Atari/Atari Games Memos and Status Reports 1990 Jed Margolin
___________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Hardware Status
Dt: January 4, 1990
DSP32 Board
-----------
When we get the prototype boards in, the DSP32C PQFPs will have to be attached.
We have two DSP32Cs. I propose that we arrange with Manufacturing to have the
DSP32Cs attached by the company that we will use for production.
Because the DSP32C has 164 pins on 25 mil centers (fine pitch) I believe
they should only be attached by a company that uses an automated system.
I would like to be present at the vendor when these protoypes ICs are attached
so I can develop a feeling for SMD.
Jed
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Hardware Status
Dt: January 5, 1990
DS II
-----
I have received the two stuffed boards and will test them when I have time.
320E14
------
From Jeff Twombly:
The current version of the 320E14 is 2.0 . There are several bugs in it but
we could probably use it anyway, although only at 20MHz instead of 25MHz.
If we wanted production quantities of version 2.0 we could probably get them
if we ordered them soon. They would cost about $35.
They are finishing version 3.0 now and expect samples in April with
production quantities in June. They are ceramic EPROM versions.
OTPs are not scheduled until 3Q '90.
We could, instead, use the 320E15, which is available now for $30 in ceramic
EPROM. The 320E15 has the same pinout as the 320C10 (40 pin DIP) but has
256 words of RAM and 4K words of EPROM.
(The 320E17 has a host interface but costs more than the 320E15.)
Jeff says the reason the 320E14 has so many bugs is that they got it as
part of the license agreement with General Instrument (now Microchip
Technology) and GI did a poor job.
320E14 2.0 April production $35. order soon
3.0 April samples
June production (ceramic EPROM)
OTP not until 3Q '90
320E15 available now $30.
Jed
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Things I am Doing
Dt: January 8, 1990
1. DS II: Bring up the board;
Writing Test Routines.
2. DSP32 Get board built when it comes in (with 164 pin SMD);
Bring up board;
Write test routines.
3. 320___ Design subsystem and do a wire wrap board;
Put it on the next DSP32 board;
Write Test routines.
4. Driver Self-Test: Rewrite Self-Test to incorporate new DSP32 Tests,
latest ADSP Tests, and change character size to 8x8.
5. Compact Self-Test: Rewrite Self-Test to incorporate new DSP32 Tests.
6. Street Driver - Rewrite Self-Test to incorporate DS II Board.
7. Driver Sound: Documentation for Hard Drivin' 90.
Question for Analog Devices about the 2101:
If a Bus Request comes along and the 2101 is stalled because it is
doing an access to external memory, will the Serial Autobuffer
also be stalled even if its autobuffer is in internal RAM?
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Things I am Doing
Dt: January 12, 1990
1. DSP32 Get DSP32C mounted onto the board;
Get rest of board built;
Bring up board;
Write test routines.
2. DS II: Bring up the board;
Writing Test Routines.
3. 320___ Design subsystem and do a wire wrap board;
Put it on the next DSP32 board;
Write Test routines.
4. Driver Self-Test: Rewrite Self-Test to incorporate new DSP32 Tests,
latest ADSP Tests, and change character size to 8x8.
5. Compact Self-Test: Rewrite Self-Test to incorporate new DSP32 Tests.
6. Street Driver - Rewrite Self-Test to incorporate DS II Board.
7. Driver Sound: Release Documentation for Rev F (with 27256s).
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Status Report
Dt: January 19, 1990
DSP32: We have had two boards built, I will start testing them next
week (Jan 22). Without a PC I will only be able to write
routines run by the 68010 through the host interface.
DS II: I am bringing up the board and writing test routines.
I have been able to have the 68010 test the 2101's memory.
Writing 2101 code is difficult because I do not have a PC.
The VAX tools only support the 2100 which means that I cannot
simulate the 2101's extra functions. Since there is also no
emulator, it makes for a lot of wasted time.
320E15: I will complete the design next week (Jan 22).
Presumably, I will write some Test routines.
DS II RAM: We have received the Rev 1 boards and will have some built.
Driver Sound: We have officially released Rev F (with 27256s).
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Status Report
Dt: January 26, 1990
DS II
-----
I have written several test routines and the board seems to work.
1. Either the 2101 manual is incorrect or the 2101 has a bug in it that
causes the DO UNTIL counter to be loaded with the wrong value. Instead
of being loaded with N, it gets loaded with INT(N/2) + 1, so that,
for example, 8192 only does 4097 interations.
2. I need to change two resistor values to adjust the levels.
3. We need to hook two speakers up to make sure the amps don't oscillate.
(They are ok with just one, which is all I have.)
4. I need to program a sinewave sweep to evaluate the filters.
DSP32
-----
I am almost ready to start testing it.
320E15
------
I designed the circuit and had a WW made. I will test it right after the
DSP32 board. We need the 320E15 adapter socket in order to program it.
DS II RAM
---------
We had some built. I will test them when I have time.
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Status Report
Dt: February 2, 1990
320P15
------
40 pin DIPs in ceramic windowed packages are available now.
40 pin DIPs in plastic OTPs will not be available until 3rd Q.
44 pin Ceramic Quad windowed packages are available now. (They
are ceramic but are mechanically compatible with PLCCs.)
44 pin PLCC OTPs are expected to be available in TMX in April
with TMS in May. (They expect to be able to supply us with
enough TMX parts to meet our needs in April.)
In order for the die to fit in a PLCC package, they had to do a
die shrink. They expect to get the first parts Friday, and, by next
week, find out if they work. If they don't work it is unlikely they
will be able to supply PLCCs in April.
TMS320E15FZL 44 pin, Ceramic Quad, windowed
TMS320P15FNL 44 pin, PLCC, OTP
TMS320E15JDL 40 pin DIP, windowed
TMS320P15NL 40 pin DIP, Plastic, OTP
DS II
-----
I lost several days' work last week because the 2101 has a bug in it that
Analog Devices did not inform me about. Contributing to the waste of my time
was the failure of Atari Games to supply me with a PC to run the 2101 tools.
This forced me to use the 2100 tools on the VAX. When the 2101 program failed
to work properly I spent (wasted) a lot of time trying to figure out if the
problem was because of the tools.
DSK Board
---------
I have started writing the test routines.
320E15
------
I designed the circuit and had a WW made. I will test it right after the
DSK board. I will ask Karen to make a programming adapter socket for the
40 pin DIP.
Since we will be using the PLCC package, Jeff has changed our order to
PLCCs for the parts as well as the programming adapter socket.
I got them to send me the schematic for the R Bit programmer. It is
very primitive. The programming pulse width is determined by how long
you press the switch.
DS II RAM
---------
We had some built. I will test them when I have time.
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: February 9, 1990
Slapstics
---------
We have been assigned Slapstic 115 for the DSK Board. This is in addition to
Slapstic 117 which is assigned to the MultiSync Board.
As far as getting a few from another series for special purposes you need to
see Don and Chris.
DSK Board
---------
I have written several test routines and I believe the board works. I have
turned one over to Max.
The other board also came up with no problem.
Mike Felice has given us two more DSP32Cs.
Manufacturing Meeting on Tuesday, Feb 6
---------------------------------------
Attendees: Jim Freitas, Norm Maeder, Pat McCarthy, Tom Clark, and myself.
Results: In order to collect data on the usefulness of doing burn-in, they
will do the following test on 100 Klax boards:
a. Build the boards;
b. Test the boards;
c. Repair failed boards and log failures before going to the next step.
d. Burn-in the boards using Erik Peterson's test rack that allows the
video to be switched to each board during the test. The boards will
be powered by supplies other than the ones currently with the ovens.
e. Test all boards again and log failures.
The object of the exercise is to determine how many early failures were
found by burn-in. (Atari has never tested boards before burn-in.)
Depending on the results of the test they will either continue 100%
burn-in or go to 10% sampling. Either way, they will get rid of the
old burn-in power supplies.
2 of 2
320P15
------
We have received eight windowed units in the ceramic PLCC package.
Tom Clark has found a nice burn-in socket for me and will take care of the
production sockets. According to his research, the ceramic parts really do
need a different socket than the PLCCs or they pop out, unless we use a
universal socket that comes with a clamp.
Assembly has made for me a programming adaptor socket for the 40 pin DIPs.
I have started writing software to test the 320E15 WireWrap board. The
VAX tools that Mike Albaugh wrote for me do not work anymore. Dave Shepperd
is working on it but I have lost critical time that cannot be regained.
In addition, the lack of a simulator or an emulator that I can run makes it
impossible to debug code. Also, I was not able to program a part on the
IMI. I was able to program one of the two 320E15s on the Data I/O after
Shepperd fixed a bug in the Data I/O VAX program.
DS II
-----
We are limited to one working board because we have only one 2101.
Margot says that the real 2101s will not be available until June and
that even the phony parts won't be available in PLCC until April-June.
It seems they are going to redesign the memory section because their yield is
so poor. The parts they got from their own fab were as bad as the parts they
got from Intel which apparently fabs a lot of their chips.
I asked her to change our order from 8 PGAs and 2 PLCCs to 10 PGAs.
DS II RAM
---------
We had some built. I will test them if I have time.
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: February 16, 1990
320P15
------
On Monday I talked to Jeff. It seems that the wafer with the die-shrink
for the Plastic package did come out of the oven last week but were not probed
because nobody thought it was important. Jeff has vowed to build a fire under
them.
