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

Any clue if TMS1400 can be read this way as well? (Used in Coleco Talking Teacher and its 5 million variants, which I'm due for a blog post about)

Oh yeah, that's right, we can get the PLAs from that pic.

With the top metal view, there's so little contrast between the 1s and 0s in the ROM array that my program can't differentiate them, though. That pic is so clear that with a program could look across each row and column and pick out the bits, but it would be a lot easier with the metal layer removed.

I picked up a TI55 that has the same chip, part # TMC1503.

About a year ago, Kevtris told me his method for dumping TMS1000s/TMS1100s didn't work on the TMS1400s that he tried it on. But his method also didn't work on the G die revision TMS1100s that we found, and had issues with the E die revision.

Ha so perhaps he's tried that for the "MP7324 ML/_\51802" inside the Talking Teacher then (he has a few IIRC).

TMS1400 is derived from TMS1100. If we have a decap of it, it should be fairly easy to emulate. I'd suggest starting with a straightforward 1-chip device, like a calculator.

S&S-related: I'll start on 'extracting' the ROM from the language tutor patent. All those 1s and 0s may look dazzling, but it should be a fun side project making a custom helper tool to read it out.

I couldn't find any calculators that use the TMS1400, but I should get a Split Second in a few days. I'll decap it and upload the pics. Maybe entering test mode requires some other pins to be held high or low, and we might get a hint from the die shot.

I decapped the TMS0980 from the Ideal game Electronic Detective. I think it was the first product to use the TMS0980- the die says 0980B-00. The ID # on the chip is MP6100A; Stop Thief, a somewhat similar game, is ID MP6101B, and has 0980B-01A on the die. http://www.seanriddle.com/0980.html

Thanks, I'll have a MESS driver for it in a week or so.

Don't forget TI-30 was obviously the 1st TMS0980 product. =p The new hardcoded opcodes in it were specifically added to make calculation loops faster (especially in the case of sin/cos/tan)

Yeah, they must have set the high bit for the internal products: the TI calculators have die IDs 81, 82 and 83, and the games are 00 and 01. But the TI30 came out something like 3 years before the games.

Speaking of calculation loops, I had forgotten how slow the TI30 is- it can take a couple of seconds to do sin/cos/tan, depending on the argument. At least they did a cool segment twirl on the rightmost digit to show it's thinking.

I uploaded a die shot of the TMC0350NL CD2302 from the Speak and Spell Vowel Power chip that I fried: http://www.seanriddle.com/speakandspell.html