ECE4803B: Theory and Design of Music Synthesizers

Final Project – Fall 2006

For the project,
you will design and build a synthesizer module to expand our existing modular
system. You will be allowed to
use existing schematics you find from the web and elsewhere, but you cannot
just build a given schematic verbatim;
you must provide an interesting “twist” to the
design or designs you borrow from. In particular,
you can create a design that combines elements from two different existing designs
(the
“crossbreed”), makes a fundamental change in an existing design (the “mutant”),
or both (the “crossbreed mutant”).
You must be sure to give credit to
the original inspirations of your designs wherever
possible. If you
eventually consider marketing your design (say, starting a business
to sell boards), karma demands that you discuss sharing a portion of your profits
with the original sources of your design. (As mentioned in class, schematics
can be
copyrighted,
but circuits can’t, so if a circuit isn’t patented, you can legally
use it
and sell it however you like – but what’s morally the right thing to do
is a different story.)

You aren’t constrained to use any existing designs if you don’t want to. If you
are feeling adventurous and want to
try making something
up out of thin air, or some combination of thin air and existing
designs, that’s
OK. I’m allowing you to start from existing designs since most of the
synth
designers
I know of actually got their start from tinkering around and modifying older
designers.

Your module must either make audio (i.e. an oscillator) or process audio
(i.e. a filter or nonlinear waveshaper) or sometimes both (i.e. a filter that
self-resonates when the Q is turned up really high).
Since this will be a component of a synthesizer, some aspect of your module
should be controllable via a control voltage. (For instance, a filter that’s
whose cutoff is solely controlled by a pot won’t do.)

Here is an extensive
list of
project ideas.
You shouldn’t by any means feel restricted to pick something on
that list (which I’ll keep adding to as random ideas strike me);
you can use the list to
get an idea of the sort of things I’m looking for.

On working with a partner: I tried this last semester (I gave people
the option of working either solo or in pairs), and it worked
OK in some instances, but not so well in others. I wound up sometimes seeing
just one of the two people in lab all the time, so I had little idea of how
much the other person actually knew or contributed to the project.
Although I’m hesitant, I will
entertain groups of two if the project is particularly ambitious and if
there’s a clear division of labor where each member of the team is responsible
for designing and debugging specific parts of a larger circuit. It’s important
that such potential teams talk to me early and make a case for their idea.

Collaboration policy: People are strongly encouraged to help each other
out, both with bantering around design ideas and particularly with debugging.
Often, the
mere act of trying to explain what’s on the breadboard to someone else will cause
you to discover the source of a bug. You can draft people from outside the
class to help you debug too. Feel free solicit advice on online forums (the
SDIY list that I’m going to make you join will be particularly helpful.)
The
only caveat on all this is that you must cite and thank the people who aided you
and how they aided you in your report. (Also, don’t be a complete idiot,
i.e. don’t try to con someone online into designing the entire circuit for
you.)

Timeline: How about the following tentative timeline:

By Nov 3, send me an e-mail about what you want to do. I want to have a
diversity
of projects, so if someone tells me they want to do a project someone has
already
picked, I’ll probably suggest they try
something else (or suggest some additional twist).
First come, first serve, more or less. Put “4803 project idea” in your e-mail
header
so I can easily sort through them. The sooner you get this to me, the
better; I have an encyclopedic knowledge of obscure synthesizer related
tidbits (you may have noticed), and I can probably suggest some things to
look at for you to get a head-start.
By Nov 10, get me a draft schematic (hand-drawn is fine – I can run your
schematic through the digital sender),
which I’ll post on a website and invite members
of the SDIY list and the rest of the
class to provide feedback and suggestions
By Nov 17, show me something put together on the beadboard.
By Dec 1, show me something working on the breadboard. (Notice I left two
weeks between “put together” and “working.” Debugging circuits is tricky; it
requires a detective’s mind! There may be a problem in the hookup on the
breadboard, or there may be a problem with the underlying design in the
schematic, and you never know which is the case at first.)
By Dec 8, show me a working module “in solder,” i.e. a piece of hardware we
can add to our existing systems.
You
can put this together on perfboard,
stripboard, or, if you’re ambitious and want to try putting together an
actual PC board,
we can look into getting some boards made. I’d only recommend the latter
option if
you already
happen to familiar with a PCB layout program, since they typically have
a steep learning curve.

Documentation:
In addition to the project, I want a brief user’s manual (in electronic
form – e-mailing is fine) explaining
what your module is,
what the
controls do, etc., and a few basic notes about how you came up with the
design, any calculations you made (for instance, to get something
to match the MOTM standards), etc. This should be reasonably slick, i.e. the
text should not be
handwritten, although I will accept scans of hand-drawn schematics. I’d like to
emphasize the word “brief.” I am more interested in PRODUCT
than I am in PROSE. Do not write something lengthy, since
I will not have time to read it anyway. The user’s manuals will be posted on
an archival website. Many people may build your circuit!

Due date, sort of:
The above timeline is just a guideline to help you gauge your progress; I won’t
be carrying a specific checkoff list. When is the project due? Well, I will
accept boards and reports up to and including the weekend after finals; I
need to have grades figured out and turned in by noon the following Monday.
However, you should treat Dec 8 as the deadline so that you can focus on
studying for your finals.
Do not complain to me or complain on evaluation form about me “making”
you work on projects during finals week. I’m giving you that extra week as a
just-in-case “s*** happens” buffer – it’s up to you to make sure you won’t need
it.

Specifications: Your module should satisfy the MOTM standards
(taken from the MOTM website):

1/4″ jacks for connections (if you look in
the second-from-top drawer in the
set of long drawers in the middle of 4006 lab, you’ll find a bunch of jacks)
1 volt-per-octave response for VCOs and filters (you may want to include
an additional linear input, if appropriate)
Positive-going GATE voltages (+1.5 V threshold)
Positive-going triggers
Audio levels of 10 V peak-to-peak (i.e., -5 V to +5 V)
VCAs respond from 0 to +5 volts
Use the standard
MOTM power connectors. There’s two kinds, depending on what you
need. The modules you built all used a four-pin connector. If the long tab
is to the
“right,” then reading from top to bottom, the MOTM power supply gives -15 V DC,
two ground pins, and +15 V DC. (The MOTM website says these are “standard
AMP MTA-156 4 position” connectors.)
If you are using some digital logic (TTL, CMOS, whatever),
you can use the six-pin connector. The bottom four pins are the same as the
for the four-pin connector, and the topmost pin gives +5 V DC
and a “digital ground”
(so you can try to keep digital switching noise away from
the “analog” part of your circuit.) The ECE lab staff stock these; ask in
the stockroom. Note that the connectors don’t fit into the standard
perf board; the easiest solution seem to be to run some short wires from the
board to the connector.

Parts:
See
Aaron’s stocklist
for a list of specialized synth-related parts I have personally stockpiled.
If you want something to
experiment with or put in
your final design, just ask. A lot of more standard
parts, including pots and
TL08x type op amps (which are a good workhorse audio
op amp) are
available form the ECE. If you need some other part that I don’t currently
have, let me know and I’ll see if I can get it.

Grading:
Your project grade will be based on how impressed I am.

I consider
the project to be the most important thing in the class; hence, your course
grade will max out at whatever your project grade is, e.g., if you do B work
in the exams and homeworks but turn in an A project, you might get an A
for the class,
or you might get a B; but if you do A work on everything else but turn in
B level
project, your grade won’t be an A.

There will be no slacking on the final project.