[sdiy] My first schematic, me too

[harrybissell at copper.net]

This is a "band-aid" ... filtering the
glitch is not a valid way to remove it
imho. You can even see from the drawing that
the triangle is rounded at the bottom..
the fixed filter assumes that the glitch is
WAY higher in frequency than the desired
triangle.

If you wanted to process that triangle and
needed a constant amplitude... the fixed filter
would hose you. High frequencies would be
attenuated.

In my case I wanted to process a (ahem) guitar
generated waveform. The tiny glitch would catch
on the PWM circuit and be a real pain-un-the- at ss.
otoh I could not stand to have amplitude variation
vs frequency either. Don't have a good solution
at this time... but this is not it.

PAiA (iirc) used a similar method to remove the
glitch (as Moog)... as did EFM and just about
everyone else. You'd be surprised at just how small
a glitch is really noticible !!!

If you have the CV available (in the guitar case
I do not)... you could use a tracking filter
to optimize glitch removal

H^) harry



>
>
>
>---- Original Message ----
>From: tim102 at tstinchcombe.freeserve.co.uk
>To: synth-diy at dropmix.xs4all.nl
>Subject: Re: [sdiy] My first schematic, me too
>Date: Tue, 2 Jan 2007 21:52:44 -0000
>
>>> Did you ever find a way to get more of the glitch out of it?
>>
>>The answer to this is essentially 'no', and what Paul points out...
>>
>>> By the time anything complicated is done to fix it, you might 
>>> as well redesign the oscillator.
>>
>>... seems pretty true in the Doepfer case, as the saw-to-tri
>topology uses
>>just the one transistor, and this limits the possibilities for a
>'simple'
>>fix (at least it does with my experience - others better qualified
>might
>>find a way!). In Moog's paper 'Voltage-Controlled Electronic Music
>Modules'
>>(available at <http://moogarchives.com/>: 'articles', 'AES 1964') he
>gives
>>what looks to be a good de-glitching method on a saw-to-tri
>converter, but
>>which has 'more to it' than the Doepfer circuit, and so seems to
>have more
>>scope for doing something about it.
>>
>>Tim
>>__________________________________________________________
>>Tim Stinchcombe 
>>
>>Cheltenham, Glos, UK
>>email: tim102 at tstinchcombe.freeserve.co.uk