RE: VSM troubleshooting, Asteroids

John,
   I'd love to see what you have, I'd even volunteer to assemble
any pieces into a pdf.
I've used the test switch as well with simple scope patterns to
work threw the state machine, but I've also had problems with the
vsm communicating with the main cpu(halt/go) and no test pattern
shows. Thats where manually loading the vsm rams with commands
helps as I can force a pattern to see where the vsm is within the
code. I'd like to see both in a doc.

If I get some time today, I'm going to try to mess with the code
for the vsm test pattern, and at least document the commands.

ToMW

> -----Original Message-----
> From: owner-vectorlist@synthcom.com
> [mailto:owner-vectorlist@synthcom.com]On Behalf Of
> John Robertson
> Sent: Tuesday, February 12, 2002 11:34 PM
> To: vectorlist@synthcom.com; vectorlist@synthcom.com
> Subject: Re: VECTOR: VSM troubleshooting, Asteroids
>
>
> If we are talking of the vector generator area here, I
> did create a
> signature map of it using the simple idea of turning
> on the TEST switch.
> This creates a repeating pattern that you can then
> grab with the SA and use
> for future troubleshooting. Works for me, and sooner
> or later I'll get
> around to posting the list. It is a continuation of
> the short SA list on
> Asteroids and does make it much easier to troubleshoot
> the vector generator.
>
> John :-#)#
>
> At 10:29 PM 12/02/2002 -0600, TomW wrote:
> >Would it be possible to cause a reset the counter=0
> at a specific point ?
> >What I'm thinking is loading the vram with valid - no
> draws. Then Step one,
> >let the counter point to the first address, and
> reset, we check that addr
> >line for activity, any others should only pulse on
> reset. Then let it
> >count/point to address 2, check the address line for
> activity, etc. Its not
> >the cleanest but in theory it should allow us to
> verify the program counter
> >is sequencing correctly. I've got a huntron tracker
> 2000 that helped me find
> >one, but it seems to catch me off guard.
> >
> >If I can find the bottom of my workbench, I'm going
> to have to start working
> >on this again.
> >TomW
> >
> > > Vector Generator Program Counter
> > >
> > > Yes I had an issue with this on a board too,
> turned out to be a bad
> >74LS367,
> > > a logic comparator found this but I was trying to
> find the problem with
> >the
> > > 670s & 193s as I new the program counter wasn't
> working properly....
> > >
> > > Scale is an output of DDMA7-DDM4 clocked by
> !LATCH3 + !LATCH0 * !ALPHANUM
> > >
> > > So round & round we go, if the 74LS42 Decoder
> could be removed &
> >controlled
> > > then I could step thru various chips of the state
> machine one at a time?
> > >
> > > Kev
> > >
> > >
> > >
> > > > Not having any of that in front of me.. ;-) I
> should have said
> > > > the vector
> > > > state stack... Top left hand side of the
> schecmatic... Darn,
> > > > can't remember
> > > > what it was 'offically labeled'. Yeah thats it,
> loaded the ram ok
> > > > and the stack had
> > > > the offests into ram. If the pattern was all
> localized, the stack
> > > > never got full enought
> > > > to excersie it. So in game mode it would die, in
> test mode it was
> > > > fine..
> > > >
> > > > Also the the whole thing needed to excersize the
> Scale0-3 as
> > > > well.
> > > >
> > > > TomW
> > > >
> > > >
> > > > >
> > > > >
> > > > > So the 9010A didn't catch the Vector Ram failure?
> > > > >
> > > > > Curiosity forced me to pull the schematics, the RAM
> > > > > failure should be caught
> > > > > by the 9010A, unless you have a problem with one of
> > > > > the 273 latches
> > > > > (right?).
> > > > >
> > > > > The 3 signals that didn't work properly for me using
> > > > > the Flukescript were
> > > > >
> > > > > !OPTS
> > > > > !SINP0
> > > > > !SINP1
> > > > >
> > > > >
> > > > > Kev
> > > > >
> > > > > > It puts a medium sized PLUS on the screen. I jumper
> > > > > the watchdog
> > > > > > and let if fly. I was going to put scope patterns in
> > > > > a doc for
> > > > > > the various test points during the draw but never
> > > > > got around to
> > > > > > it. What I'd like to do is increase the scaling, and
> > > > > have the vsm
> > > > > > code put in three different places in the vsm ram.
> > > > > I've run into
> > > > > > a few cases that it ran fine, but only because it
> > > > > was located low
> > > > > > in the vsm ram (its such a small pgm it does not
> > > > > fully test the
> > > > > > vsm address space). The game would die, but
> the vsm was fine.
> > > > > > Turns out that one of the high order bits (I forget
> > > > > the chip) was
> > > > > > out in the vsm ram. So to make this run in low vsm
> > > > > and then again
> > > > > > in high vsm memory space would be nice. Did anyone
> > > > > ever document
> > > > > > the vsm command structure ? That would be helpfull,
> > > > > this PLUS is
> > > > > > from the atari catbox documentation and I just
> > > > > adapted the tests
> > > > > > to the 9010. On of these day's I'll make it pretty..
> > > > > >
> > > > > > TomW
> > > > >
> > > > >
> > > > > -------------------------------------------------------
> > > > > --------------------
> > > > > ** To UNSUBSCRIBE from vectorlist, send a message with
> > > > > "UNSUBSCRIBE" in the
> > > > > ** message body to vectorlist-request@synthcom.com.
> > > > > Please direct other
> > > > > ** questions, comments, or problems to
> neil@synthcom.com.
> > > > >
> > > >
> > >
> > >
> -------------------------------------------------------
> -------------------
> > > -
> > > > ** To UNSUBSCRIBE from vectorlist, send a
> message with "UNSUBSCRIBE" in
> > > the
> > > > ** message body to
> vectorlist-request@synthcom.com. Please direct other
> > > > ** questions, comments, or problems to neil@synthcom.com.
> > >
> > >
> -------------------------------------------------------
> -------------------
> >-
> > > ** To UNSUBSCRIBE from vectorlist, send a message
> with "UNSUBSCRIBE" in
> >the
> > > ** message body to
> vectorlist-request@synthcom.com. Please direct other
> > > ** questions, comments, or problems to neil@synthcom.com.
> > >
> >
> >------------------------------------------------------
> ---------------------
> >** To UNSUBSCRIBE from vectorlist, send a message
> with "UNSUBSCRIBE" in the
> >** message body to vectorlist-request@synthcom.com.
> Please direct other
> >** questions, comments, or problems to neil@synthcom.com.
>
> -------------------------------------------------------
> --------------------
> ** To UNSUBSCRIBE from vectorlist, send a message with
> "UNSUBSCRIBE" in the
> ** message body to vectorlist-request@synthcom.com.
> Please direct other
> ** questions, comments, or problems to neil@synthcom.com.
>

---------------------------------------------------------------------------
** To UNSUBSCRIBE from vectorlist, send a message with "UNSUBSCRIBE" in the
** message body to vectorlist-request@synthcom.com. Please direct other
** questions, comments, or problems to neil@synthcom.com.
Received on Wed Feb 13 06:48:15 2002