mike@artsvax.UUCP (Michael Czeiszperger) (10/09/87)
We recently purchased a SB180 development board from a company
called MicroMint. The company is owned by the person who
writes the Circuit Cellar in BYTE. After reading a column about
a SCSI project using this board, we decided to use the
board to make a SCSI controller for a music product. To make
the product work correctly, we relied on the information from
MicroMint about the board, which indicated that it would
perform DMA operations from memory to it's SCSI port.
When the board arrived, the documentation and schematics that
came with it were poorly done, and in some cases, just plain incorrect.
The schematics break every rule in the book, and are extremely hard
to follow. The rule seems to be "mismatch logic signals where ever
possible". Top that off with several wrong pin numbers and
mislabed signal lines, and you get the complete picture.
Even so, we decided to go ahead with the design, because despite
the problems with documentation, it still appeared that the
board would function correctly.
After spending 3 months writing code, (part time), we discovered
that the DMA just won't work with the SCSI chip. (NCR5380)
We called MicroMint, and they told us, "Oh, no one has ever
been able to get that to work". They then recommended we
call the 'users group'. The users group turned out to be a
guy who wanted a consulting fee to discuss our problem.
After painstakingly going over the schematics, we found several differences
between what was on the schematic, and what was actually on the board.
There were also a couple of design mistakes in the 5380 interface. After
doing a fair amount of hardware redesign, and placing jumpers over
the mistakes in the circuitboard layout, we got the DMA to the 5380
to work in limited circumstances. It is still not working as expected,
but we're working on it.
I think that someone who plugs his own product in BYTE should
plainly give the limitations, as well as the advantages of his
product. BYTE should be responsible to see that no conflict
of interest exists between their writers, and the products
they recommend. It is unthinkable that something like this should
happen, and I will never, ever believe at face value anything
that is published in that magazine.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Michael S. Czeiszperger | Disclaimer: "Sorry, I'm all out of pith"
Systems Programmer I | Smail: Room 406 Baker (614)
College of the Arts | 1971 Neil Avenue 292-
Computer Lab | Columbus, OH 43210 0895
The Ohio State University | UUCP: {decvax,ucbvax}!cbosgd!osupyr!artsvax!mike
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~