I have gotten an interface test program running on the 320 that tests the
interface timing. In order to get the Checksum program running I will need
a Simulator.
DSK Board
---------
I have given Leon the inputs for the 320 section.
The burn-in socket has a different footprint than the production socket. We
will lay out the board for the production socket and figure out a kludge
for the burn-in socket.
DS II
-----
Margot is working on getting us more X-rated 2101s.
DS II RAM
---------
We had some built. I will test them if I have time.
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: February 16, 1990
320P15
------
On Monday I talked to Jeff. It seems that the wafer with the die-shrink
for the Plastic package did come out of the oven last week but was not probed
because nobody thought it was important. Jeff has vowed to build a fire under
them.
It turns out that they are doing the E17s first.
We received the CLCC Programming Socket Adaptor, but no R-Bit Programmer. Jeff
is working on it.
I have gotten an interface test program running on the 320 that tests the
interface timing. The 320 assember that Albaugh wrote ran under MAC65 and
no longer works. Shepperd has converted it to MAC68K. Unfortunately, it can
no longer tack itself to the end of an existing 68010 program; its address must
be explicitly stated. This does not matter for the DSK Board, but if it were
necessary to redo Self-Test on the Driver Sound Board it would be a problem.
There was a problem in programming the E15 on the Data I/O. It seems that
even when you give it a Family Code (in this case for a 27C64) the Data I/O
tries to get the Manufacturer's device code from the part. The E15 does not
support the Manufacturer's device code so the Data I/O would report it as an
error and would refuse to program the part. Shepperd fixed it. After being
programmed, the VAX reports an incorrect checksum, so it requires going into
local mode to get it.
This is for the Data I/O in the DIO Room. The Data I/O in the Alpha Lab refuses
to program it. It just times out.
The IMI in the DIO room might work if the socket locking bar didn't force the
E15 adaptor socket out of the IMI socket.
DSK Board
---------
I gave Leon the inputs for the 320 section. The first plots are bad because
on the DSP32C, SCI-CARDS connected all adjacent unconnected pins.
The burn-in socket for the E15 has a different footprint than the production
socket. We will lay out the board for the production socket and figure out a
kludge for the burn-in socket.
I will not approve the layout until I check it against the 44 pin PLCC socket
that I spent Friday getting. Tom was supposed to get me a socket but didn't.
2 of 2
DSK WireWrap
------------
I will mark up the latest SCI-CARDS schematic for WireWrap II.
DSP32C
------
Mike Felice has delivered a new DSP32C Data Book (January 90) and will get us
more to replace the old ones (December 88). We also have a bug sheet now.
DS II
-----
Margot is working on getting us more X-rated 2101s.
We are having more DS II boards stuffed.
DS II RAM
---------
We had some built. I will test them if I have time.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: February 23, 1990
320P15
------
No News
DSK Board
---------
We have film and Fab for Rev 2 and are ready to buy boards.
I have a 44 pin burn-in socket from AMP that has the same footprint as the
production socket (the rows line up).
DSP32C
------
In going from version F32 to F33 they fixed some bugs and introduced others.
We can live with the bugs in the F32 but not with the ones in the F33.
With the bug in the F33, host interface auto-increments do not work properly
without extra hardware.
The AVL listing had called out the F33 part. I have asked Tom Clark to change
it to F32 and have also told Mary Burnias.
DS II
-----
We have had more DS II boards stuffed. I will start bringing them up
when we get more 2101s.
Other
-----
I have submitted a CER for the 27C010 (1 Meg EPROM).
I have started working on the part numbers for Hard Drivin' The Race.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: March 5, 1990
DSP32C
------
I have asked Tom Clark to disqualify the F33 part. (The F32 is on the AVL.)
DSK Board
---------
Rev 2 boards are due in this week.
I have gotten two more 44 pin PLCC burn-in sockets from the AMP salesman. He
will try to get me more, later this week. He also gave me an extraction tool.
This tool appears to be superior to the Burndy tool we are now using and
will probably extract the device without breaking off the corners like the
Burndy tool does. We sould buy more. Also some for the 68 pin. (Different size
PLCCs require different size extraction tools.)
320P15
------
No News. No RBIT Programmer yet.
DS II
-----
We have had more DS II boards stuffed. I will start bringing them up when we
get more 2101s.
Part Numbers
------------
I have received two new programmed part series. One is for Hard Drivin' The
Race (Cockpit) and one is for Hard Drivin' The Race (Compact).
I have prepared a list of proposed programmed part numbers which I will show
to Rick Owens before giving them to Art.
Other
-----
I have started updating the Driver Main Self-Test to include DSK tests. I have
also converted the characters from 16x16 to 8x8.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: March 9, 1990
Hard Drivin' The Race Part Numbers
----------------------------------
I reviewed the part numbers with Jeff Bell, made some appropriate changes,
and gave the list to Art.
I gave Art the corrected parts list for the DSK Board.
I gave Art the list of MSP parts that will come out for The Race.
DSK WW Board
------------
Have received the board.
ADSP Graphics RAM WW Boards
---------------------------
Emmette is making two more ADSP Graphics RAM Boards for the Eds.
DSK PC Board
------------
Have received ten Rev 2 boards and are having 8 stuffed.
Once they are built and verified we can release Rev A.
320P15
------
No News. No RBIT Programmer yet.
DS II
-----
We are still waiting for more 2101s.
Other
-----
I have given Max the first preliminary version of the new Driver Self-Test.
There is lots more to do.
Boards with Rev D Fab are Assembly Rev E and are completely ok to use.
Boards with Rev C Fab are Assembly Rev D and need a Mod. The Mod we did
in production was not the mod we incorporated into Rev D Fab.
(The production Mod was a lot easier to do and worked almost as well the one
incorporated on Rev D Fab.)
I believe the Production Mod was:
Change 80S from 74LS04 to 74F04.
Change R89 from 1K Ohm to 470 Ohm.
Use only 74LS244 by Texas Instruments at 20K.
Both these boards use only 27C512s for sound wave ROMs:
136052-1123 5A Programmed EPROM, 137448-200, 27C512-200
136052-1124 20A Programmed EPROM, 137448-200, 27C512-200
136052-3125 30A Programmed EPROM, 137448-200, 27C512-200
136052-1126 45A Programmed EPROM, 137448-200, 27C512-200
The board is documented for 250 ns EPROMs in these positions but MANMAN shows
none in stock. There are, however, 42335 200 ns parts in stock.
The program ROMs are still:
136052-3121 45N Programmed 27C256, 200 ns
136052-3122 70N Programmed 27C256, 200 ns
Jed
_____________________________________________________________________________
To: Don Wrightnour
Fr: Jed Margolin
Re: Driver Sound
Dt: March 12, 1990
Driver Sound Rev E Fab (Rev F Parts List):
Build to the released documentation, no mods required.
Driver Sound Rev D Fab (Rev E Parts List):
Build to the released documentation, no mods required.
Driver Sound Rev C Fab (Rev D Parts List):
A deviation was issued to:
Change 80S from 74LS04 to 74F04.
Change R89 from 1K Ohm to 470 Ohm.
Use only 74LS244 by Texas Instruments at 20K.
The Mod we did in production was not the mod we incorporated into Rev D Fab.
(The production Mod was a lot easier to do and worked almost as well the one
incorporated on Rev D Fab.)
This Mod is NOT required Rev D and Rev E Fabs. (But if it is done, the boards
will still work.)
Jed
cc: Rick Moncrief
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: March 16, 1990
DSK PC Board
------------
We have had one Rev 2 board built up. It passes the test programs I have
written for it. We need to decide how to verify it so we can release Rev A.
ADSP Graphics RAM WW Boards
---------------------------
Emmette made two more ADSP Graphics RAM Boards for the Eds.
One worked right off, the other worked after a minor correction.
Hard Drivin' The Race Part Numbers
----------------------------------
I have reviewed Art's parts lists, made corrections, and returned them.
Production
----------
1. Production has some Rev C and Rev D Board Fabs for Driver Sound. I explained
to Don Wrightnour what the differences are and what documentation applied to
which board.
2. On the MultiSync Board I called out the Opto State PROM to be 35 ns.
Production found some old parts (bought in October '88 under the old 37-
number) that were 50 ns. I verified the circuit timing and told Jim Freitas
that 50 ns parts were ok (and signed the deviation).
Tom Clark has sent a memo warning of the dangers of using old parts. I told
Jim Freitas that the Programmer that burns the parts also verifies them and
I believed that if the parts verify, then they are good. Unless the parts
were stored in a high humidity and/or high temperature environment (they
weren't) they should still be ok. Presumably some PROMS from that batch
were used in games and are presumably still working in games that have been
powered up for 18 hours a day, which is harder on the part than it being in
storage.
3. I went over the list of parts to be scrapped and found several that we could
use, especially the VRAMs (quantity 1706).
DSK WW Board
------------
Have received the board.
320P15
------
No News. No RBIT Programmer yet.
DS II
-----
We are still waiting for more 2101s.
Other
-----
I am continuing to work on the new version of Driver Self-Test.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: March 23, 1990
DSK PC Board
------------
The Rev 2 board appears to be totally working. I have given Art the
corrections for the Silkscreen and the parts list in order to release Rev A.
