20 Aug
2001
20 Aug
'01
6:51 p.m.
I've got two different Apple][+'s and they have different keyboard/encoder
arrangements. One uses the GRI keyboard I mentioned earlier, and the other uses
something I can't see because it's not out of the box (yet). If the MM5740
(I'll look it up later, in an old NSC MOS LSI databook) is just a ROM lookup
table, an exacto knife and a few wires will allow it to be adapted to a 2732 or
some other common device. If the control, shift, and caps-lock keys (and ALT if
there is one) are simply grounded, they can simply be used as high-order address
inputs (with pullups) and the scan code as the low-order bits, the EPROM can be
programmed to cough up the correct codes.
It doesn't have to be a big deal. The business of scanning a keyboard, given
what the existing circuitry produces and given what the desired outputs are, is
really quite straightforward. My interest is not sufficient to move me to do
the work, but if someone wants to generate a schematic (or, perhaps, refer me to
a page in the Apple DOS manual I could easily come up with a hack to the
hardware to make a common EPROM do the translation. Likewise, if a
keyboard-resident scanner is required, I think I can come up with a recipe using
an MC68HC05 of something of that ilk. Of course, if the fellow whose keyboard
is on the fritz does the research and subsequent work, he'll gain a lot more
than if someone else does it.
Dick
----- Original Message -----
From: "Tony Duell" <ard(a)p850ug1.demon.co.uk>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, August 20, 2001 2:44 PM
Subject: Re: Apple II+ Keyboard Encoder
My Apple II+ Keyboard Encoder card is faulty and I would like to know if
it's fixable. I've tried several different keyboards and 2 motherboards so
There have been several different encoder boards used in Apple ][s.
The one shown in the Reference Manual uses 5 chips :
U1 7400
U2 7404
U3 555
U4 7400
U5 MM5740
If it is that version, then the control key has one side grounded. The
other side is pulled up by R1 (4k7) to +5V. That side of the key also
goes to the inputs of U1d (inputs tied together), which inverts the
signal. The output of that (which will be high when the control key is
pressed, low otherwise) goes to pin 19 on U5, the MM5740 keyboard encoder
chip.
If you have that version, it would be worth checking that the output of
U1d (and pin 19 of U5) do go high when 'control' is pressed and are low
otherwise. If not, the components are probably trivial to replace.
Also, you might check that the sections of U2 (low 6 bits of the output
word) and U1c (again, inputs tied together) are working correctly. Losing
the top few bits of the keyboard data will make it appear that control is
pressed.
Most likrly, though, the MM5740 is dead, and that is non-trivial to
replace. I read somewhere that Apple used a custom version in some of
their machines, because the keyswitches are in a non-standard matrix
pattern (the MM5740 contains a ROM which converts the X,Y coordinates of
a key in the matrix into its ASCII code, and it was possible to get
versions with custom ROM patterns if you wanted 10000 of them).
Many Apple ][s at least in the UK use a different encoder board. I would
guess the circuitry is similar in concept, and all you need to do is
trace the connections of the control keyswitch (most likely one side is
grounded) to the logic, and make sure the signal changes state, at least
up to the LSI encoder chip. Again, if that chip has failed, you've got
problems.
-tony