MOTM

The oscilators on my MOTM never drift, and that is very nice, except when
I want them to drift around.  I put the random output of the SH101 module
into an oscillator FM in, but it's just not the same.  Randomness tends to
still center around zero and doesn't get into long drifts.  What is needed
is some sort of accumulative module.

This module will have knobs controlling the speed of the drifting and
whether it has an upward or downward tendency.  What happens is that the
module outputs some DC voltage.  Periodically, depending on the speed
setting, it will look at some randomness and add or subtract some number
of milivolts from it's output depending on what the random number was.  It
will then ramp it's output up or down to the new value.

With a 2 u module, you ought to be able to get 4 or more drift outs to
route to ocillators and filters.

Super bonus feature - a "warm-up" drift.  Drifting varies based on time
since module was reset, in a way that mimics "real" drift.

I have a little MAX patch that mimics drifting for me, but it seems like
there  ought to be a nice harware solution, and I have more important data
for my MIDI/VC convertor to crunch.

TIA,

Celeste, who doesn't have enough money for a Moog to solve this problem...

--
Celesteh@...     http://www.casaninja.com/celesteh
                                 http://www.mp3.com/celesteh

--- In motm@y..., <celesteh@c...> wrote:
> 
The oscilators on my MOTM never drift, and that is very nice, except 
when I want them to drift around.  I put the random output of the 
SH101 module into an oscillator FM in, but it's just not the same.  
Randomness tends to still center around zero and doesn't get into 
long drifts.  What is needed is some sort of accumulative module.
> 
       
Hmmm....., perhaps some PAIA equipment is required.   I do believe 
that the PAIA oscillators are sensitive to the background radiation 
from outer space and produce pitch changes accordingly.    :-)

Andy

One thing I like to do (which is kind of related) is to patch noise (or LF 
random) to a VC lag processor to the SH to the CV input of a VCO. With 
careful adjustment of the VC and other controls one can get the pitch to roam 
the whole 10v (ten octave) range, but moving by very small steps thereby 
taking a long time to get there -- actually, I tend to use patches like this 
more often for things other than VCOs.

I guess that patch is pretty obvious, and has nothing to do with oscillator 
drift; something which seems less novel when one has to tweak the tuning of 
the VCOs of a 2600 every three hours of use (and those sliders make it even 
more fun).

JB


In a message dated 1/16/02 1:34:24 PM Pacific Standard Time, 
celesteh@... writes:

> 
> This module will have knobs controlling the speed of the drifting and
> whether it has an upward or downward tendency.  What happens is that the
> module outputs some DC voltage.  Periodically, depending on the speed
> setting, it will look at some randomness and add or subtract some number
> of milivolts from it's output depending on what the random number was.  It
> will then ramp it's output up or down to the new value.
>

Since others have already mentioned slow LFO's to S&H through a lag processor 
to accomplish this, which is what I do sometimes, then my other suggestion 
would be not to install the Tempco temperature compensating resistor.  Your 
VCO's may drift around plenty after that.

-Elhardt

>The oscilators on my MOTM never drift, and that is very nice, except when
>I want them to drift around.  I put the random output of the SH101 module
>into an oscillator FM in, but it's just not the same.

Are you using a Roland SH-101 (which isn't a module) or an MOTM-101??  The
"random" setting on a SH-101 is a quantized voltage -- it's not that random.

>Randomness tends to still center around zero and doesn't get into long
>>drifts.  What is needed is some sort of accumulative module.

You can patch an output of the 320 into the FM input of the 300 -- sine
might be a good choice.

>Super bonus feature - a "warm-up" drift.  Drifting varies based on time
>since module was reset, in a way that mimics "real" drift.

The 320 also has a saw output, and a reset feature.  You may patch the saw
ouput back into the 320's FM input, so as it "warms up" the "drifting"
settles down, or you could also use a 110 to control the 300 modulation
depth.

Well, Maxim makes chips that are thermometers. They output 2mv/C. That's about in the range of
Moog drift (they are about 15mv/C). Cost a quarter. You could put one into the end of a 1/4"
plug, and stick it right into FM2 of a MOTM-300. Use 2 hearing aide batteries to power it.

Paul S.

>      I am new to this forum and fairly new to analog synthesis, so please
> forgive me if this question sounds stupid, but why would you want your
> synthesizer to drift? I thought the oscillators were carefully  designed so
> that they wouldn't drift.

Well, they *are*. My attitude is: it's a lot easier to ADD drift than to try to remove it/live
with it.

Some folks like the "random beating" that was a by-product of older VCO temperature drift. If so,
here are ways to emulate that, without sacrificing the glory of the MOTM-300/310 VCOs:

a) *ever so slightly* de-track the VCOs by adjusting the 1V/Oct trimmer. Say you have 3 VCOs.
First, set all 3 for 'perfect' tracking, having zero beat frequencies as best you can. Then, on
VCO #1, turn the tracking pot about a 1/8th turn CW, and on VCO #2 turn 1/8th turn CCW. Leave VCO
#3 alone.

b) use "bad" control voltages. Modern MIDI-to-CV converters have 16-bit DACs that are very
accurate. Instead, use the....errrrr....."lowly" PAiA MIDI-to-CV for that great 7-bit accuracy.
So, even though the VCOs are 'perfect', the CV feeding them is not. Use the CV/Gate of an SH-101
(gack!). The older, the better.

c) Use the MOTM-320 LFO with RATE = 0, AND with -5V fed into the 1V/Oct input. Feed TRI into a
MOTM-850 input on '1' setting, then to a FM input on '1' setting (i.e. heavily attenuated). This
simulates the air temperature slowly rising and falling.

At NAMM, we left the modulars on 9hr/day (6 days in a row). The drift was about 0.2Hz per day. To
me, that's GOOD.

Paul S.

> At NAMM, we left the modulars on 9hr/day (6 days in a row). The drift was
> about 0.2Hz per day. To
> me, that's GOOD.

No s***... This doesn't surprise me at all. I'm running a 440 into a 420
with the Cutoff and Res in each set "just so" to create a nice beating
effect. This is for a theatre production, and the modular is left on 13
hours a day (at 230v btw) for rehearsals, soundcheck and performance... and
so far I've only had to account for drift once in 1 week. And that's VCF
drift in two modules, not a VCO!