DSK WW Board
------------
The board works so far.
320P15
------
Stephanie, Max, and I converted my VAX 320 program to PC format, decided how
to integrate it with Stephanie's code, and figured out how to use the new
software.
We received the RBIT Programmer. It is very primitive.
DS II
-----
We received more 2101s along with a new bug sheet. Margot says they have
new totally working working parts but will not give us (or anyone else) any
until June.
I have fired up DS II #2 and it works.
Other
-----
I am continuing to work on the new version of Driver Self-Test.
Aces
----
I updated the ACES' Self-Test to include the DUART Test and the display of
the 12 Bit A/D in the CONTROLs screen.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: March 30, 1990
DSK PC Board
------------
I have signed off the DSK Board.
DSK WW Board
------------
The board works and I have turned it over to Max.
Mike Felice have given us a ME-DSP32C-R33 (the one without the bugs). The one
we have is a 'BL' that supposedly contains the bugs (although it works in the
board).
I have done new GALs for the separate DSK and E15 WW Boards so they can
coexist on the bus for development.
320P15
------
I will start the RAM Test for Internal RAM soon.
DS II
-----
I have a working board set consisting of a DS II Board and a DS II RAM
Board. I have documented each one appropriately for Brad.
We have now received the ten 2101s that we ordered.
Max and I have implemented the system wherein the 21010 Serial Download
can be selected from the DS II Menu. It will download from a PC and
start the 2101.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 6 April, 1990
DSK Board
---------
Mike Felice have given us a ME-DSP32C-R33 (the one without the bugs). We
need to try it to see if it works. The one we have is a 'BL' that supposedly
contains the bugs (although it works in the board).
Slapstics
---------
According to Don Pauuw we will probably get 3,000 - 5,000 good Slapstics
after they are packaged. Someone needs to make sure they get packaged on time.
320P15
------
I am almost done with the RAM test for Internal RAM. The next step is to
integrate it into Self-Test. Then comes the P15 programmer.
DS II
-----
Erik has attached the DS II RAM board to the DS II Board with appropriate
connectors and Max and I have worked out the software so that after
selecting the Download option from the Self-Test Menu one can download
code from a PC to the 2101 and run it.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 12 April, 1990
Pat Brosnan
-----------
Pat Brosnan sent me a fax about two MultiSync boards (one Stun Runner and
one Compact Driver) with problems he could not fix. I faxed him a list
of suggestions.
Harnesses
---------
Stevie is on vacation this week so I talked to Carole. She had already
documented the Ribbon Cable Harness but did not know that the cable from
Main Board to Sound Board needed to be longer. She promised to tell Stevie
when she comes back.
320P15
------
I have finished the RAM test for Internal RAM and integrated it into Self-
Self-Test. Next comes the P15 programmer.
Self-Test
---------
The memory test for DSP32C External Memory runs through its Host Interface
and is too slow. I need one that runs in the DSP32C.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 20 April, 1990
Harnesses
---------
Stevie has taken care of the documentation and released the harnesses.
320P15
------
I am almost done with the design of the P15 programmer. I expect to have one
section wirerapped next week to test the design. The PC Board version will
have four mostly independent sections.
Other
-----
I am getting more information from Direct Imaging on their PC Board Printer.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 27 April, 1990
320P15
------
I am done with the design of the P15 programmer and am doing the GALs.
I expect to have one section wirerapped next week to test the design. The
PC Board version will have four mostly independent sections, although it
could have eight if necessary.
I have started writing the programming package.
DSK Board
---------
I have given Art the inputs to change the 8Kx8 RAMs to 32Kx8 RAMs. This is
a parts list change only; the board was laid out to accept either one with
no mods or jumpers.
According to Art, Purchasing has only just now asked for the tape for the
DSK Board. (We now longer supply film to the board vendors; they make their
own. We do have film made, but it is for our own use, in Receiving.)
Other
-----
At Max's request I am having two ADSP II Graphics RAM WireWrap Boards made.
It seems that we had given ours away.
Cyclotron
---------
Bonnie is using a MultiSync Board with a Sitdown Motor Amp (which requires
a different software driver than the Compact) so I changed Self-Test and
gave her a new set of ROMs so she can test the motor.
She later complained that the CONTROLS Self-Test screen no longer worked with
the 8 bit A/D inputs. The problem turned out to be that the Strain Guage Amp
was installed and saturating the A/D inputs. I unplugged the amp.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 7 May, 1990
320P15
------
I have received the WW board and expect to start working with it this week.
I have started writing the programming package.
DSK Board
---------
I have written tests to track down the problems Max and Stephanie had with
Board #1. The board totally works on my system but does not totally work
on Max's system. Erik and I are working on it.
There is only one PAL Programmer and on Thursday it was broken.
Other
-----
At Max's request I am having two ADSP II Graphics RAM WireWrap Boards made.
On Thursday, Max and I met with Margot and Dan from Analog Devices. The
2101 with all the bugs is available now in production quantities. The
fixed version should be available in June. The 2105 should also be
available in June.
Jed
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Tom Clark
Dt: 14 May 1990
Components has reactivated 179250-044 for the socket for the 44 pin PLCC.
320P15
------
I have started working with the WW board and writing the programming package.
Other
-----
According to the AVL, the change in part numbers did not occur until April 18,
nearly two months after Tom's memo.
When I signed off on the DSK board on March 29, 179250-044 was still active.
(This transaction placed the AMP part in the inactive file)
VENDOR
PART PART TRANSACTION
NUMBER GENERIC NUMBER DESCRIPTION DATE
179250-044 Socket,IC,PLCC,44pin
AMP_INC 821575-1 3 18-Apr-1990
VENDOR
PART PART TRANSACTION
NUMBER GENERIC NUMBER DESCRIPTION DATE
179237-044 Socket,IC,PLCC,44pin
BURNDY QILE-44-P410T 0 7-Jul-1987
179237-044 Socket,IC,PLCC,44pin
ITT_CANNON_ELECTRIC LCS-44-02 0 7-Jul-1987
179237-044 Socket,IC,PLCC,44pin
JAPAN_AVIATION_ELECTRONICS PLPS-44-T 0 7-Jul-1987
179237-044 Socket,IC,PLCC,44pin
METHODE 213-044-001 0 7-Jul-1987
179237-044 Socket,IC,PLCC,44pin
AMP_INC 821575-1 0 18-Apr-1990
Jed
From: BERT::CLARK 27-FEB-1990 08:51:10.15
To: MCCARTHY,CLARK,HOFF,MARGOLIN
CC:
Subj: DUPLICATE PLCC PART NUMBERS
We are presently carrying three part numbers for PLCC sockets. they are
P/N 179252-XXX, 179250-XXX and 179237-XXX. Only, part number 179237-XXX shows
any WHERE-USED, while, part numbers 179237-XXX and 179252-XXX shows any stock.
I would like to inactivate for new designs part numbers 179250-XXX and 179252
-XXX. This would require that the stock for part number 179252-084 be sold off
or exchanged for credit. this stock is from 3M/TEXTOOL and needs a special
tool to remove the PLCC from the socket (Note: the sources for part number
179237 all use the same tool and 3M/TEXTOOL is changing over to the same style
socket). We presently have 1250 pieces in stock of the 84 pin socket. If we
want to use them we can, but remember they need a special tool.
On part number 179252; I believe that this was established for SHOWVIEW and
only called out the AMP sockets. After Geoff Barker and I ran shock tests for
Morgan, it was determined that other manufactures could be used in his
application. I would add AMP to part number 179237-XXX.
If everyone is in agreement, I would like to do this, this week. Any
questions, please feel free to contact me.
THANKS
TOM C
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Status Report
Dt: 15 May 1990
320P15
------
I have started working with the WW board and writing the programming package.
Components
----------
Components has reactivated 179250-044 for the socket for the 44 pin PLCC.
Other
-----
I met with Don Mullikin of TI and got the data sheet for the new 128Kx8 VRAM.
I also got the new hardbound 34020 manual.
The data sheet for the VRAM does not give enough detail to determine if it
can be easily retrofitted into Driver hardware. The reason is that the
34010 address bus is multiplexed for VRAMs with Row and Column addresses
separated by 8 bits. (64K x n requires 8 Row address bits and 8 Column
address bits. VRAM A0 is processor address A0 for /CAS and A8 for
/RAS.)
The 256K x 4 VRAMs are definitely not easily compatible since the address
requires 18 bits, 9 for /RAS and 9 for /CAS. VRAM A0 is processor address
A0 for /CAS and A9 for /RAS.
The question for the 128K x 8 is: how are the address lines apportioned
between /RAS and /CAS? If they are A0-A7 during /CAS and A0-A8 during /RAS
then it will work. Process address lines A0-A7 will go to VRAM A0-A7 during
/CAS; processor address lines A8-A16 will go to VRAM A0-A8 during /RAS.
The address line will be 8 bits apart.
Note: The 128Kx8 VRAM will not be available until the end of this year.
The 34020 is neat. The address is available both as a complete bus and as
a multiplexed bus for VRAMs with the address offset controlled by an
internal register. It can be set up for offsets of 8, 9, 10, and 11 bits.
Jed
_____________________________________________________________________________
To: Rick Moncrief
Fr: Jed Margolin
Re: Status Report
Dt: 21 May 1990
DSK Board
---------
I have released the GALs (136077-1025 and 136077-1026) to Rick Owens.
