Up: Executive Summary
For me, DACS is a project of longer-term than an MQP can provide. The MQP
has provided a solid foundation upon which the remaining unimplemented
firmware and software designs may stand. Even though the project did not
meet the initial goals to the letter, it still seems that they were well
met. Given the scope of the work completed on the project, and the degree
to which it is complete, this makes sense.
DACS is a project that, at least in concept, began for me over two years ago.
Considering the potential I see for this system, it is quite likely that
development will continue beyond the scope of the MQP. The technology of DACS
is viable, affordable, and directly applicable to many applications. Over
time, it is hoped that it can be evolved into a marketable product.
The following is a list of known problems or incomplete tasks:
(note: not all of these may make sense until this design document has
been read through.)
This may seem like a large list, but given the sheer number of things that
went correctly, it is a relatively small list. These items are among
the first that will be addressed as the project progresses beyond being
- There is a noise floor approximately 50dB down in the mixer unit.
This has not been investigated thoroughly, but it is suspected that it may
be due to a grounding issue with the SSM2163 mixer chips.
- Digital ``hash'' noise can be heard in the outputs of the mixer unit
during periods of heavy microcontroller activity. This is likely due to the
complete lack of shielding in the unit.
- The bus combining/switching logic in the mixer unit is noisy during
switching periods. This is heard as a loud ``pop'' on the outputs of the
unit. It is suspected that this is due to the large switching currents
used in the relays, and/or residual DC offsets between buses.
- The mis-designed input trim section requires a re-design and engineering
fix. This is a relatively important piece of functionality that is missing
in the prototype mixer unit.
- The microcontrollers chosen for the task may not be sufficiently
powerful. Early tests are inconclusive, but it is a distinct possibility
that a more powerful microcontroller needs to be used.
- The firmware and software are still at a relatively immature stage,
from an implementation point of view. For the system to truly be useful,
these areas need to be addressed.
- The control board requires another LCD, three buttons, and a data
wheel to be mounted. Painting and silkscreened artwork also need to be
Up: Executive Summary
Table of Contents
[Whole document in PDF 1.9MB]
[more photos and information]