SV: SV: [sdiy] X-mas gift...

[Rönnberg Niklas]

Thanks for your answer Magnus!

I see the problem with the circuit, to bad I allready has made two for my
modular project... :-(

Anyway, does this look better?
http://www.itn.liu.se/~nikro/diy/PS-3_mods.pdf

Best regards!

/Niklas

-----Ursprungligt meddelande-----
Från: Magnus Danielson [mailto:cfmd at bredband.net]
Skickat: den 28 december 2004 14:51
Till: Rönnberg Niklas
Kopia: heitor.alves at gmail.com; synth-diy at dropmix.xs4all.nl
Ämne: Re: SV: [sdiy] X-mas gift...


From: Rönnberg Niklas <nikro at itn.liu.se>
Subject: SV: [sdiy] X-mas gift...
Date: Tue, 28 Dec 2004 13:02:08 +0100
Message-ID:
<82BEDACF1CE44E45AB82F167BD8C07D1A078FB at merlin.epost.itn.liu.se>

> Yes, I have looked at this page too. I was thinking of making an external
> box containing the extra circuits and potentiometers. If the external box
> isn't preset the "synth" will work as original. It would look something
> like this (on the inside):
> http://www.itn.liu.se/~nikro/diy/PS-3_mods.pdf
> 
> What do you think? How many basic (and stupid) mistakes have I done?

Mistake or mistake... your input mixer is a bit non-optimal in that when
you
change the volume of one sound you also change the volume of other sounds.
If you instead would have connected them to the negative input of the
op-amp
you would have made them independent of eachother. The reason is that the
other four inputs will load the output of one input in parallel to the 47k
at
the op-amp. The other inputs loading depends on where on the pot-scale you
are.
In practice this effect may not be very extreme unless you turn all other
inputs to 0 except one, then you have four 22 k resistor in parallel with
47k
giving roughfly a load of 5k instead of expected 47k. A more normal case
would
have you running with a load of four 27 k resistors in parallel with the
47k
giving roughfly a load of 6k. With lower values of the mixing resistors you
would have had a greater influence. Just shifting the input toss this
little
problem aside very neatly since the negative input with feedback resistor
forms
a virtual ground with current input rather than voltage input. Regardless
of
the contribution from the other inputs will the virtual ground stay (well
ideally, it will move but so very little that we can for our normal audio
aspects forget it) at about the same voltage all the time, more or less the
same as the positive input which is grounded.

So, it may not be a big mistake, but I thought you could learn something
from
it never the less. If you are worried about inverting the waveform, waste a
full TL072 at the task, one for the mixing and one to invert the whole
thing
back again.

> In this way I could choose waveform and the mix between different
waveforms,
> as well as in which filter the new waveform will be post-processed.

Fair enought.

> Next part I must consider is how to change the amplitude envelope, if
this
> is possible at all.
> 
> Maybe I could insert a VCF and an envelope follower, as complement to the
> original filters...

Soon will the add-ons be larger than the original box!

Cheers,
Magnus