SynthModules

hi all. i see that there are various bits of software available which 
turn the PSIM into an LFO/LFO's. i have a couple of questions about 
this. 

firstly, how smooth are they at very slow rates, e.g. if you put a 
really slow sine on the pitch of an oscillator, can you hear any 
stepping?

next up, at the other end of the frequency spectrum, how do these 
LFO's sound when driven into the audio rates? i know the PSIM is not 
designed to do audio, but it must be possible? how does it sound? how 
easy would it be to implement PWM etc.

cheers,

gregg.

Hi All, 

Since it is summer, I have been working on my Vespa's and not 
spending time in the dungeon. But I did get some results before 
the nice weather set in.

The limitation on the Basic Atom Pro at the design time of the 
PSIM was the number of pins. Brice only had access to the 24 
pin version of the processor. That mean't the implementation 
had to use a very low I/O pin count. That fact (and economics) 
mean't the use of a serial DAC.

The transmit time for the serial DAC is the primary bandwidth 
limitation on the PSIM. The math coprocessor does 32 bit 
floating point math in one clock cycle. The actual calculation is 
very fast.

When the 40 pin version of the Basic Atom Pro comes out, you 
could add parallel DAC output. That will up the output rate by a 
whole lot and get the processor up into the audio range.

The 12 bit resolution of the DAC means each output step is 2 
millivolts. For most applications, that amount of zipper noise 
should be inaudible (inaudible stepping).

--- In SynthModules@yahoogroups.com, "mrboningen" 
<darkflametwentythree@h...> wrote:
> hi all. i see that there are various bits of software available 
which 
> turn the PSIM into an LFO/LFO's. i have a couple of questions 
about 
> this. 
> 
> firstly, how smooth are they at very slow rates, e.g. if you put a 
> really slow sine on the pitch of an oscillator, can you hear any 
> stepping?
> 
> next up, at the other end of the frequency spectrum, how do 
these 
> LFO's sound when driven into the audio rates? i know the PSIM 
is not 
> designed to do audio, but it must be possible? how does it 
sound? how 

> easy would it be to implement PWM etc.
> 
> cheers,
> 
> gregg.

----- Original Message ----- 

From: "grantrichter2001" <grichter@asapnet.net>
>
> Since it is summer, I have been working on my Vespa's and not
> spending time in the dungeon. But I did get some results before
> the nice weather set in.

As the South Florida summer intensifies, we *try* to spend more time in the
cool dungeons. Anyway, I don't get it -- what's a PSIM doing on a Vespa? ;-)


> When the 40 pin version of the Basic Atom Pro comes out, you
> could add parallel DAC output. That will up the output rate by a
> whole lot and get the processor up into the audio range.

The PSIM-2, perhaps?


> The 12 bit resolution of the DAC means each output step is 2
> millivolts. For most applications, that amount of zipper noise
> should be inaudible (inaudible stepping).

There is 2mv resolution at full strength. When you attenuate the PSIM's
output, though, you also decrease the variation per step, i.e. you increase
the resolution. Attenuating by 50% yields 1mv increments, for example. A
PSIM-generated LFO routed through a mod wheel or joystick will typically be
so attenuated that the steps will be fractions of a millivolt.

When you don't need the full output range of the PSIM, you can control the
output in 2 ways: within the PSIM (by scaling the output values within your
program), and externally (using an attenuator or mixer). Each method has its
place.
--
john

It seems an attenuator with a hbit of slew might work well.  Just a 
one-pole filter would do pretty good, or a single CyIndustries 
Lowpass Gate (from the quad)....  Then you would get better 
resolution (or quasi-resolution, as it were) and not have to 
attenuate everything quite as much (in case you wanted a wide-assed 
sweep range)....  owning cakes and eating them too anybody?

Jonathan

--- In SynthModules@yahoogroups.com, "john mahoney" <jmahoney@g...> 
wrote:
> ----- Original Message ----- 
> From: "grantrichter2001" <grichter@a...>
> >
> > Since it is summer, I have been working on my Vespa's and not
> > spending time in the dungeon. But I did get some results before
> > the nice weather set in.
> 
> As the South Florida summer intensifies, we *try* to spend more 
time in the
> cool dungeons. Anyway, I don't get it -- what's a PSIM doing on a 
Vespa? ;-)
> 
> 
> > When the 40 pin version of the Basic Atom Pro comes out, you
> > could add parallel DAC output. That will up the output rate by a
> > whole lot and get the processor up into the audio range.
> 
> The PSIM-2, perhaps?
> 
> 
> > The 12 bit resolution of the DAC means each output step is 2
> > millivolts. For most applications, that amount of zipper noise
> > should be inaudible (inaudible stepping).
> 
> There is 2mv resolution at full strength. When you attenuate the 
PSIM's
> output, though, you also decrease the variation per step, i.e. you 
increase
> the resolution. Attenuating by 50% yields 1mv increments, for 
example. A
> PSIM-generated LFO routed through a mod wheel or joystick will 
typically be
> so attenuated that the steps will be fractions of a millivolt.
> 
> When you don't need the full output range of the PSIM, you can 
control the
> output in 2 ways: within the PSIM (by scaling the output values 
within your
> program), and externally (using an attenuator or mixer). Each 
method has its

> place.
> --
> john

But doesn't the PSIM and it's Basic support a function (name escapes
me at the moment) to produce square wave in the audio range through
the Aux out?  As I recall, you can get it to actually output two
waves, each at it's own frequency, mixed.  Some futzing necessary at
the output for buffering and filtering required, but maybe good enough
for fun noises?

