CATWEAZLES Fraktal / CellilarAutomata Synth

[catweazle]

 * PICsynth Betaversion03a bugfix *

- FM-Synth OSC A/B oddity fixed. Now both are change values in the same direction
  http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_Betaversion03a.hex

I have deleted the old betaversions, only the newest is online (prevents chaos on server and organizing it)

That pot prob (step selection) seems to be mechanical. I've printed out the values from the
PIC internal ADC(analog to digital converter) via PC RS232 terminal and measured the
real voltage on the ADC input to compare them.
There is an area on the left and right side where the values (real and adc converted) are not changing.
This problem could be removed by software (scaling with a lookup table)  but this kills performance
and it depends on the pot used. Some sort of calibrating would be needed then.
... maybe we have used cheap shitty pots (I used very old ones)
... or I do a big mistake in thinking/software

Be sure that you are using linear pots, not audio (logarithmic) pots!!!

I add one 7-segment LED display next, connected directly to the PIC to show the selected step.

[catweazle]

ok, another update and a demo of the pseudo WSG (synth 4).
The pitch/speed range of the FM and WSG synths are expanded...
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_Betaversion03b.hex

Long WSG Demo (external filter and delay )
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_WSG.mp3

and a new demo from the FM-Stepsequencer (lower sequencer speed)
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_FM-Stepsequencer_%233.mp3

enough for today...

[Circuitbenders]

... maybe we have used cheap shitty pots (I used very old ones)
... or I do a big mistake in thinking/software

Be sure that you are using linear pots, not audio (logarithmic) pots!!!

I add one 7-segment LED display next, connected directly to the PIC to show the selected step.

I've tried it with 3 different brand new linear pops of different types and you get the same effect so i'm thinking it might be in software.

The LED display is a great idea if it can be added in.

One slight problem with the FM speed knob is that it now seems to have a logarithmic response, in that the speed difference for a small movement as higher speeds is a lot more than at slower speeds. Maybe it was always like that but because there are slower speeds available the faster speeds have all been pushed into a smaller pot rotation at the top.

[catweazle]

I try to fix that with the strange scale of the step-select pot. But in the moment I have no idea.
The LED display should fix that for now.

Speed/logarithmic isue:
problem is: every pot is converted into values 0..255
that means the whole speed range is divided into 256 steps, (not much)
But you are right, the old scaling for the FM-sequencer speed was better.

here the firmware, fm speed old scale.
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_Betaversion03c.hex

[catweazle]

For those who want to know what can be done with the PICsynth FM-Stepsequencer,
here is some upspeed techno stuff. (no drum-machine or other noise-makers were used)
"BassDrum" and "hi hats" result from FM modulation (and internal math errors in software).
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_FM-Techno.mp3

... schematic and software for display and extra switches is in progress, coming soon ...

[catweazle]

PICsynth LED display and switches added...

Schematic:
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_V11_schematic.JPG
old schematic version (REV1.0) will no longer be supported.

Software/Firmware Betaversion 04:
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_Betaversion04.hex

[catweazle]

photo of the picsynth ...

[Circuitbenders]

Thats looking good.

How do you mount one of those number LED displays, i've never had to do it before? Looks like i'm going to have to build another one with a bigger box and more space for additions on the board. Whats the plan for the other two switches?

I assume that everything will still work with the updated firmware and the play / reverse switches but without the LED number display actually installed. I suspect there will be some people that may want to build the thing but might find the display unnecessary or too complex to add. Also i'd like to update the firmware of my first prototype to the version with the switches but theres no space to mount the display in the box i put it in.



UPDATE: i had Betaversion03c loaded into mine and it started doing something weird. On the FM step synth if i set the step value and then turn the step selection knob up and down quickly all the steps become the same note. If i change the step value and turn the selection knob again all the steps become the new note. Its as if the step value knob is constantly outputting data. I upgraded to Betaversion04 and added the play and reverse switches and its still doing the same thing, maybe my pot is dying or dirty although i'm using sealed pots so it seems unlikely. It seems like it happens at random but its a bit difficult to program if you don't know if all the steps between the one you are on and the one you want to change are going to change as you pass through them.   