I have given him actual programmed parts as well a 5 1/4" Floppy containing
the JEDEC files.
320P15
------
I have lost three days because Assembly was doing a 'Pilot Run' of
15 games for Gumball Rally and could not work on my board.
Now that I have it, I am almost done stabilizing the Power Supply
switching circuits.
DSP32C
------
Max has been using the new R33 and it seems to work.
The DSP32C has had its AVL entry fixed and I have had the F33 added to it.
(The R33 is in the PGA package used by our emulator; the F33 is the PQFP
that we will use in production.)
137661-080 AT&T WE-DSP32C-F32-080
137661-080 AT&T WE-DSP32C-F33-080
137661-100 AT&T WE-DSP32C-F32-100
137661-100 AT&T WE-DSP32C-F33-100
AVLNOTES will specify:
" This part to be marked with the Atari part number only"
Driver Interface
----------------
The Cable that Erik got at Fry's confirms Max's theory so I have given
the information to Stevie.
Driver Sound
------------
The 2Kx8 RAMs used the part number for the NMOS part (137211-001) in order
to use them up. We must have done so because I keep signing deviations
to use the CMOS part (137278-002). Therefore I have asked Art to ECN the
parts list to use the CMOS part.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 1 June 1990
GAL Programmed Parts
--------------------
The vendor does not have a 1.2MB Drive and cannot read the files on the disk
I made for her. I have made one formatted for 360KB ('FORMAT A: /4') but I
don't know if it will be successful. My DOS book says that even though a disk
can be formatted for 360KB on a 1.2MB drive it is possible that a 360KB drive
will not be able to read it.
320P15
------
I have successfully programmed a part and set the Security Bit.
After more testing I will mark the schematics to have four units on one
board and will work with PC to have the board done.
Driver Voltmeter
----------------
Self-Test will have the version of Line Calibration where the calibration
changes as long as you hold the switch down (instead of having to press
the switch for each change).
The voltage can be calibrated to a maximum of 237 volts.
Driver Interface
----------------
I am planning on adding a Link Connector Test to self-test.
Driver Manual
-------------
I have started documenting the changes to Self-Test for Jean.
2 of 2
DSK Board
---------
I have incorporated Max's DSK Test into Self-Test and Erik, Rick, and I
have started working on it.
The Problem:
When doing a Read from the MOSEL 32Kx8 RAMs, a large glitch is
generated that gets into the expansion bus and makes the GALs go crazy and
select addresses that it shouldn't. It does not occur on Writes.
The glitch is the result of the RAM sense amps turning on and
propagating through the F245 bus drivers.
So far we have found:
1. With slower GALs, the board will pass the test although the glitch
is still evident.
2. The test runs and the glitch goes away with 8Kx8 RAMs instead of the
32Kx8 MOSEL RAMs.
3. The problem also goes away with SONY 32Kx8 RAMs. The RAMS were 45 ns. I
don't know if SONY 85 ns parts are ok. The circuit should work
with parts as slow as 120ns. I chose 85 ns because there are already
four 85ns 32Kx8 RAMs on the board (on the DSP32C).
The bus lines are clean when there is no RAM plugged in.
The bus lines are clean when the Buffer control is disabled.
Plan: A. Replace F245s buffers with LS245s.
In order for the 170 ns EPROMs to work we will have to continue
using the fast GALs. Timing will still be close. I will do an
analysis.
If this does not work, the other things to try are:
B. Do not use MOSEL RAMs.
C. Use slower RAMs.
D. Use faster RAMs.
E. Get a job at TOGO's.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 3
Fr: Jed Margolin
Re: Status Report
Dt: 11 June 1990
DSK Board
---------
The Problem:
When doing a Read from the MOSEL 32Kx8 RAMs, a large 30ns glitch is
generated that gets into the expansion bus and makes the GALs go crazy and
select addresses that it shouldn't. The extraneous Reads reset the P15
Interface Flag.
It does not occur on Writes.
The glitch is the result of the RAM sense amps turning on and
propagating through the data bus drivers.
So far we have found:
1. The RAMs on the ADSP Board also generate a glitch but it does not affect
the bus address lines.
This suggests that the problem is related to PC Board layout. Doing a
new PC Board layout is not an option. (We have already bought boards.)
2. Wiring the data buffer inputs low does not result in bus noise. The
problem appears to be due to the glitch, not to a normal signal
transition.
3. The problem does not occur when using 45ns SONY RAMs (instead of the
85ns MOSEL RAMs) because the glitch is much shorter and does not make
it down to a logic '0'.
Solution:
1. I programmed the GAL to activate the RAMs on the early strobe
(/EXT) so that the effects of the glitch have gone by the time
the P15 signals (strobed by the late strobe) are activated.
2. I have also changed the GAL program so that it uses an input that
is normally tied high to its own 10K pullup resistor. When left
in this state (pulled high) the address decoding remains as it
was and is fully functional.
When this signal is connected to /ERD (Read, active low) it changes
the data buffer enable on Read accesses to ROM, RAM, and the P15
so that the buffer is enabled by the late address strobe (/RVAS)
instead of by the early address strobe (/EXT). This way when the
glitch from the 32Kx8 RAMs occurrs it does not get through the
data buffers.
status report, 11 June 1990 2 of 3
The ZeroPower RAMs also produce a glitch from the sense amps and
since these RAMs are slower, some of the glitch does continue
into the late address strobe. The noise from this glitch
appears to be within acceptable limits.
Reads from the ROMs and the P15 interface are not affected, since
the data buffer is enabled long before the data is required by the
68010.
Plan A: 1. I will ECN the DSK Board to change the F245 data buffers to LS245s.
2. The GALs have the new program and will probably work with no
further action.
Plan B: If there is a problem, we will have Production add the single jumper
which activates the change in data buffer timing.
Plan C: We can use faster 32Kx8 RAMs. (The glitch from the sense amps is
too fast to generate a logic '0').
Plan D: We can change from the CMOS GALs to Bipolar PALs (PAL20L8) if the
GALs cause problems like they have in the past.
GAL Programmed Parts
--------------------
The vendor does not have a 1.2MB Drive and cannot read the files on the disk
I made for her. I have made one formatted for 360KB ('FORMAT A: /4') but I
don't know if it will be successful. My DOS book says that even though a disk
can be formatted for 360KB on a 1.2MB drive it is possible that a 360KB drive
will not be able to read it. If there is a problem I will put the files on
a 3 1/2" floppy (Joan says they have a 3 1/2"drive.)
I have given the 5 1/4" floppy to Rick Owens. The plan is that Joan will make
some GALs from the files and Rick will verify them.
Since they will be working from the JEDEC files supposedly the programmer will
use the test vectors in it to test the programmed device. When we supply only
a device to be copied they can only copy the GAL programming, they cannot do
a functional test on the programmed device. They will apparently also set the
security bit.
status report, 11 June 1990 3 of 3
Driver Interface
----------------
I have added a Link Connector Test to self-test.
Driver Manual
-------------
I have documented the changes to Self-Test for Jean.
To Do
-----
I will do more tests on the P15 Programmer; then give it to PC.
While they are doing it I will:
1. Finish the new Driver Self-Test;
2. Write the P15 Programmer Code.
I will try to squeeze in a Motor Amp test during the automated self-test.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 18 June 1990
DSK Board
---------
1. I have ECNed the DSK Board to change the F245 data buffers to LS245s.
2. The GALs have the new program and I have verified the parts from the
vendor.
3. If there is a problem, we will have Production add the single jumper
which activates the change in data buffer timing.
Driver Self-Test
----------------
I have finished the North American version except I may need to change the
Program ROM checksum switches depending on how many ROMs we end up using.
I have put in a simple automated Motor Amp test at the end of the automated
self-test.
Driver Manual
-------------
I have been working with Jean on the changes to the manual. The Company
has underestimated the amount of time required and Jean will not be able to
make all the changes.
X Programmer
------------
I have gone over the schematics, made corrections, and they were waiting
for Leon when he came to work Monday. I did it over the weekend.
To Do
-----
Work with Leon and Joe on the PC Board for the X Programmer.
Write the X Programmer Code.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 25 June 1990
DSK Board
---------
Production is having ASIC 61 mounted.
Driver Self-Test
----------------
I have finished the North American version except I may need to change the
Program ROM checksum switches depending on how many ROMs we end up using.
Driver Manual
-------------
I have not heard from Andrea about my last corrections.
X Programmer
------------
I received schematics and PCB plots, checked them over the weekend, made a few
changes and corrections, and had them ready for Leon on Monday morning.
On Monday afternoon I approved his changes and asked him to send out for film.
I have set up the menu structure and am filling it in. The menu includes
hardware tests so I will be able to bring up the PC Board faster.
Other
-----
I will start the slapstic encryption program as soon as I have time.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 10 July 1990
DSK Board
---------
We have received some Rev A boards from production.
They split the order between two different vendors.
The boards are not the same. I am not thrilled with either one. Both
have exposed traces not covered by solder mask, only in different
places. Also, the alignment pads were done differently. One use a sold
pad, they other used a doughnut.
The boards do seem to work; however we have not tested the Helper
circuitry yet.