Mike

--- In SynthModules@yahoogroups.com, "john mahoney" <jmahoney@g...> wrote:
> ----- Original Message ----- 
> From: "grantrichter2001" <grichter@a...>
> >
> > Since it is summer, I have been working on my Vespa's and not
> > spending time in the dungeon. But I did get some results before
> > the nice weather set in.
> 
> As the South Florida summer intensifies, we *try* to spend more time
in the
> cool dungeons. Anyway, I don't get it -- what's a PSIM doing on a
Vespa? ;-)
> 
> 
> > When the 40 pin version of the Basic Atom Pro comes out, you
> > could add parallel DAC output. That will up the output rate by a
> > whole lot and get the processor up into the audio range.
> 
> The PSIM-2, perhaps?
> 
> 
> > The 12 bit resolution of the DAC means each output step is 2
> > millivolts. For most applications, that amount of zipper noise
> > should be inaudible (inaudible stepping).
> 
> There is 2mv resolution at full strength. When you attenuate the PSIM's
> output, though, you also decrease the variation per step, i.e. you
increase
> the resolution. Attenuating by 50% yields 1mv increments, for example. A
> PSIM-generated LFO routed through a mod wheel or joystick will
typically be
> so attenuated that the steps will be fractions of a millivolt.
> 
> When you don't need the full output range of the PSIM, you can
control the
> output in 2 ways: within the PSIM (by scaling the output values
within your
> program), and externally (using an attenuator or mixer). Each method
has its

> place.
> --
> john

> But doesn't the PSIM and it's Basic support a function (name escapes
> me at the moment) to produce square wave in the audio range through
> the Aux out?  ...

The name is about all that I *do* remember: freqout.
Freak out! :-)
--
john

That's it!  I'll play around with it some and see if I can get it to
work...

--- In SynthModules@yahoogroups.com, "john mahoney" <jmahoney@g...> wrote:

> > But doesn't the PSIM and it's Basic support a function (name escapes
> > me at the moment) to produce square wave in the audio range through
> > the Aux out?  ...
> 
> The name is about all that I *do* remember: freqout.
> Freak out! :-)
> --
> john

The highest frequency I can get out of a tight loop toggling the aux 
jack is 39 KHz.  Note the waveform is only 0 - 3 volts on the aux 
output.  The rise and fall time is nearly 50% of the duty cycle.  
Adding a single goto statement drops the frequency to 34 KHz so there 
is some room for very simple programming.

The highest frequency I can get out of a tight loop using a all four 
DAC channels is 860 Hz. If I drive only a single channel I can get 
this up to 3.2 KHz but there is jitter. There may be a better way to 
drive a single channel with less jitter.

Using the sound command I can generate sounds up to about 20 KHz. It 
appears that the 'note' parameter does not really work correctly. I 
get 50% of the frequency that I program it to use. However, using the
sound command to generate sequences of notes is kind of interesting.
I'll have to play around with this some more.

Here's an octave sequence:
sound p8,[500\130,500\147,500\165,500\175,500\196,500\220,500\247]

Dave
--- In SynthModules@yahoogroups.com, "mrboningen" 
<darkflametwentythree@h...> wrote:
> hi all. i see that there are various bits of software available 
which 
> turn the PSIM into an LFO/LFO's. i have a couple of questions about 
> this. 
> 
> firstly, how smooth are they at very slow rates, e.g. if you put a 
> really slow sine on the pitch of an oscillator, can you hear any 
> stepping?
> 
> next up, at the other end of the frequency spectrum, how do these 
> LFO's sound when driven into the audio rates? i know the PSIM is
not 
> designed to do audio, but it must be possible? how does it sound? 
how 

> easy would it be to implement PWM etc.
> 
> cheers,
> 
> gregg.

Freqout works reasonably well when followed by a low pass filter up
to about 1000 to 1500 Hz.  After that it starts to degrade. The dual
frequency capability is nice, though.

Dave



--- In SynthModules@yahoogroups.com, "Mike Marsh" <michaelmarsh@s...> 
wrote:
> That's it!  I'll play around with it some and see if I can get it to
> work...
> 
> --- In SynthModules@yahoogroups.com, "john mahoney" <jmahoney@g...> 
wrote:
> > > But doesn't the PSIM and it's Basic support a function (name 
escapes
> > > me at the moment) to produce square wave in the audio range 
through

> > > the Aux out?  ...
> > 
> > The name is about all that I *do* remember: freqout.
> > Freak out! :-)
> > --
> > john