[sdiy] Teaching our Senior Design class this Fall

[Tom Farrand]

Dr. L,

Mission 1 Possibility: Dawn of the Dead
Design a top-octave generator using a Xilinx FPGA.  Handout an MK50240
data sheet (the dead) as the basis for the design criteria.  Who has
the patience to grind out a divider chain design where some of the
divisors are an odd number ... but a 50% duty cycle output is
required?  (That eliminates simply doubling the reference clock and
tacking on a /2 at the end of each chain.)  This could be the basis
for a string synth.  Could also be productized for the ambitious.

Mission 2 Possibility: Revenge of FFT
Design waveshaping circuitry for harmonic control for Mission #1
project.  Could include DSP for animation (or a JH-like BBD animator)
if appropriate for the audience.  Maybe implement Walsh functions in
another assemblage of FPGAs?  Did I mention presets?  Yeah, I want
preset capability with a method to store and recall those presets.
Presets must survive power loss and long-term storage.  Look at
stinged instrument spectra and figure out how you are going to get
similar results in a limited timeframe.

Mission 3 Possibility: Guitar Interface from Hell
Synths also process audio from other sources.  This guitar interface
to a synth is different ... it is digital.  I want to see the audio
from the the pickups digitized to at least 16 and preferably 24-bit
data at a sample rate of at least CD quality.  The output will be a
clocked bit-stream that still is usable at the end of a 20' cable.
The power for this interface will be provided by a source of 5V at
600mA maximum.  Piece of cake, right?  Well, there is one more
requirement ... the 5V power you get comes from AC mains voltage -
that can be dangerous if something fails.  The interface, located
inside the guitar, must be insulated from the power source by a
barrier that provides a minimum of 1500V RMS of isolation.  And the
output data must also be isolated by the same amount.  We don't want
to electrocute the next-gen Joe Satriani.


Off the top of my head that's about all I can come with.  I think
these exercise are fairly representative of real-world engineering.
One other requirement I would add for all projects henceforth, is
design using RoHS components only.  That is reality and has been for
some time.  Build them using leaded solder of course (because
lead-free solder sucks) but design for RoHS compliance.  Assembly for
RoHS is a production issue for the board stuffers so don't torture
yourself, but everything today must be RoHS.

Just my $0.02

Tom Farrand



On Sun, Aug 17, 2008 at 1:54 AM, Aaron Lanterman
<lanterma at ece.gatech.edu> wrote:
>
> Howdy gang,
>
> I'm signed up to run a section of our Senior Design class this Fall. This will be my first time doing Senior Design. The class splits up into teams of 4 or 5 students, which then each try to design and build something interesting. The class has a lot of structure, with various proposals and progress reports that are graded both by me and our staff that teaches technical writing.
>
> Guess what my section will be doing. ;)
>
> The class will have students from a wide variety of backgrounds - a few will have taken my "Electronics for Music Synthesis" class, but not many. So I'll need to be sure to steer people towards projects that could be reasonably completed in a semester by teams that size.
>
> My own main interest is in synthesizers - I'm thinking of a computer-controlled analog-signal-path sort, monophonic for simplicity - but alternative interfaces also tickle my fancy.
>
> Anyway, if anyone has any wild suggestions for undergrads to tackle, let me know. ;)
>
> - Aaron
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy