Mathbox Theory... Help!

Okay, Im being good and asking this on Rasterlist since my question is
for I,Robot but I will repost on vectorlist if I don't get any good
answers... J

Im getting an error in the I,Robot Mathbox, the specific error is 'Bad
Y' which apparently in I,Robot the mathbox has X,Y,R and C functions (I
don't know what those mean tho) which is probably exactly the same as
Star Wars!!??

I have done all the basics..

1. ROM and RAM are all known good.

2. Replaced all Mathbox PROM's with known good ones

3. Swapped out theMathbox Bit Slice Processors with known good
ones

4. Made sure I don't have any trace damage on the PCB.

So, I think Im going to have to dive into the TTL stuff now and I really
don't have much understanding about how the mathbox works or how to
start troubleshooting it.

Here is what I 'know' now..

1. The mathbox does 3-D vector calculations (rotations)

2. It uses common RAM with the CPU so the CPU can load up data and
then perform calculations on that data.

3. Im sure it works like the Vector Generator in that the RAM gets
loaded up with a handful of needed calculations, then the mathbox is set
to 'Go' and it runs until it is done, then it tells the CPU that it is
done. The CPU reads the data and then sets up the next batch of
instructions.\

Since Im just getting a 'Y' function error in self test, then I am
either very lucky or there is something really bad and it is only
showing me a Y Error since it is happens to be checked first. My
assumption is that the Self test routine is doing a known set of
variables and the Y axis data is wrong in the resulting calculation?

Does anyone have any insight on how to start testing this in more depth?
Im fairly limited in my test bench tools... logic probe, oscilloscope,
VOM, hammer

I guess I do have the luxury of a working boardset too, just thought of
that as an A/B exam via logic probe (maybe looking for non-pulsing
pins)

Anyway... any suggestions would be helpful!! Thanks!

Jess

---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Sun Sep 8 00:46:33 2013