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

Spelling B did come with a book. It looks a little different from that pdf, but the pics look like they are the same: 26 is a bed, 64 is a dog, and 85 is a kite. Must be another bad bit, huh? I'll check again.

spellingb1984raw.bin byte 2B1: 8D should be 8F

There's as much chance of bad emulation too. Unfortunately it's still broken after correcting that bit.

Bummer. I wrote a new algorithm for detecting bits in the die shots, and then compared its results to the old ones and found that 1 bit in the 1984 dump, and none in the 0272 dump.

Does it find any in mrchallenger 0273? That game almost works. I can see it going to the correct address in the word ROM for MYSTERY, then read the letters. And then it shows junk on the screen instead :p

I had scanned that one with the new algorithm already, so I just now rescanned it looking for bits where the average value of the red component of the color in the middle 4x4 of each bit was > 180. After fixing up some bits that it missed, that dump is identical to the first one. So that gives me pretty good confidence that the dump is good.

I think I found the problem for Mr. Challenger, the game expects reads to be valid after setting indirect address on the VSM.

Cool. I read about that in the data sheet, but didn't know anything used it. I've got some VSMs; I can write some PIC test code and capture waveforms if it turns out that you need them.

After fixing a stupid bug, Spelling B emulation is playable, I didn't get any lockups. There's still a bug where the game gives the wrong word sometimes, maybe mcu<->mcu comms emulation timing problem.

Also FWIW Mr.Challenger patent with source code: 4421487

Hmm, since the Mr Challenger patent has binary object code, it might not be too hard to OCR to compare to my file.

I got the TI calcs- a TI-1200 with an 0952 and a TI-1250 with an 0972. I got a good pic of the 0952 top metal, but I damaged the 0972 die. The ROM array is OK, though.

The 1250 has memory functions and CS (change sign, or +/-) that the 1200 doesn't have. They are in the same case, with just those 5 buttons covered up on the 1200. I pressed the button where CS is, and it worked as it should. The memory buttons were actually missing the metal domes that make contact with the wires, but I used a clip to connect the wires, and all the memory functions work as well!

I haven't removed the top metal layers yet, so I haven't transcribed the ROM bits, but I visually compared the first few bytes and they are different.