Cyclotron
---------
I received a call from Bonnie requesting help in reading the Opto-Coupler
inputs because she is using the Center input independently of the Quadrature
inputs. She also requested new self-test ROMs to deal with it.
I went over the code in self-test so she will be able to put the appropriate
code in her program. I explained that I would not have time to do new
self-test ROMs until August. I told her where Cyclotron self-test lives and
invited her to use it as she wished.
DSK GALs
--------
I was out of 15ns 20V8 GALs so I ran MANMAN to see if the programmed parts
had come in yet. They had not. They had not even been ordered. They were
not in MANMAN. They were not even on the AVL even though I released them
several weeks ago and have even approved programmed first-article parts from
Merit.
It turned out we ordered blank parts from Marshall and they have come in,
although they were rejected because the Lattice part number did not match
the part number on the AVL. Mary said she had submitted a CER to correct
the number but that no action had been taken. The parts will be sent to
a different vendor for programming which I find strange.
Thanks to the kind assistance from Erik (Durfey), Mary Burnias, Rob Bryant,
and Mark Hoendervoogt I have received the parts I need for the field test
units.
Driver Manual
-------------
I went over my last corrections with Andrea.
Status Report 7/10/90 2 of 2
X Programmer
------------
We received the boards and had two stuffed.
I have brought one up to the point where I have been able to test most of the
software, and I believe the hardware is working.
(The software totally works on the wire-wrap board.)
The problem is that the 3M Burn-in sockets DO NOT WORK. They do not make
reliable contact with the PLCC pins. I have tried a brand new Helper and it
is no better.
I can run a program where a data line is viewed on the 'scope and it goes
from floating to logic high depending on how the helper is pushed while it
is in the 3M socket.
The PLAN:
1. Determine if a Yamaichi socket will fit. If so, see if I can
locate and buy some. (TI used a Yamaichi on their program
adapter.)
2. Call the 3M rep and make him come over to look at it. Make him
feel really bad so he will buy the Yamaichi sockets for us.
(They are sure to be expensive. This assumes they will fit
and are available.)
3. Make an adaptor to use my two remaining AMP sockets. The problem is
that the board was layed out for the staggered rows used by the
3M (and Yamaichi) sockets. The AMP sockets have their rows
aligned. The other problem is that AMP stopped making them and I
only have two left. I will see if I can get more somehow. The AMP
salesman has always been very helpful and would probably make an
even greater effort to find them after I tell him the 3M sockets
don't work.
We might be able to remove the AMP burn-in sockets from existing
DSK boards and replace them with standard production sockets.
(I would really like to avoid doing this.)
Plan B: Program the parts one-at-a-time on the wire-wrap board.
Plan C: Program the parts on the EPROM Burner with the TI Adaptor socket.
Other
-----
I have done some preliminary work on the slapstic encryption program. I hope
I get time to work on it again sometime.
Jed
_____________________________________________________________________________
_____________________________________________________________________________
7-14-90
Erik,
On the board with the AMP socket,
1.PD15 seems to be shorted to 1.PA0 where 1.PD15
passes under 8C on its way to 11C pin 2.
Please see if you can fix it.
There may be more shorts under 8C.
I have added a new feature to the Program called 'DUMP' that
dumps information to the screen.
First Column Second Column Third Column
Word Address What the data Data read from
should be Device A
. . .
. . .
. . .
. . .
If the data read from Device A is correct, the entry is Green.
If it is not correct, it is Red.
Right coin increments the address,
Left Coin Decrements the address,
Start returns to the menu.
Jed
PS: The square collar around the AMP socket could short some pins when
it droops.
_____________________________________________________________________________
To: Art Jackson
Fr: Jed Margolin
Re: Sound Board ECN
Dt: 23 July 1990
Please ECN the parts list for Race Drivin' Sound PCB Top Parts list
A046491-02.
Change
Item 4 from: 136052-3121 1 IC, Prog 27256, 200ns 45N
to: 136077-1033 1 IC, Prog 27C512, 200ns 45N
Item 4 from: 136052-3122 1 IC, Prog 27256, 200ns 70N
to: 136077-1032 1 IC, Prog 27C512, 200ns 70N
Item 29 from: 179178-002 1 CONN, RCPT, 2CKT E1
to: 179178-002 1 CONN, RCPT, 2CKT E2
Thanks,
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 23 July 1990
Production
----------
I have given Jim Freitas and Jim Buchanan the updated test procedures
for the boards.
When I went over them with Jim Buchanan the ZRAMs on two DSK Boards
did not work. As I was preparing to MTO one so I could find out what
was wrong, Jim discovered that his +5V was low. Adjusting the supply
fixed the problem.
X PROGRAMMER
------------
3M sent three guys (Larry, Moe, and Curly) to investigate the problem with
the Quick-Lock socket. I was able to very effectively demonstrate that the
devices worked when they were first inserted into the socket but stopped
working once they had locked in.
1. After measuring E15s and P15s they were unable to find any
reason for it not to work.
2. They were consistently unable to reach anyone in their own
company because of their electronic messaging system.
3. They make a socket similar to the Yamaichi socket which Tom Clark
did not tell me about. This other 3M socket has the same footprint
as the Yamaichi which means it is not the same as the 3M
Quick-Lock. (They had to have a distributor fax the data sheet
because they couldn't reach anyone in their company.)
4. A few days later I received a fax from Chris Schmolze from 3M in
Austin. He claims the problem is that the socket was designed
for plastic parts (PLCC) and I was using ceramic parts.
a. I tried a plastic part and it failed, also.
b. There is no way, from reading the 3M and TI data sheets,
that I would have had any reason to believe that the
socket wouldn't work.
5. I left a message for Rick Putnam (3M sales rep) to this effect
but he has not returned my call.
6. The three wise men who visted me were actually Rick Putnam and
Neal Fouts from 3M in Foster City, and Byron Westerman from 3M
in Austin, Texas.
2 of 2
The board that we modified to use the AMP sockets seems to work, although
only three of the four sockets are useable. The fourth socket has
successfully resisted all efforts to find out why it will read parts
but won't program them.
I will try to have another board built-up and modified to use the Yamaichi
sockets.
Slapstic Encryption Program
---------------------------
I have written a specification for the program and have started implementing
it. I will test it as much as I can on the PC before installing on
the 68010 for a functional test.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 30 July 1990
X PROGRAMMER
------------
Out of the 170 devices I have programmed for production there have been
2 failures. Both appear to be due to address line problems in the
devices.
3M has taken back the Lock-In sockets and given me their version of the
Yamaichi socket. Their version is junk.
Emmette has mounted one of the Yamaichi sockets on the board that
originally had the 3M sockets. I haven't tried it yet.
Slapstic Encryption Program
---------------------------
I am now installing a simple version of the program on the 68010 in order
to perform a functional test.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 3 August 1990
X PROGRAMMER
------------
I have successfully programmed 617 devices which I have MTO'd to Production.
In the process there were 10 failures. Of these:
7 appear to be due to bad address lines;
1 was not blank;
1 could not set the RBIT;
1 could set and verify the RBIT, but the ROM could still be read!
Rob Bryant wants a total of 635 devices, so I need to make 18 more. There
are 468 parts still in inventory so we should MTO about 20.
We are due to receive another 2000 on September 13.
I have kept 10 blank parts for our use.
Slapstic Encryption Program
---------------------------
I have gotten the 68010 program working, exercising the slapstic with
encrypted data in the simple slapstic mode.
Next comes the dreaded shell mode.
To Do for Race Drivin' Release
------------------------------
1. Fix the ROM7 Checksum problem.
2. Get Max to add Watchdog to his Helper test routine.
3. Test everything once all the ROMs are properly checksummed.
4. Add the Slapstic encryption routine if possible.
5. Make sure all the ROMs get documented.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 10 August 1990
X PROGRAMMER
------------
I have asked Karen to finish the X Programmer board by installing the
Yamaichi sockets. (The board had originally been built with 3M sockets.)
I have given her the information she needs to buy more Yamaichi sockets.
This will give the company a backup board for programming Helpers.
Slapstic Encryption Program
---------------------------
I have gotten the 68010 program working, exercising the slapstic with
encrypted data in the complicated shell mode.
After discovering (with Don) and compensating for errors in the
Slapstic design, the program seems to work. I will let it run for
awhile.
To Do for Race Drivin' Release
------------------------------
1. Fix the ROM7 Checksum problem. [done]
2. Get Max to add Watchdog to his Helper test routine.
3. Test everything once all the ROMs are properly checksummed.
4. Add the Slapstic encryption routine if possible. [finished]
5. Make sure all the ROMs get documented.
Other: I paid $121.02 cash for the Yamaichi sockets on July 13 and have not
been paid by Atari Games.
I will never again loan Atari Games money in this fashion, no
matter how desperately the parts are needed.
Never ever.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 21 August 1990
X PROGRAMMER
------------
I have received the board from Karen.
Slapstic Encryption Program
---------------------------
In addtion to the complicated Slapstic encryption program, I have done a
simple Slapstic detection program that runs at Power-Up.
Race Drivin'
------------------------------
We have released the game.
Somehow, Slapstic did not appear on the Main Board parts list.
I have asked Art to ECN all the cockpit and compact versions to include it.
The Slapstic also did not appear on Ralph's kit parts list; there were
also several other omissions. I have asked him to ECN it.
