At 08:51 PM 1/18/03 -0800, Sellam Ismail wrote:
>On Sat, 18 Jan 2003, Tothwolf wrote:
>
>... (re: Apple 1 schematics)
>I asked Woz, and he said as far as he knows, the original masks for the
>Apple-1 PC board no longer exist.
>
>A schematic would have to be made from an original board. Unsoldering the
>chips and sockets would be very intrusive and I doubt you'll find anyone
>willing to do that. Ideally there would be some other non-intrusive
>method. X-ray perhaps?
All you need are your eyes, an ohmmeter, and a lot of patience. I have
reverse engineered some double sided PC boards from my Wang 2200. I have a
lot more to go before I recreate all the schematics. This is the procedure
I have used so far.
I create a drawing of the board and all the chips, including pin
numbers. Then I pick up the board and start with a pin or connector. I
ferret out all the devices/pins on that net by just looking at it (on the
back). Any time you get to a via or other through hole, you must check the
top to see if it goes anywhere on top. Even though the chip obscures part
of the tracework on top, you can still make out some of it from the back
(especially with the right lighting), and if that fails, just buzz out all
pins of the chip and any traces that pop out from under the chip (this is
pretty easy on the Wang since there is no soldermask -- you can buzz out a
trace anywhere).
Once you know everything on a net, double check it by buzzing it out again,
then check off all of those pins/connectors/devices on the diagram. Then
pick another pin and see what it is connected to.
Then the next bit of fun is taking your netlist and turning it into a
coherent schematic.
Here is a pretty regular board -- it is the microcode store for a Wang
2200-T CPU. It took about ten hours to buzz everything out and build a
schematic.
The board diagram (a copy of which I used to make sure I hit all the pins):
http://www.thebattles.net/wang/2200tech/7025-outline.pdf
The recreated schematic:
http://www.thebattles.net/wang/2200tech/7025-sch-p1.pdf
I have the printer interface card & schematic online too, and I have the
netlist for one of the three cards that implement the CPU. I haven't had
any time in the past few months to push this project along, but I hope to
get all of the CPU done some day.
If there had been any buried layers, it would have been much harder -- hard
enough that I wouldn't consider investing the time trying to figure it out.
-----
Jim Battle == frustum_at_pacbell.net
Received on Sun Jan 19 2003 - 01:47:42 GMT