this with square or pulse waves, since the amplitude isn't continuous.  You
>> could use a phase locked loop with a frequency divider in the feedback loop
>> (so the frequency can be multiplied by any arbitrary number), but this needs
>> time to adjust to changes in frequency.  It won't be instantaneous like the
>> sawtooth or triangle-derived, and it can have objectionable glitches and
>> artifacts.  But you could potentially get all the octaves you want in one
>> go, deriving them from the frequency divider in the feedback loop.
>>
>> ----------------------------------------
>>>
>>> Date: Mon, 24 Aug 2009 19:37:36 +0100
>>> Subject: Re: [sdiy] Divide down question
>>> From: cheater00 at gmail.com
>>> To: synth-diy at dropmix.xs4all.nl
>>>
>>> Tom,
>>> extremely interesting! I should think about that for a while. This
>>> 'multiply up' as I would call it is especially cool. Any sound
>>> examples?
>>>
>>> Which makes me wonder: can you do that with pulse waves..?
>>>
>>> I think you could almost certainly do that with sawtooth waves.
>>>
>>> D.
>>>
>>> On Mon, Aug 24, 2009 at 5:53 PM, Tom Wiltshire wrote:
>>>>
>>>> On 24 Aug 2009, at 16:00, cheater cheater wrote:
>>>>
>>>>> Hey guys,
>>>>> in a divide-down architecture, every key has a divide-down 'level'
>>>>> which then gets used in the same note but an octave lower.
>>>>>
>>>>> Does any amplification happen at each stage?
>>>>
>>>> Not really, or at least not in the analog sense. The division is done by
>>>> flip-flops, e.g. digital circuits, which means the output is the same
>>>> level
>>>> as the input, so no amplification is required.
>>>>
>>>> As an aside, I once considered doing a "divide up" scheme, using a
>>>> triangle
>>>> wave master oscillator at a low frequency fed to a series of precision
>>>> rectifiers. This provides triangles in octaves going up, and does
>>>> require
>>>> gain, since the rectifiers cut the signal level in half. I hoped that
>>>> the
>>>> reduced harmonic content of triangles when compared with square waves
>>>> would
>>>> give me a smoother tone. I built the circuit and proved the concept, but
>>>> never did anything with it.
>>>>
>>>>> If I press multiple keys from the same 'class' (e.g. C1, C2, C3) will
>>>>> any 'amplitude stealing' happen?
>>>>
>>>> No. Essentially all the tones are fed to a big mixer. Imagine a simple
>>>> inverting op-amp mixer with lots of 100K input resistors and 100K
>>>> feedback
>>>> resistor for unity gain. Now imagine what happens if you feed the same
>>>> tone
>>>> into two of the input resistors. Those two resistors are effectively
>>>> paralleled, which will halve their joint value and cause the signal fed
>>>> to
>>>> them to double in level - exactly what is required.
>>>>
>>>>> Where can I read more about divide-down synths? I am not yet good
>>>>> enough to read the schematics just like that; I tried googling around
>>>>> but couldn't find much of substance.
>>>>
>>>> You might have better luck looking at some combo organ sites, since
>>>> they're
>>>> mostly done this way.
>>>>
>>>> Good luck!
>>>> T.
>>>>
>>>>
>>>>
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at dropmix.xs4all.nl
>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>
>> _________________________________________________________________
>> Windows Live: Make it easier for your friends to see what you’re up to on
>> Facebook.
>>
>> http://windowslive.com/Campaign/SocialNetworking?ocid=PID23285::T:WLMTAGL:ON:WL:en-US:SI_SB_facebook:082009
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>
>
>
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>