TI-81 Hardware and ROM Versions

[GalacticPirate]

Little question : how did you manage to dump the ROM of a calculator that hasn't got any link port ?

[critor]

@STV
Il faut utiliser une faille pour exécuter du code asm qui affiche la ROM à l'écran, puis filmer cet écran, puis faire de l'OCR sur la vidéo produite.

[Ti64CLi++]

Désolé critor, c'est une 1.1

[CVSoft]

TI-Planet has documentation on the TI-81 ROM dumping procedure here.

[critor]

Désolé critor, c'est une 1.1

Pas grave.
Garde la précieusement - la 2ème version de la toute première calculatrice graphique de TI, c'est une pièce de musée !

[Ti64CLi++]

En même temps, ma mère la eu pour le lycée, c'est-à-dire fin année 80, début année 90
Pourquoi n'est elle pas née 1 ou 2 ans plus tôt

[CVSoft]

After looking at Datamath a bit more carefully, I noticed that what I have been calling 'variant 1' contains two distinct hardware variations:

• Variant 0 uses the Toshiba T6A49 ASIC, and has a 74HC14 IC near the ASIC. These are expected to use ROMs 1.0 or 1.1. The mainboard is 'TI81MB-31G'.
• Variant 1 uses the Toshiba T6A49A ASIC, and the 74HC14 area is unpopulated along with the pads for supporting hardware. These are expected to use ROM 1.1K. The mainboard on the Datamath I-0990 model uses 'TI81KB-32F' -- is this related to 1.1K?
• Variant 2 uses the Toshiba T6A43 ASIC, and the PCB has been redesigned to '9081NMB-30C'.

Sample size is, at the time of this posting, limited to the few calculators listed in Datamath's collection and my own. The transition from variant 0 to variant 1 happened between July and September 1990, and from variant 1 to variant 2 between November 1990 to January 1991.

[CVSoft]

A few days ago, the TI-81 0600008 I-0690 arrived. To no surprise, it has ROM 1.0, and the hardware is very consistent with Joerg's ~0690 TI-81.

Internally, it is virtually identical to the 0790 TI-81 in my collection; a Toshiba T6A49 ASIC and Sharp LH5164N RAM with an unknown ROM, labeled 267422 © 1990 TI and presumably a 32kB or 64kB PROM. The display board shows some work or messy soldering around the three Zener diodes. CPU speed is somewhere around 5.4 MHz; I will measure this value precisely soon. Interestingly, Joerg's two early TI-81s use Hitachi RAM whereas my two early TI-81s use Sharp RAM.

Unfortunately, I found that the ROM dumping utility did not work on ROM 1.0; it would freeze on a busy screen when the final step is performed.

[critor]

ROM 1.0 at last !
Congrats !

The other dumping way is unsoldering the ROM and connect it to an EPROM programmer.
I've already done this with a TI-80 using an LH5359 ROM.

Many TI-81 starting from version 1.1 also use that same ROM.

The problem is that this version 1.0 TI-81 ROM is currectly unindentified, as we can only see the "(c)1990 TI" sticker.
(same problem with the Datamath 1.0 TI-81)

Can you check if something's written on the ROM chip under the sticker ?
If not, we could just pray that this ROM chip is pin-compatible with the LH5359 ROM... checking the PCB circuits and comparing them with PCB 1.1 may help us figuring it out.

And it would be wise to get a 2nd TI-81 1.0 before attempting such potentially destructive dumpings.

[critor]

So apparently, ROM 1.1 chip is a Toshiba TC54256AF :


As the TI-81 1.0 is using the same PCB, we can assume the ROM chip is either the same, either pin-compatible.

It's not in my current USB EPROM programmer compatibility list, but there are other programmers which clearly list it as supported like the TL866A or TL866CS :
http://www.autoelectric.cn/minipro/mini ... rtlist.txt