[voodoolikeudoo]

What way round are the pots wired up on the schematic there?
Which pin is pin 1 depends on which way you are looking at the pot and there doesn't seem to be any normal way of numbering them.

[catweazle]

@voodoolikeudoo:
if you look at the potentiometer, axis facing to you, then connect left pin to gnd, the right to +5V and the middle pin goes to the microcontroller (AN0..7). The schematic/layoutsoftware (cadsoft eagle) has that numbering in the libraries.
Voltage on the middle pin should rise if you turn the knob clockwise. -> see picture

@circuitbenders:
I like that transparent boxes you use!

Display:
I have simply cut out a hole fitting the LED display and then glued it with hot-melt adhesive.
The series resistors (current limiting for the LEDs) were soldered behind the display (I used small resistors).
Or you can solder the Display on a piece of veroboard and mount the board with screws.
There are little LED displays available, I found one with dimension 10mm x 8mm
The display can be omitted if not neccessary. But if used it have to be a "common anode" (CA)  type !

Switches:
Play and Reverse can also be omitted, but you have to do the following:
  - connect PLAY       to +5V  (always play)   [PIC pin B5]
  - connect REVERSE  to GND (play forward)  [PIC pin B4]
  - connect the other 2 inputs to GND          [PIC pins B0 and B1]

The "other 2 switches":
have no function yet. I have some ideas like a global sequencer controlling all synths for example.
But ignore them (ground them) for now if you have some keep the space for them.

Sequencer step selection bug:
oh oh, I see what you mean... really not nice ... will be fixed.
Thanks for investigating and reporting the bugs!
well, maybe a bit annoying all that bugs and extensions,
but hey, it's all free and I don't want to sell a ready made perfect (and expensive) high-end synth.
I keep in mind that I do no more hardware changes if not really necessary.

[catweazle]

*********
* BUGFIX *
*********
FM step selection problem:
Pot values toggle on some positions between two values, what is normal and no error.
This could be seen on the display if the Step-select-pot is "between" two steps.
The display shows a mixture of both values (flickers).
Therefor the selected step toggles.
I've tried different possible solutions to solve that.

I've used  a common method often used in commercial synths/sequencers.
If you change the Step, the Step-value pot is locked.
To unlock it you have to fetch the old value with the Step-value pot.
Turn pot from left to right (or in the other direction).
If the pot-value and the stored value (old value at the selected step)
are the same the Step-value is unlocked and the Step-value-pot
changes now the selected step value.

It is a bit stressy to fetch the old value with the pot to unlock the value but
this has some advantages against other methods, and it is very easy to implement in software.
When the pots get older there could be jumps in the values (like crackling on audio amplifiier pots)
which ignores the "fetch old value"-method.
Since the performance is limited I have to implement all as simple as possible. It's no DualCore Pentium

Another new feature is that the displays shows symbols for the selected synth:

Fraktal1   Fraktal2   FM-Seq    WSG
                 _            _             _
  |  |                       |_|           _
   _           |  |          _            |_|

   
Here's the bugfix
http://etecnic.fh-pforzheim.de/~catweazle/F-Synth/PICsynth_Betaversion04a.hex

[Circuitbenders]

Just discovered you can get some insane loops & sounds with the fraktal1 algorithm running at maximum speed and adjusting knobs 5,6 and 7. The reverse switch can turn then inside out as well.

Heres a couple of demos of full speed fraktal1 action

http://www.circuitbenders.co.uk/demos/PIChispeedfraktal1demo1.mp3
http://www.circuitbenders.co.uk/demos/PIChispeedfraktal1demo2.mp3

[yogi]

Very cool, catweazel! Ill be building one very soon.

[oscillate_wildly]

Hi, could you please re-up the .HEX file as the link is dead. thanks x

[Circuitbenders]

catweazle seems to have disappeared off the face of the planet and so has his website. He hasn't signed in here since march 27th.

I think i might have most of the hex files in a backup somewhere. I'll take a look and post it if i can find them.