finalproject_sp06.html

ECE4803B: Theory and Design of Music Synthesizers

Final Project

See
Aaron’s stocklist
for a list of synth-related parts I have stockpiled. If you want something to
experiment with or put in your final design, just ask.
I don’t want to just leave them out in the main ECE4006 lab,
since it’d be tragic if a naive student in one of the ECE4006
wasted the special qualities of one of these parts when a lesser part would do.


For the final project, you can work by yourself, or with a partner – it’s up to you.
If you work with a partner,
you must explain in your proposal why it makes sense to do so, i.e. what a natural
division of labor is – and I don’t want it to be “person A builds the circuit,
person B
writes the report.” Something like “person A will build the expo converter, and person
B will build the triangle core, and we’ll share in the writing responsibility”
will make me happier.
Although I don’t have a specific rubric in mind right now, I expect the projects
of two-person teams to be more ambitious than that of one-person teams.


I might entertain the possibility of three-person teams,
but that must be for an unusually ambitious project for which
such a large team would make sense. You will have to convince me it’s a good idea,
and there’s a reason it’s so complex it needs three people.


I don’t want any teams with four or more people – our experience with the MOTM kits
(and what I’ve heard from in 4006) is that winds
up with too many cooks in the kitchen.


Different teams are strongly encouraged to help each other out.
In particular, feel free to help each other debug. 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!


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


Please choose one of the following options.

Option 1: A Genetic Crossbreed Mutant Module

If you choose this option,
you will design and build a module to expand our existing modular
systems that you have already helped to build. 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 interest “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.)


BTW, 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.


Here is an extensive list of
project ideas for Option 1
.
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.


How about the following timeline for Option 1:

  • By April 14, 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.

  • By April 21, 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 April 28, show me something working, or at least put together, on the breadboard
  • By May 5, 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.

    In addition to the project, I want a brief user’s manual 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.” For this option, 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!

If you are in the Atlanta section, your module should satisfy the MOTM standards
(taken from the MOTM website):

  • 1/4″ jacks for connections
  • 1 volt-per-octave response for VCOs and filters (you may want to include
    an additional linear input, if appropriate)

  • Positive-going GATE voltages (+1.5V threshold)
  • Positive-going triggers
  • Audio levels of 10 V peak-to-peak (i.e., -5 to +5)
  • 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.)
    I recall James telling me they had those kinds of connectors;
    if not, let me know, and
    I’ll pick some up.

If you are in the Savannah section, your module should satisfy the PAiA 9700 series
standards, i.e. use 1/8″ jacks for connections… Unfortunately, I don’t have
the instructions to the PAiA kit, so I’m not sure what standards it uses – please
refer to the documentation you have. Ping me if the documentation seems unclear.

Option 2: Detailed Laboratory Study

In this option, you will conduct a detailed comparison of different parts and/or
circuits for basic synth subsystems (such as OTAs, linear-to-exponential
voltage-to-current converters, etc), and make recommendations regarding price vs.
performance of different options. You will need to figure out what the important
performance metrics for your task are, and formulate and document a detailed
and reproducable experimental protocol to measure those metrics.

Here is an extensive list of
project ideas for Option 2
.
As with Option 1, 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.

How about the following timeline for Option 2:

  • By April 14, send me an e-mail about what you want to do. As with Option 1,
    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.

  • By April 21, write up your suggested experimental protocol and e-mail it to me
    (plain text would
    be easiest to manipulate at this stage),
    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 “experimental protocol,” I mean what you’re going to measure, how you’re going
    to measure it, etc.

  • By April 28, show me your raw data.
  • By May 5, give me a final report. It can get right to the point – i.e.
    what parameters you measured, how you measured them, why you measured them,
    what you found, and what you concluded. (For instance, I have zero interest in an
    explanation of why OTAs are useful in synths, since I already know that. I want to
    learn something new!) Try to keep the exposition short and to the point, as I won’t
    have time to read anything overly long.

    This should be a reasonably slick document that I will be proud to put on an archival
    website. Many people will benefit from your results!