[sdiy] MG-1 Mods (patching it out)

[WeAreAs1 at aol.com]

In a message dated 7/1/05 10:50:20 AM, mmanard at columbus.rr.com writes:

<< I'm not sure what you mean. A diode off pin 7? Will that remove 
the .5v? Will it affect the output of the EG in other ways? >>

Hello Michael,

Phillip's diode thing is quite simple, and would work and even help a bit, 
but it's not the optimum way to deal with the problem, in my opinion.  

FYI, here's how the diode idea works:  Standard silicon diodes have a 0.6 
volt "voltage drop".  That is, they reduce whatever voltage is running though 
them by 0.6 volts.  If you put 5 volts into the diode, you will get 4.4 volts out 
of it.  The voltage drop also works in such a way so that the diode will not 
conduct at all until the input voltage rises above the 0.6 volt point.  
Therefore, if you run your EG output through a diode, the output of the diode will 
stay at 0 volts until the EG output rises past 0.6 volts.  Sounds good, no?  

Yes, it would take care of your .5 volt offset problem, but here's what's not 
so good about this solution:  When your EG starts to rise, it will start at 
0.5 volts, but it still has 0.1 volts to go until the diode will allow the EG 
signal through.  This means you will get a rather strange attack phase -- it 
will have a short step or bump in the beginning of the attack.  Depending on 
your attack time setting, it may also feel like a kind of delayed attack.  You 
will also get a decay/release phase that falls nice and smoothly, then abruptly 
drops in stepwise fashion at the end.  These attack/release bumps will be 
*especially* audible when using the EG as a pitch modulation source.

Another problem the series diode will cause is that it will reduce the 
overall EG maximum output level by 0.6 volts.  This may give a noticeable kind of 
lackluster or weak EG performance, especially if this EG's normal maximum level 
is already relatively low -- say around 5 volts or so.

In my opinion, the more elegant solution to this problem is to add an output 
buffering stage with an offset trimmer.  The offset trimmer would be used to se
t the "at rest" EG output to exactly 0 volts.  If you wanted to be really 
clever, you could also set up your output buffer to have both positive-going and 
negative-going (inverted) outputs.  This would allow you to make a whole bunch 
of new types of sounds that would not be possible with just the standard 
positive-going EG output.  

Keep in mind that your MG-1 circuits may have been designed to expect that .5 
volt offset from the EG, and trimming it out may give you a slightly "weaker" 
EG sound.  This wouldn't be much of a problem when using it to drive VCO 
pitches, but might be apparent when using it to drive filters and VCA's, because 
you will lose .5 volts at your peak, and they may have been expecting that 
addtional .5 volts.  Hey, there are tradeoffs everywhere in life.  It would, 
however, be possible to add a small amount of gain to your buffer amp, to make up 
for the peak voltage lost in the trimming process, giving you the best 
solution.  The whole buffer amp/trimmer thing could be done with just one opamp stage, 
or maybe with two stages if you need some extra gain and also want an 
inverted output.

Peter Grenader has designed a very nice little circuit he calls the 
Gizmotron, which would make a very nice output buffer for your EG.  It has the added 
benefit of having an attentuator pot that gives both positive and negative 
ouput, with zero output at the center of the pot's travel.  This would be very cool 
as an EG attenuator.  The Gizmotron has its own offset adjustment trimmer for 
calibrating its own zero point, but I'm not exactly sure if the trimmer would 
also take care of your MG1 EG offset problem.  You might need to add a second 
trimmer for that, but it certainly would not be at all complicated.  You can 
find the Gizmotron at Peter's site:  http://www.buzzclick-music.com

Michael Bacich