[sdiy] Square to Sine

[Elaine Klopke]

>What do you want a "dirty" fixed frequency sin for?

> taking this even further, aren't we going to have a bank of carefully
> tuned sine oscillators soon?

The plan, which is coming together nicely because of all you kind folks, is to make a module that has some passing resemblance to a drawbar organ tone generator using ATM Wiltshire's NoteDiv chips. The working name at the moment is the Organ Donor. Is it going to be perfect? No. But I've always thought a little (or a lot) dirt adds character and we'll see how it sounds when everything is plugged in and doesn't puff smoke. :)

>> On 05.04.2017 09:45, Richie Burnett wrote:
>> If you are going to the trouble (and slight increased cost) of making an active filter, or putting a buffer after a passive filter, then you would be better off introducing some "corner peaking" into the filter response.
>> 
>> You'll get a much cleaner sinewave by increasing the Q of the filter when the cutoff frequency is tuned to the fundamental frequency of the input squarewave. The resonant gain will also help to make up for any signal loss if you decide to cascade it with one or more passive stages.
>> 
>> The higher you design the Q factor to be, the purer the sinewave output will be, but the downside is that the tuning of the filter becomes more critical re component tolerances, drift with temperature, etc.
>> 
>> -Richie,
>> 
>> Sent from my Xperia SP on O2
>> 
>> ---- Andrew Simper wrote ----
>> 
>>>> On 5 April 2017 at 13:46, Magnus Danielson <magnus at rubidium.dyndns.org> wrote:
>>>> 
>>>> The backside of that would be that the 2 pole gives only 12 dB/oct compared to the 3 pole giving 18 dB/oct, thus giving less suppression of the square overtones. Their roll off would be 18 dB/Oct vs 24 dB/Oct from the filters. You would have to add a passive RC-link to achieve the same performance.
>>>> 
>>>> The simplicity of the RC link setup is good starting point from a learning perspective.
>>>> 
>>>> Cheers,
>>>> Magnus
>>>> 
>>> 
>>> The original passive 3 pole filter was drooping a bit much for Elaine,
>>> I was just proposing a very simple way of swapping in a BJT sallen key
>>> section cascaded with a passive low pass to keep the signal level a
>>> bit higher while still delivering the same 3 poles of filtering. I'm
>>> not following what you're talking about with the 12 dB/oct vs 18
>>> dB/oct vs 24 dB/oct, can you please let me know where the fewer / more
>>> poles are coming from?
>>> 
>>> Cheers,
>>> 
>>> Andy
>>> 
>>> 
>>>> 
>>>> 
>>>>> On 04/05/2017 04:41 AM, Andrew Simper wrote:
>>>>> 
>>>>> One way to avoid some of the volume drop is to use a 2 pole sallen key
>>>>> BJT filter with a cascaded passive stage. This is a common circuit
>>>>> used around BBD chips, eg:
>>>>> 
>>>>> http://www.hobby-hour.com/electronics/s/schematics/boss-ce2-chorus-schematic.gif
>>>>> 
>>>>> Andy
>>>>> 
>>>>>> On 4 April 2017 at 19:50, Elaine Klopke <functionofform at gmail.com> wrote:
>>>>>> 
>>>>>> Thanks for the info Tom!
>>>>>> 
>>>>>> Magnus, the sources are fixed square waves coming from a set of PIC chips. So the fixed filtering doesn't bother me a bit. One filter for one frequency. What worries me is the volume drop.
>>>>>> _______________________________________________
>>>>>> 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
>> 
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy