deskthority

Hi everybody,

Searching for some help about this problem I found this fantastic place! So let me introduce myself and the problem,

my name is Gabriele and I'm the director of "Museo dell'Informatica Funzionante", a computer museum in Palazzolo Acreide, Sicilia, ITALY (http://museum.freaknet.org)

Yesterday a friend come to us with this "Citrus II", an Apple 2 clone; he made a mistake trying to put his ASCII keyboard into an APPLE 1 replica.

The keyboard itself is handled by a Toshiba TMP8048 cpu; the result of his mistake was about -12V connected to pin 15 (D3) of that CPU. This data line is now burned

So, the Apple 2 clone power up and work flawlessy but the keyboard doesn't. Some keys are working correctly, some not, as they print other characters.

We traced this, seeing that pin 15 is not changing its state at all.

We also found that pressing a key and pulling the pin 15 to ground, the correct key will show up! Also, pressing the working keys with pin 15 to ground result in printing out the wrong characters...

So, that's our idea: add some logic to this keyboard to "sense" the keys and pulling the pin 15 to ground when those keys are pressed.

Those keys are tied to the input lines of the 8048: for example, the "1" key is connected to pin 21 (P20) and pin 31 (P14) of the 8048; but I can't measure any kind of logic signals there, so... how can we "sense" the key contact to reach our external logic and do the trick?

Our hope is to avoid changing the 8048 using some external PIC but maybe this will be the only way to recover this problem

Sorry for my bad English, I hope this will be clear and readable by anyone

Thank you for any kind of help,

Gabriele

Welcome. You did an excellent job of describing a confusing situation!

A quick search for the Toshiba TMP8048 online seems to show they are still available. The fried logic you're experiencing probably needs replaced rather than worked around. If it can be removed without much difficulty, I would replace the chip. Of course, things get more interesting if it is hard to pull out.

Other people here know far more about digital electronics than I do. Typically, we replace old controllers with custom logic of our own for supporting USB or (soon) Bluetooth. But as a museum your priorities are different. Perhaps there is a way around it.

Thank you well, the IC is socketed but, due to the pin 15 burned out, we can't read the internal programming properly, so there's no way to program a new 8748 (the EPROM version of it)...

I have an update: the pin 21 connected to the key is not "floating", is a square wave, I saw it right now with an oscilloscope.

A picture of it is here:

http://zaverio.com/~asbesto/flo.jpg

Maybe I can sense the key using some TTL ?

All I need to do is open my mouth to prove my lack of digital skills! This is quite an interesting problem, then.

The keys themselves will be laid out in a two dimensional matrix. If you want to sense which keys are pressed, you need to work it out. Here's a thread going through the process on a modern keyboard.

Beware ghosting! Keyboards often use diodes to get around the problem posed by every matrix: what happens when several keys are pressed at once? Your logic will need to take this into account too.

Actually, your oscilloscope signal reminds me of those from this current thread about a similarly old IBM keyboard. In that one, the signal is a probe. The matrix is scanned by one axis as the other is read. Naturally, the probe looks like a square wave.

But beware my naïveté, just because I saw two oscilloscope traces posted lately!

I suggest you ask on some Apple II forums (for example here or here), including pictures of the keyboard and the whole machine. Most clones were quite direct and used commonly available parts, so I think it is likely that someone will know enough to help you find or program a replacement controller.

Most people on this forum would be more interested in replacing the controller to use the keyboard on a modern computer, e.g. here.