The parts list for the DSK Board needed some work. I have asked Art to ECN it.
I have started the German version.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 27 August 1990
Race Drivin'
------------
The German Self-Test is complete, subject to finding out from our
German Distributor what he thinks of my translations.
Art has started the process to ECN all the cockpit and compact versions
to include Slapstic.
I have ECNed the DSK Board to its final Production configuration.
There is a new Self-Test (Version 2.1) that has a menu entry for
clearing the ZeroPower RAM. It is in [Margolin.GSPTST]FLTST_R2_1.vlda .
It should be assigned ROM Numbers of: 136077-2001 and 136077-2002 .
It should be used in all future North America, British, and Japanese versions.
The German version also contains the menu entry for clearing ZeroPower RAM.
It is in [Margolin.GSPTST]FLTST_G2_1.vlda and should be assigned ROM numbers
of 136077-1201 and 136077-1202 . Source code is in [Margolin.GSPTST.NEWGER] .
To Do
-----
1. Give Art the inputs for the Compact version of the DSK Board.
2. Ask production if they need more programmed helpers.
3. Ask production if they want me to look at the bad DSK boards.
4. If I have time I will work on the North American Compact version of
Self-Test.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 31 August 1990
Race Drivin'
------------
I have made the changes suggested by Nova-Apparate and now German Self-Test
is complete.
German Self-Test is in [Margolin.GSPTST]FLTST_G2_1.vlda and should be
assigned ROM numbers of 136077-1201 and 136077-1202 . Source code is in
[Margolin.GSPTST.NEWGER] .
There is a new version of Self-Test (Version 2.1) that should be used in
all new versions of North American, British, and Japanese.
It is in [Margolin.GSPTST]FLTST_R2_1.vlda and should be assigned
ROM Numbers of: 136077-2001 and 136077-2002 .
Production
----------
I have looked at two of the 21 bad DSK Boards from Production.
They both had the same symptoms:
ASIC 61 chip was alive and its internal RAM tested ok.
Two of the external RAMs tested bad even when replaced
by known good parts.
Apparently most of the 21 bad boards exhibit this problem.
Either ASIC 61 has a design weakness that makes it susceptible to this
kind of failure or there is a SMD attachment problem or there
is a PC Board problem.
In order to properly test the circuit I would need a modern logic
analyzer and several weeks to program it. I have given the boards back
to Production.
Don is going to send two of the boards to Fine Pitch to have the ASIC 61s
replaced. That should help decide whether to have the remainder of the
21 boards fixed or scrapped.
Jim Freitas wasn't sure if these boards had gone through the Zehntel Board
Tester. I explained that in the last few months I had experienced several cases
where I had boards with a trace bridge that interfered with circuit function
but which were cleared with the current from a buzz light. I asked him to run
the bad DSK boards through the Zehntel again.
I talked to Jim Buchanan who told me what he had found. With the RAMs unplugged,
data lines on a good board read 2 MegOhms. Some of the data lines on a bad
board read completely open. He determined that the ASIC 61 pins were properly
soldered and concluded that the ASIC 61 outputs were blown.
I agree.
2 of 3
We might be looking at a board handling problem. With most ICs, once they are
installed on a board they are pretty much protected from Electro-Static-
Discharge (ESD) because they are wired to other ICs, most of which also have
ESD protection. (There is safety in numbers.)
With ASIC 61, each external data line goes only to one external RAM pin.
Until the RAM is installed, the line to the ASIC 61 is floating naked and is
therefore very susceptible to ESD.
I recommend that the ASIC 61s that are removed be sent back to the manufacturer
for failure analysis.
I do not know how they handle boards after the ASIC 61 is attached but before
the external RAMs are installed. They may need to observe full ESD precautions
until the RAM is installed.
Cyclotron
---------
I have made new Self-Test ROMs for them with some new features they had
requested.
Bonnie had a problem where she could not get a board to run standalone.
There were two problems:
1. This MultiSync board with this AMS required the 20pF cap on the
Address Strobe line to eliminate the PEEK Errors.
2. Her program contained old Driver code that looked in the ZeroPower
RAM at the stored values for the control setup and if it was bad
would go to a control setup mode.
The MultiSync board that worked came from Driver Compact and still
had valid data in the ZeroPower RAM. Her new board came from
Production and had not been set up for the Driver.
The software detected that the control setup was wrong and tried
to setup Driver controls, which don't exist in Cylotron.
Patching out the routine solved the problem. Bonnie is going to
look at the program in more detail to make sure she doesn't need
this function.
NoName
------
I have made new Self-Test ROMs for them with some new features they had
requested.
Other
-----
I will be on vacation for September. I am scheduled to return on October 1.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 8 October 1990
Race Drivin'
------------
According to Ray Sherman the company does not have the R2R resistor networks,
the lead time is 9 weeks, and is there anything else we can use?.
I told him it is a custom part with two sources and that the only
alternative would be to build them, at enormous cost, out of 0.1% resistors.
He said that one source (Mepco/Centralab) disclaims any knowledge of the
part. I gave him copies of my files and a Centralab part.
Art is coming out with the parts lists with the new cockpit ROM numbers that
we have given him for North American, British, and German versions. Note that
this will change again for the Link feature.
I have made more PLD65s; Emmette will presumably do the rest. I have gotten
more labels from Jim Freitas.
DS II
-----
I have gotten the second board to download from the PC. It is ready for Matt.
Only one of the two 2101 Rev 3 parts works. The oscillator does not work on
the bad one; there may be more wrong with it also. I tried different crystals,
even 10 MHz, but to no avail.
It looks like the only way to tell the difference between rev 2 and rev 3 parts
is the date code. Both are marked ADSP-2101 G-50X. Rev 3 parts should have a
date "greater than 9034C" which should mean that rev 3 starts with 9035.
Other
-----
The pod on my system needs its target system to have a 33 pF capacitor on the
/AS line. This pod seems to get moved around alot because systems tend to not
work with it. Now we know why.
Next
----
Compact Driver Self-Test
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 15 October 1990
Race Drivin'
------------
I showed Emmette how to program the P15s and he did the remainder of the
325 parts.
Out of 325 parts we had 13 failures for a failure rate of 4%.
Ten of the 13 failures were from the last batch of 68.
In the first batch of 617 devices the failure rate was only 1.6%.
Both of these 'batches' are the original TI Pre-Production parts.
DS II
-----
I have given Matt a DS II board along with new Self-Test ROMs containing
the DS II downloader.
We received two more Rev 3 2101s from Margot.
The 2101 IC in the IC is labeled 'EZE-ICE' indicating that it is probably a
special part bonded out for the ICE, but it has a date code of 9028 which
normally means it is the old Rev 2 mask, the one with all the bugs.
If it is a Rev 2 part, the EZE-ICE will not be very useful.
Compact Driver Self-Test
------------------------
I am making progress.
Cyclotron Unwanted Helicopter Sound
------------------------------------
Brad's SA II Audio Board makes an unwanted Helicopter Sound during game play.
The sound is not really loud but once you key in to it you can hear it even
during game sounds.
When Farokh powered the SA II board with a separate +5v the problem went away.
The +5V line from the MultiSync Board has about 50 mv of noise on it. This is
normal for a MultiSync Board. In fact this is normal for anything powered by
a switching power supply.
The problem on the SA II board appears to originate in the OKI sound circuit.
(The OKI IC is connected directly to +5V.)
The real problem is that the SA II Board is a two layer PC Board. TWO LAYERS!
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 22 October 1990
Race Drivin'
------------
Some of the defective P15s appear to be due to a worn out socket on the
Programmer. I have brought up and installed the Programmer board with the
Yamaichi sockets.
DS II
-----
The 2101 on our EZE-ICE is indeed a Rev 2 part. We are supposed to be getting
a new ROM and a new 2101 for the EZE-ICE.
Note that the 2101 used on the EZE-ICE is a special part bonded out for
the EZE-ICE only. This is because:
1. There are extra signals bonded out for the EZE-ICE;
2. The EZE-ICE PC Board was laid out for the old PGA pinout and is
not compatible with the new PGA pinout. (The new PGA pinout is
compatible with the PLCC pinout; the old PGA pinout is not.)
The bottom of the EZE-ICE socket goes to a pin-changer adapter so it will
correctly plug into a standard 2101 socket.
The part used in the EZE-ICE could not, for example, be unplugged from
the EZE-ICE and plugged directly into a standard 2101 socket.
Likewise, a standard 2101 cannot be used on the EZE-ICE board.
2101s
-----
According to Margot:
Lead times for the 2105 are 12 weeks although we might be able to get
a small quantity before that.
We could get ten 2101s in 2-3 weeks and another ten 2-3 weeks after
that.
Supposedly, both parts are available in either PGA (KG) or PLCC (KP).
I suggest we order: (10) 2101 KG (PGA)
(10) 2101 KP (PLCC)
(20) 2105 KP (PLCC)
Compact Driver Self-Test
------------------------
I am making progress.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 29 October 1990
Race Drivin'
------------
I have checked the ROM Parts List for the North American Race Drivin' Kit.
When I asked Art to change the Compact Race Drivin' DSK Board to the -02
version, he accidentally changed the Cockpit version instead. He is working
on fixing it.
DS II
-----
We are waiting for a new ROM and a new 2101 (Rev 3) for the EZE-ICE.
