ComputerVoltageSources

I have uploaded into the Files>Hardware section my set of schematics. 
I did not post these previously because I did not originally generate 
them.  I only took other's ideas and draft schematics and enhanced 
them with ideas from this group and some of my own. I believe everyone 
is in agreement that they should be posted now.  I would like to thank 
Grant Richter, Harry Bissell, John Loffink and others who have posted 
comments and made input.  I also want to thank John Loffink for his 
contributions to the LCD_Support module.

I originally built all portions of these schematics to validate them. 
I purchased an AtomPro28 to evaluate and understand it since 
BasicMicro still has not released a datasheet.  I also purchased many 
of the parts to prototype and validate functionality such as the I2C 
interface.  I also purchased a front panel to verify the LCD_Support 
module mounting.

Like many of you, I wanted an advanced module and decided the only way 
to get one was to build it myself. Since I already has well over $200 
invested, I decided to complet a PCB design over Christmas and built 
it last month. I have written a 16 step sequencer program (1762 source 
lines) that uses 75% of the program memory. I have spent hours with a 
scope and protocol analyzer determining how the I2C and serial ports 
really work.  Details of my module are documented on my web site at
http://modularsynthesis.com/cvs/cvs.htm

Having completed my module, I am now on to other projects. One of the 
many is a remote I2C interface with a keypad and additional controls 
for improved user interface.  There are several videos on my website. 
I would suggest you view the videos of the LCD display. It adds 
awesome capabilities to the module and is why I am now designing a 
better user interface.

My PCB is not for sale. I designed a 4x6" PCB with a ground plane and 
SMT components. There are over 50 wires to the front panel and another 
50+ wires on the front panel. The ground plane clearance is 0.012 
inches. The PCB is very difficult to solder and once all the wires are 
connected it is nearly impossible to access for troubleshooting.

I would, however, be happy to offer suggestions to anyone who wants to 
carry forward a PCB design that is more reasonable to build.  Here is 
my initial list of suggestions:

1. Decide on a group consensus for the feature set. This particular 
set of features cost over $500 in parts alone.

2. Decide on a panel and do a PCB for the jacks and controls. 
Soldering over 100 wires was not fun nor practical for may DIY'ers.

3. Decide on an appropriate DAC. I chose the DAC7715U since it was 
less expensive than the DAC8420 with a savings of ~$40.00.  This DAC 
is in shortage and distributors are not expected to have parts until 
after August 2007.  You may also find that stock has been depleted on 
the DAC8420 as well. You probably want to use the DAC8420 so you can 
get it in the DIP package. Laying out dual package styles is not 
really feasible.

4. The LED drivers are simple current sinks and I used amber LEDs.  
Basically they are on, dim, or off so they only tell you that an 
output is doing something. This portion of the design could be 
improved.

5. My power supply noise comes right through into the input buffers 
when viewed on a scope. I can see digital ambiguity of +/- 1 count on 
my A/D conversions. This could be due to the noise, the inexpensive 
potentiometers, or the accuracy of the H8/3664 processor.  I can't 
troubleshoot it any further because I can't access my board (due to 
all the wires). You might consider adding 12 volt regulators for the 
op amps. I adapted my software to input average the A/D conversions 
over the last four cycles and is reasonably stable.

6. The display module is a must-have (watch the videos). The I2C works 
very well. The hardware interrupt serial support works very well for 
MIDI. Having programmable MIDI capability enables a whole new set of 
features and having serial communications enables synchronizing 
multiple modules (there is a video of this as well).

7. Standardize on a design that has software compatibility. The 
advantage of a group design is the ability to build upon other's 
software contributions. Anyone wanting a 1 to 16 step sequencer has my 
application as a starting point.

8. Use my analog input layout for a starting point. I compressed the 
circuit as small as possible with standard power buses so I could step 
and repeat this for all the channels.

Best of luck to anyone who volunteers to complete the design of a PCB. 
I suggest you be the final "voice" in determining the feature set.

Dave