Home Page Forums MAME MAME TMS-09xx/1xxx thread (was New Dumps)

Christmas come early

Indeed!

Welcome back Sean, seems you have a ton of work ahead of you

Welcome back Sean

Tomy Pinball and Mattel Dungeons and Dragons board game are both TMS1100, so they should be pretty easy. VTech Mini Wizard looks like a 4-bit MCU, but I don't recognize it. There's no name or logo on the die other than a V, so maybe it's their own custom. Anyway, I can't distinguish any ROM bits in the array, so it's moot.

Is anyone working on Rockwell PPS-4/1 in MESS? I remember Kevin having some chip, and we found the patent with MM77 assembly code, but I'm not sure if any coding has been done. Patent US4339134 for a blackjack calculator has assembly code for the Rockwell A48xx calculator chip, (I have a blackjack calculator and a regular calculator with that chip) and it looks very PPS-4/1-ish. The mnemonics are different, but the opcodes are mostly the same, and the patent has a pretty good explanation of each instruction and I was able to map them to PPS-4/1 fairly well.

Patent 4322074 for the game Gravity mentions that an object code listing was submitted with the patent application, but nothing is on the USPTO site or Google. So I called the USPTO and asked, and they said I could order the "patent wrapper", which includes the application and correspondence between the patent examiner and applicant. But the older ones are only available on paper, and cost $225 to copy! I bit the bullet, and several weeks later I got a bound folder with several hundred pages of application and correspondence. It's pretty interesting to read; the examiner rejected or objected to all the claims in the application, so they had to make a lot of changes, some of which were rejected again. I was disappointed that the object code listing was just that, a 4-page binary dump of the ROM.

Gravity uses the same chip as Mattel Football and Baseball (B61xx), and I had decapped both of those, but didn't have Gravity. After watching ebay for several months, one finally came up for a somewhat reasonable price. I decapped the chip on Friday, took pics and composited them, and transcribed the ROM bits. A count of 1s versus 0s showed I had swapped them, so I fixed that and then wrote some programs to figure out how to map the bits on the die to the bits in the patent. The layout on the die is 64 rows by 112 columns- the 112 columns are made of 8 groups of 14 columns for the 14 pages of 8 bit bytes. First I figured out how the 8 groups of columns mapped to bits in each byte, then how the 14 columns mapped to banks, then how the 64 rows mapped to the sequence in each bank. Eventually I got an exact match to the patent, so I should be able to use the same mapping on Football and Baseball.

The older games (Auto Race, Battlestar Galactica) use the B60xx chip which has less RAM and ROM, but it looks like they use the same opcodes. I've decapped those, so they should also be dumpable. And there are newer Rockwell chips in the same family (that have more RAM and ROM, and some with better sound circuitry) that probably also use the same opcodes.

$225 for the rom dump? Wow, talk about dedication!
Also, you should (maybe, if you feel like it) bug the USPTO to attach the rom binary pages to the patent hosted on patft.uspto.gov since they're clearly meant to be part of the public record as the patent mentions them...
(there's probably some stupid form that has to be attached to it too, explaining that the pages were initially omitted, though, and its probably too much hassle to generate/make said form, since it probably has to be done via the same lawyer/firm as generated the patent in the first place. and its useless since the patent is expired anyway.)

LN

I don't know how long it would have taken me to figure out the ROM bit mapping, so $225 might be very cheap! But I was hoping for more 'inside' information in the wrapper. The patent wrapper is public information, but I guess the USPTO isn't going to go back and digitize all those patents that were submitted on paper.

I confirmed that the "ROM word address decoder" (to use the TMS1000 terminology) is the same on Football, Baseball, Gravity and Battlestar Galactica. Bit decoding is the same on Football, Baseball and Gravity, but BSG only has 8 banks instead of 14, so I'll have to figure it out from the decoding circuitry.

Tandy Basketball is PIC1655A and Coleco Electronic Quarterback is TMS1100, so those will be easy. Scrabble Lexor is Panasonic MN1405, like the MN1400 in Computer Perfection, so I'll try to dump it.

Tandy Golf is HD38135S, which I don't know, and Mattel Gin is NEC D650C, which I don't know (it also has a version of the LCD controller like in Microvision). Dungeons and Dragons LCD is Toshiba T6818S, which I don't know.

NEC upd650c aka 'D650C' is dumpable electronically with no decap; send to kevtris.

HD38135S I don't know, it may be similar to some of the other Hitachi NMOS parts, which are also dumpable electronically. Do we have a datasheet for it?

LN

I looked through my notes for info about the HD38135, and a couple of Mattel games have similar chips: Sub Chase has HD38132 and Armor Battle has HD38133. I sent Armor Battle to Kevin, but we couldn't find a data sheet and he didn't have high hopes.

sean: I think the MN14xx can be dumped electronically, I saw something about it in the data sheet I think, it's been awhile. I remember looking into it when you did the computer perfection.

I can dump the D650C with the D55x rig I already have. I dumped one I found cheap on ebay to test.

I don't know much about the toshiba part- there's an LCD driving toshiba MCU in that vtech learning window as well (and SP0256 speech chips and ROMs that I have dumped).

Kevin- I've found a few datasheets, but they don't really have any info about the test pin. I'll try some simple stuff and see if I get lucky.

You want me to send you the Mattel Gin game that has a D650?