Compact Driver Self-Test
------------------------
I am making progress.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 26 October 1990
Race Drivin'
------------
I have checked the ROM Parts List for the North American Race Drivin' Kit
and they are correct.
When I asked Art to change the Compact Race Drivin' DSK Board to the -02
version, he changed the Cockpit version instead. He is working on fixing it.
DS II
-----
We are waiting for a new ROM and a new 2101 (Rev 3) for the EZE-ICE.
I received a 2105 but it is a Rev 2 part which is basically worthless.
Compact Driver Self-Test
------------------------
I have finished the North American/British version and am working on the
German version.
Counterfeit Compact Hard Drivin'
-------------------------------
We received the board set and Erik hooked it up. There are several
interesting things about it.
1. The boards are all two layer.
2. It uses a Jamma connector.
3. Instead of ZeroPower RAM it uses regular RAM with an external
Battery.
4. They use different parts in the Color RAM and output circuit.
(The color RAM does not pass self-test.) It does not look like
they use 8 bits of RGB. It is not noticeable in the game.
5. The two 34010s are both -40 but are run at our clock speeds
(48 MHz. and 50 MHz.).
6. The ADSP Board is an ADSP II but uses the 2100 PGA instead of
PQFP. The date code of the 2100 is 9037 so it is not from the
batch 3-KOAM bought last year.
7. One set of Main Board Program ROMs was altered but the only
thing different we noticed is that the attract screen copyright
message says 1990. Our compact says 1988.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 2 November 1990
Race Drivin'
------------
I have signed off the Compact parts lists calling out the -02 version of
the DSK board.
I have generated the preliminary ROM Release document which can be found in
[MARGOLIN.RACE]MS_ROMREL.DOC
Self-Test ROMs for North American and German versions are ready to be released.
VRAMs
-----
Erik and I tested the Micron 100 ns VRAMs in a board and ran the VRAM Verify
Loop Test overnight. It ran 375 tests and there were no errors. (The VRAMs
also ran the game.)
According to the partial data sheet that I received on the NEC 64K*8 VRAMs,
they still do not support Shift-Register-To-Memory Transfers, making them
unuseable in the game. Erik installed some and they were, in fact, not
useable.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 9 November 1990
OrCad
-----
I talked to Rhonda in Purchasing who said she did order the software from
Elcor. I talked to Tanya from Elcor who said she had found that we had
ordered it from her and everything was fine.
The new version of OrCard is supposed to start shipping around the first week
in December. We will get it as part of the year's free update. People
buying the new version will pay $595. (We paid $495.)
Production EPROM problem
------------------------
Don brought over two of the thirteen ADSP II boards with the problem that they
passed Self-Test but would not run the game.
One board has a hardware problem. The other board had a bad EPROM.
1. It passed the ADSP checksum test, which does not run the EPROM at
full speed. EPROM Programmers also do not verify it at full speed.
2. It was plugged into a Main Board to run it at full speed. The
checksum was correct.
3. It passed the Verify Test in the EPROM Programmer against the
ROM master. As I mentioned, EPROM Programmers do not verify parts
at full speed.
4. The EPROM is, nonetheless, bad. Replacing it in the ADSP II allows
the board to run the game.
I suggested to Jim Buchanan that, if he could get the time, to separate out
the bad ADSP II boards that were bad because of bad EPROMs. This may help
determine how much of a problem it is.
Part of the problem may be due to the EPROMs that were sent out to be erased.
There does not appear to be a specification for erasing EPROMs. (Cleaning
adhesive off the window, erasure intensity and time, etc.) A simple blank
check is not sufficient for determining whether an EPROM has been properly
erased.
It has come out that the EPROMs going into the Race Drivin' kits are not
being tested in games. At the meeting with Bob Stewart you worked out
something with him, but I don't quite remember what it was because of the
really loud banging going on next store in Dan Van's new area.
2 of 2
74AS823s
--------
Because the 74AS823s are not available TI suggested we use their new
Bi-CMOS part the 74BCT29823NT).
The Bi-CMOS from TI (the 74BCT29823NT) not only works but produces
a slightly sharper picture on the monitor.
I have submitted a CER and when I get the Atari pn I will ask Art
to ECN the parts list for Compact and Cockpit to use this new part.
I have evaluated the Signetics 74F823 and it will also work. It is
as good as the 74AS823 but is not as good as the 74BCT29823. The
Signetics part is already on the AVL under
The preferred part will be the 74BCT29823NT. The 74AS823N (TI) and
74F823N (Signetics) can also be used if necessary.
I have informed Mary of this.
The 74BCT29823NT costs $2.50; the 74AS823 cost $1.70 .
Compact Race Drivin' Manual
-----------------------------------
I have reviewed, corrected, and returned to Publications the schematic package
for Compact Race Drivin' manual.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 3
Fr: Jed Margolin
Re: Status Report
Dt: 21 November 1990
Compact Self-Test
-----------------
Because we took off the Slapstic and two ROMs on the DSK Board, Auto Self-Test
reports a DSK ROM error.
I fixed it, added it to [Driver.Raceversions.ROMREVB], updated MS_AROMREL.DOC,
gave a copy of it to Rick Owens along with the new ROMs, and gave Joe a marked
up MultiSync North American Parts List and asked him to ECN it.
I also asked Joe to ECN the DSK -01 board, used for the Cockpit version.
This took most of Monday.
Cockpit Self-Test
-----------------
Manufacturing needed the Cockpit version right away so I did that, too. I
fixed it, added it to [Driver.Raceversions.ROMREVC], updated AROMREL.DOC,
gave a copy of it to Rick Owens along with the new ROMs, and gave Joe a marked
up Main North American Parts List and asked him to ECN it.
This took most of Tuesday. Joe was very helpful; Leon was not.
Self-Test at a Glance
---------------------
Race Drivin' Cockpit Self-Test ROMs 136077-3001 and 136077-3002 are used
in all North American and British versions:
Cockpit North American Released and ECNed
Cockpit North American Kit Released and ECNed
Cockpit British Released and ECNed
Cockpit British Kit Does not exist
Race Drivin' Compact Self-Test ROMs 136077-2001 and 136077-2002 are used
in all North American and British versions:
Compact North American Released and ECNed
Compact North American Kit Does not exist
Compact British Does not exist
Compact British Kit Does not exist
German Versions for Cockpit and Compact: Not Released.
2 of 3
Compact Opto Harness
--------------------
When Art's group documented the Compact Board Set they started with the
Cockpit version and changed the numbers. Unfortunately, they forget
(and did not check) that the Compact Board Set includes the Opto Harness.
The Opto Harness was therefore not ordered. They had 33 in inventory.
They will either have some made outside on a quick turn at enormous cost
or make them in-house at enormous cost.
74AS823s
--------
I have received an Atari pn for the 74BCT29823NT and have asked Art's group
to ECN the parts list for Compact and Cockpit to use this new part.
Production DSK GALs
-------------------
Rick Owens gave me a tube of GALs that he said didn't work in the DSK board.
Production was having a large fallout rate.
I tried them in a board and, sure enough, they didn't work. I erased and
reprogrammed one and it did work. I again erased and reprogrammed one but
with the security fuse set, and it still worked.
Last June when I released the GALs I gave the vendor a floppy disk with the
JEDEC files on them. The files included test vectors so they could test the
entire part and not just verify the fuses. They obviously have not been
using it.
The GALs need to be programmed in order to work properly.
On Wednesday, Joan Alpen from Merit brought over some guy from Lattice who
told the following story:
On one of the machines the Master part got fried so the operator
took one of the already programmed parts and used it as the new
Master, not knowing that the security bit was set. Therefore, all
parts made from this new "Master" were blank; except, of course,
for the user signature. When Manufacturing complained that the
GALs were blank, Merit verified them against the blank "Master" and
proclaimed them properly programmed.
When I told the Lattice guy that I had given them a diskette with
test vectors and that if they had used them this problem would not
have happened, he said they were using them. Joan spoke up and said,
" No, the machines were not linked to the PC."
The Lattice guy ended up admitting they were not using the test
vectors and said, "You got a problem with that?" I told him I did.
The second time he asked me, "You got a problem with that?" I told
them we were done and, being in the cafeteria at the time, walked out.
2 of 3
OrCad
-----
As an exercise I have used it to do the DSPLINK schematic; OrCad is very
useable. Also, it will produce netlists in lots of forms, including SCICARDS
and WIRELIST.
Elcor has sent me information about the new OrCad release due in December.
The major change seems to be the Graphical User Interface (GUI) so people
don't have to type DOS commands.
Race Drivin' Compact Manual
---------------------------
I have reviewed and corrected the Race Drivin' Compact manual. There were
numerous places that contained information that pertained only to the Cockpit
game. I have gone over these problem areas with Andrea.
P15 Programmer
--------------
Two of the modules have developed problems. I believe it is due to the
extensive rework done to accomodate the Yamaichi sockets.
So far we have supplied the following programmed parts: 635 in June, 325
in October, and currently 414 out of a projected run of 1144.
If the company expects to use P15s after the current run I recommend that
a new board be done to use Yamaichi sockets.
DSPLINK
-------
As a result of the Serial Meeting I will modify the test board to permit
switching the 2105 interrupt to be either when the 68010 writes to the
interface or when the 68010 reads the interrupt.
