HP-67 chipset


I am trying to repair an HP-67 that almost works. In PROG mode, everything works OK, I can enter a program and edit it. In RUN mode, things look strange. Programs appear to run OK, but the numbers on the display are all wrong: There are leading zeroes, and when you press numeric keys, they are displayed with lots of extra digits. Its not a DSP or format problem - this behavior is consistent from power on and is totally different from that of my HP-97.

The problem looks like it is localized in the RUN mode display, constants like Pi are recalled to the display, and (as far as I can tell from the weird displays) registers and numerical calculations work.

I suspect the problem is in one of the ROMs, but I can't tell which. I only have access to the HP-97 service manual but this is not much help, because the 97 has more chips.

I am including some notes I made while looking at the 67, if anybody can provide additional information, I'd really appreciate it.


8-pin chips (ROMS)

1818-0550 (also 0228) common with HP-97
1818-0551 (also 0226) common with HP-97 (this one contains constants like Pi)

Card Reader Chip
1820-1751 (largest chip) common with HP-97

1820-2530 Mostek part



Actually, the 67 and 97 are pretty similar. In the 67, the 8 pin chips are
combined ROM and RAM (16 7-byte registers). 2 of the RAMs (224 bytes) form the program memory, the other 2 form the user data registers (26 registers, stack/lastX (5 more registers) and one last register for angle mode, display format, flags. Or at least I think that's how it works.
The Card Reader Chip (28 pins) is the same as the one in the 97. The CPU is the 22 pin chip, and is an ACT. The pinout is very similar to the one in the
97 except that 5 of the pins are used for the keyboard column lines. It's trivial to trace these from the keyboard, though. The other pins (power, bus, etc) are in the same places.
The 18 pin chip is ROM0 and the display anode decoder. Again the pinout is the same as the one in the 97 (although the code in the ROM is different).
My first suspect would be ROM0, actually. I've had it fail and cause odd faults. If it's not that, then it's probably the ACT (processor), but I'd have thought a fault there would have caused more problems.

Forum Jump: