Synth Module Board - Cheezy Version

SMBC

First Posted Jan 17, 2012
Updated October  29, 2018

    It has been a long time since I had the idea of making a Cheezy Synth board.  It never happened because at the time I did not want to pay close to $400 to get a small set of boards made and also because at the time I felt that my blindness would prevent me from building the board.  But I have a lot more confidence in myself now, so I am revisiting this project.  However, since that time I have changed computers.  I am now runnning Windows 10, which will not allow me to use Protel version 2.7 any longer, so I am starting over using KiCad, which is better because it is in the public domain.  Also, I decided to make the cheezy boad a little less cheezy.  It will sport an ATmega1284 (128K flash, 16K ram) processor as well as a state variable filter and my new Blackmer VCA.  It will still have the cheezy drum circuits that were designed by John Simonton.

   

   
And this is what the PC Board will look like.



Back in October 2014:

It has been a while since I have done a synth project.  I am taking a crack at making a fairly simple synthesizer module board.  My main goal here is to keep the parts count low.  The only part that I am not going to compromise much on is the VCO.  But, the rest, I am going to try and be as minimalizst as I can.  And, in a way, I am going  to be doing a blast from the past, in some ways.  The first synthesizer  filter I ever built and used was the PAiA 2720 band pass filter.  This filter used a twin T notch filter in a feed back loop of a transistor amplifier to make a band pass filter.  It was tuned with a single diode in one of the T networks..  Well, I am going to do something similar.  Rather than use a diode to tune the thing, I thought I would try an OTA to do the tuning.. Other than that, it is pretty similar to the original Paia version from back in 1973.

    This board is going to have an ATxMega64A3 (or ATxMega128A3 or ATxMega256A3) micro controller to handle the Midi Interface and all on board logic.  This part has two analog outputs.  One output will be used to control the pitch, the other output is going to be connected to a midi control channel, such as the pitch bender, key pressure or what ever.  This board will also have two VCOs that will have my active temperature compensation and will generate either a ramp or variable pulse (PWM).  The other signal source will be  a noise generator.  I haven't decided yet weather to use a junction noise generator (breakdown of a emitter base junction) or a 32 bit pseudo random noise generator.  There will be one filter, as described above, one ADSR, and a VCA.  There will also be a glide connected to the pitch input of the VCO.   And finally, a LFO that will generate a triangle wave will provide modulation..

    The VCF will have a frequency control, as well as a control voltage attenuator for the ADSR and a control attenuator for the Midi Control signal.  The control attenuator has the property and the gain will be variable from -1 to +1 gain.  This way, the control voltages can make the pitch either go up or down.  No polarity switches required.  The same control attenuators are also on the Glide, VCO and PWM.

    At this time, the schematic is still not done.  However, here is a sneak peek at the filter.



January 19, 2012

    The schematic is not quite done yet, but it is done enough, I think, to post.  Keep in mind this is a minimalist implementation.  The schematic is only 32 pages long.  Well, maybe it is not quite as minimalist as I would have hoped, but I am afraid that this is about as minimalist as I can get.  The project also had a bit of feature creep.  After seeing the Korg Monotribe in action, I decided to add some analog drums to the thing.  I mean, what the heck.  The drums are sort of an expansion of the Paia drum oscilator that you can find on their web page.  The Cymbal was tough.  When I checked google to find out how other people implement cymbals, I discovered that most everybody punted and used samples.  But this is supposed to be cheezy.  So I opted for filtered noise.  I am using a ATxMega A3 processor on the board to do the Midi, Sequencer and on board logic as well as interfacing for an at this time undefined user interface.  I am planning to use an I2C LCD display, along with LEDs and push button switches.

Anyway, hopefully I will soon get all the details filled in on the schematic.  Keep in mind, this was drawn by somebody that is legally blind.  If you do see something that looks real weird, let me know, because I probably just didn't see it.  But you will find a bunch of places where there are no component values, no power supply terminals, etc.  I am going through the thing page by page filling in those details.

Feb 7, 2012

    Progress is slow.  It looks like the schematic for the synth board is just about done.  It is generating a BOM and a NetList now.  And, I think I have just about most, if not all, of the details filled in.  I am still trying to locate parts for the board.  It is sort of silly, but connectors are proving to be difficult for me to find.  As of today, I have updated the schematic.  I will hopefully soon be able to post the ORCAD files for this.

Preliminary Schematic for a Cheezy Synth Module Board  (updated May 14, 2012)

March 21, 2012

    Progress has been slow on this project.  I have been doing the PCB layout now for quite a while.  I am still in the phase where I am placing parts on the board.  This has always been the most difficult part of a PC board project, but man, it is really tedious now.  While plaing the parts I have found several omissions   and boo-boos.  These have been fixed.  The schematic is still not quite complete yet.

    Also on the drawing board is the PC board that will handle the user interface functions for this board.  That is still pretty much in the works, and will probably get posted soon.

April 5, 2012

    I have started laying out the voice card.  This started nearly a month ago, but I finally got all the parts placed and am now starting to route the board.


**UPDATE  The above picture is pretty much the final version.  Posted July 1, 2013

As you can see, it is going to be sizable.  10x8 inches (80 square inches).  I will be updating this photo from time to time.

May 14, 2012

    Well, I have changed my mind a bit.  Originally, I was going to use an ATxMegaA3 part for the cpu.  I decided to keep the same basic architecture across several projects, so, I have reverted back to the ATmgea2561.  Another reason is to potentially make this project Arduino compatible.  This will also allow me to add a couple more control voltage outputs as well.  The midi controller is going to use my new library, so it is going to look a bit strange.

December 18, 2012

    Well, I haven't changed my mind since March yet.  I have almost got all the componentws replaced on the board.  I have posted the latest schematic.  I discovered last night that there was a twenty pin part on the board that was doing next to nothing.  It was the address latch for the external memory bus.  I made a slight change to the  circuit, and was able to eliminate that part.  That was a big help.

July 1, 2013

    I am itching to get this done.  Yesterday, I went through the schematic and fixed all the problems I could find.  There were a few to say the least.  Such as, Omitted a couple of resistors, a couple of opamps had their inputs swapped, and a few signals were not taken out to a connector.  I have added a lot of stuff to the silkscreen as well.  Gonna sit on it again for a few days, while I do other cleanup, such as checking pad sizes, and hole sizes.

October 16,. 2014

I am still working on this project, but not at the rate I would like to.  I am also working on the NOOTA project right now, and I have given that a higher priority.  However, if I am stuck on the NOOTA, I work a bit on this one.  I am doing another redesign of the board, AGAIN!  And I am also analyzing the circuits so that others will have an easier time making mods to my design.  The first circuit I am doing is the Twin T filters.  The Twin T is not an easy circuit to analyze.  And I had to back off a bit and start out with a special case for the twin T whiich is not actually very useful for us.  This is the case where R1=R2=R3/2 and C1 = C2 = 2C3.  Now that I have that case done, I will work on the math for when R3 and C3 are arbitrary.  The analysis can be found here.


Anyway, here are the first file releases.  This is an open source project.

    Orcad 9.2 Schematic File
    Protel 2.7 PCB file
    Schematic file (PDF format)


All for now.