I have not made much progress because I keep getting interrupted here.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 30 November 1990
Race Driven' Self-Test
----------------------
I have signed off several parts lists that were ECNed for the new Self-Test
ROMs.
P15s
----
The two PC Boards used for programming the parts have essentially crapped out.
The WireWrap board we were using has also crapped out.
I believe the problems are due to:
1. The AMP sockets wear out after a few hundred insertions.
(AMP no longer makes burn-in sockets so we cannot replace them.)
2. The board with the Yamaichi sockets suffered mortally from the rework.
We are working on two solutions:
1. I have given Joe the inputs to make a board with one module. I have
also asked him to use very conservative design rules such as
12 mil traces and only one trace between IC pins. As of Friday he
has been making good progress on the board.
2. Jim Freitas has just gotten a Data I/O 2900 programmer so they
can do GALs. With an appropriate adapter ($995) and the
Microcomputer Data Library ($650) it will program the part.
Whether it will set the security bit is an open question.
Jim is borrowing the adapter and library so we can find out.
The best solution would be for us to buy one so we can continue to
program them. We would also be able to do GALs which would be nice.
3. Emmette has been making parts on the PC Programmer with the TI Adapters.
There has been a high failure rate with parts not being blank.
As of Friday the 2900 has refused to work with the P15s. It reports
Device Overcurrent. The socketing system looks pretty dubious.
So far we have supplied the following programmed parts: 635 in June, 325
in October, and currently 886 for a grand total of 1846.
DSPLINK
-------
I don't get to work on it very much.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 2
Fr: Jed Margolin
Re: Status Report
Dt: 7 December 1990
P15s
----
Joe has finished the new Programmer board and has sent out for film.
We are ordering Yamaichi sockets for the boards.
When the boards are built I will need to modify some software. One of the
modifications will be so it will not set the security bit if the part
is non-blank.
Manufacturing's Data I/O 2900 refused to program a P15. The salesman
brought in another 2900 and it also refused to program the part. There
appears to be a problem with the Model 2900. Perhaps it is a software
problem that they can fix.
DSPLINK
-------
It has actually started to behave responsibly now that I am giving it a good
long time to recover from /RESET.
It is alive but needs more work.
Programmed Parts
----------------
We have received from Rob Bryant 644+500+28+263 = 1435 parts. No one saw him
deliver the batch of 263 and there is no work order tag for them. In addition,
the batch of 28 is for our own DSK boards.
As of 12-7 we have made 1322 programmed parts.
We have experienced 95 bad parts with many of them non-blank to start with.
Although we seem to have bought these parts with no warranty, we have the
right to expect that these parts have undergone at least minimal testing.
Not Blank = Not Tested.
There is a discrepancy of 1435-1322-95=18 parts that are not accounted for.
Either Rob remembered incorrectly the 263 parts or they got lost or someone
took them. Anyone with a Schlage card can get into our lab and that really
sucks.
So far we have supplied the following programmed parts: 635 in June, 325
in October, and currently 1322 for a grand total of 2282.
Emmette deserves congratulations for his fine work in programming parts
under trying conditions.
2 of 2
Manufacturing
-------------
We are using a 2% capacitor on the Main Board in the 12 Bit A/D circuit.
When the board was laid out the AVL called for parts on 0.3 centers.
Someone seems to have changed the AVL and added parts on 0.4 centers.
They bought said parts which now cannot be auto-inserted. I was asked if
I would allow 5% parts which are 0.3 and therefore can be auto-inserted.
I said no, just buy the 2% parts that fit.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 1
Fr: Jed Margolin
Re: Status Report
Dt: 14 December 1990
Programmed Parts
----------------
So far we have supplied the following programmed parts: 635 in June, 325
in October, and currently 1322+137=1459 for a grand total of 2419.
TI would like some of the non-blank parts, as well as the part that could
still be read after setting and verifying the RBIT. I don't know about
this last one; it contains the released program.
DSPLINK
-------
I have turned over to Ed Rotberg two working DSPLINK boards and some
documentation, as well as the revised Self-Test. The boards perform as
advertised: they exchange 32 word buffers at 5 MBits/sec and the 68010 can
upload and download buffers at full speed inside a DBF loop.
I have finished the schematics for the Party Line version; Emmette is
wirewrapping two for Peter.
There still remains a considerable amount of work to be done:
1. Finishing Self-Test with Memory Tests and full Link Tests;
2. Documenting things for Peter.
Peter's code will have to start at Boot Page 3. The board will come up in
Self-Test; a Self-Test Command Code will cause it to Boot up Page 3.
After that Peter can do whatever he likes. He cannot, however, change
the Self-Test code in Boot pages 0-2 unless he wants to write a lot
of 68010 Self-Test code.
I have attached a copy of the documentation to this report.
It was extremely difficult to get as far as I did because:
1. There was no working emulator. The Rev 2 emulator made the
situation look worse than it was.
2. The power to the various benches has been, and apparently still
is, erratic. It may explain some of the problems with the
P15 Programmer.
Jed
_____________________________________________________________________________
To: Rick Moncrief 1 of 3
Fr: Jed Margolin
Re: Status Report
Dt: 25 December 1990
Data I/O
--------
The Data I/O saleman brought a new 2900 in; it worked with some E15s but
failed on others.
The conductive foam in the socket adapter contains lots of vertical wires
separated by non-conducting foam. There appears to be too much play in the
adapter for it to make reliable contact.
Also, Data I/O has made a major blunder. When the security fuse is set, the
programmer refuses to try to read the device. It does not verify that the
device is not readable. The salesman didn't understand why this is a problem.
I explained to him that I had a device with the security fuse set and
verified but which could still be read. The security is in the device, not
the programmer. He still didn't understand, and suggested I call Data I/O in
Washington.
I told him that due to my experience in trying to get getting information
from Data I/O after getting through their exerable voice mail system I am
not inclined to go through that again, especially to give them information
that will help them make a better product. They should call me.
He said he would have someone call me; no one did.
I am tempted to write a letter to the TI newsletter explaining why the
Data I/O 2900 is not recommended for programming P15s.
As a result I cannot recommend the Data I/O 2900 for programming P15s or
anything else that uses that stupid PLCC adapter socket.
By the way, when they test the program for a new device they test it on four
devices.
DSPLINK
-------
With Matt's help I have found and fixed the bug in the DSPLINK software.
From talking to Dan Ash I have learned that there is another bug in the
2101. The LSB transmitted by Serial Port 0 does not work. The recommended
workaround is to send one more bit than is needed, and then drop the LSB
on the receive end. This is not a good fix if you want to send 16-bit words
because 16 is the maximum number of bits/word that can be sent.
I have been using Serial Port 1 (it is the only Serial Port in the 2105) and
it is ok.
Unfortunately, I wanted to use Serial Port 0 in the DS III board for
communications so I could use Serial Port 1 in its alternate mode which
provides two more interrupts. They would have been used to make the parallel
interface more efficient.
No word on when the 2101 will be fixed.
2 of 3
DSPLINK - Party Line
--------------------
I have received two WireWrap Boards from Emmette but have not had time to
fire them up. These two will presumably go to Peter.
The schematics for this board are available to the other groups who have
decided to use it as a standalone board. Unless they put in a Driver Expansion
Bus they will have to adapt the interface to whatever they decide to use.
There still remains a considerable amount of work to be done in finishing
Self-Test with Memory Tests and full Link Tests.
Bonnie's PC Board
-----------------
Whwn fully loaded, Bonnie's PCB will have:
1.2M Bytes of EPROM (10 x 27C010-170)
64K Bytes of RAM
4K Bytes of ZeroPower RAM with Software Write Protect.
Driver P15 System
Because the Expansion Bus only supports 2M Bytes of Address space, it will
be challange to have this thing work with other boards. I am willing to
guarantee that it will work with a DS III board, but that's it.
I have schematics ready to be reviewed.
More to Do:
Program the GAL address decoders;
Write Self-Test for this board.
Note: The GALs on this board can be replaced by standard PALs.
DS III
------
The schematics are ready for a preliminary review.
More To Do:
Program the GALs;
Write Self-Test.
Note: In one location on this board the GAL cannot be replaced by a
standard PAL.
3 of 3
New Release: the Driver LINK
----------------------------
I have updated Self-Test; the versions have to be tested in a game.
There are four versions:
1. Cockpit North American/British;
2. Cockpit German;
3. Compact North American/British;
4. Compact German.
The eighteen Parts Lists that will have to be ECNed are as follows:
Cockpit: North American A045988-05
British A045988-07
German A045988-06
Cockpit Kit: North American A048486-01
British A048486-02
German A048486-03
Cockpit Link Kit: North American A048978-01
British A048978-02
German A048978-03
Compact: North American A046901-04
British A046901-05
German A046901-06
Compact Kit: North American A048984-01
British A048984-02
German A048984-03
Compact Link Kit: Referred to in A048980-01,-02,-03
I can't find the actual parts list
Not to mention AROMREL.DOC
Driver Documentation
--------------------
I am working on it. The new QMS Laser Printers are very nice, and are supposed
to be HP LaserJet compatible, but will not do 12 CPI. It comes out as
16.67 CPI. (10 is 10, 16.67 is 16.67, and 12 is 16.67) I have told Dave and
he has informed QMS of the problem.
Jed
_____________________________________________________________________________