HP 29 & 21

[brouhaha]

(11-13-2022 01:41 PM)rprosperi Wrote:  years ago, when Bernhard confirmed my worst fears that my 29C ACT was bad, we figured (just exactly as Teenix states) that there is no harm to try it, and it will confirm once and for all if indeed all the (same process) ACT chips were created the same, as a 21 ACT (being the first) would be about as different as possible as a 29C ACT (being the very last model 4-5 years later) if they were not the same, and to our surprise, it works perfectly.

The NMOS ACT from the 27 isn't electrically compatible with any of the others. I believe there are some subtle functional differences between some versions of the PMOS ACT, because HP had to revise the microcode in the 67/97 ROM0/anode driver for a later ACT version. The newer ROM0 will work with either the older or newer ACT. That's documented in the later edition of the 97 service manual.

The 19C/29C ROM microcode is very closely related to the 67/97, so it's possible that the same subtle ACT difference might apply to them. IiRC, the 19C and 29C each have two ROM0 versions, like the 67 & 97, possibly for the same reason. I haven't yet tried to analyze the nature of the difference. Maybe someday.

That's independent of the 67/97 ROM1 and ROM2 change, which have to be swapped together as a pair. That change was to fix the trig and lastx bugs, but would work with either tho older or newer ACT and ROM0/Anode.

If I had a Woodstock with bad ACT or RAM chips, I'd buy the Panamatik ACT replacement. And in fact a friend gave me a dead 25C recently, so I should do that.