Tony Duell wrote:
> The keyboard on the HP9810 and HP9820. OK, not really a computer (they're
[....]
OMG. That is TOO COOL.
I got part-way to working it out by dismantling a machine and testing
signals. But not being able to determine the phasing of each transformer,
I thoght that the metal disk increased the coupling on the 'top' side of
the PCB. Fortunately I read the patent which explained it all.
They don't have much of a following, but I think the HP desktop
calculators up to about 1977 (that's things like the 9100, 9810, 9820,
9830, 9815, 9825) are very interesting and elegant machines. Maybe
they're not computers in the strictest sense, but they're darn close.
Totally useless trivial point about the 9810 (and none of the others). It
can be claimed to be a 1, 3, 4, 6, 8 or 16 bit machine :
1 bit is the width of the binary ALU path (bit-serial processor)
3 bits is the physical width of user program memory (it's doubled-up in
hardware to store 6 bit keycodes)
4 bits ithe width of the BCD ALU path (from the A and T registers and
back to the A register)
6 bits is the logical width of user program memory
8 bits is the physical width of user data memory (again doubled-up in
hardware)
16 bits is the logical width of user data memory, and also the size of
the CPU (shift) registers.
It's almost criminal that the suits have taken over HP and ruined it.
They came up with SO much amazing stuff over the years.
Agreed. Time was when I read about a new HP handheld calculator and I
wondered how I'd afford it. Now that worry has been taken away, I don't
_want_ it. I made the mistake of buying an HP49G as soon as I saw it.
The first ROM versions were so buggy as to be unusable (and to be
honest, while later flash upgrades have fixed some bugs, they've also
added some), it was mis-advertised, the keyboard is horrible, and so on.
I wrote a letter to HP complaining about this, they didn't even bother to
reply (and yes, I did include return postage). That is not the attitude I
expected from a once-great company.
No, I'll stick with my old 67, 41, 71, etc machines. Those work, they
behave as advertised. And I can understand them.
And that's just calculators. Their test gear was great too. As an
example, take a look at the manual for the 5245 counter (the 5243 must be
similar, probably others too). Building a decade counter, latch and
display (nixie) driver using 8 transistors and no ICs is a major hack
IMHO. Every time I read an _old_ HP service manual, I am filled with
admiration for the way it was designed and constructed.
Alas not any more. Their modern stuff (printers, digital cameras, etc)
seems to be cheap and nasty. Probably no worse than anybody elses, but no
better either.
-tony