The current page has been visited 24676 times.
Visitors today: 2
Hits to all pages: 8092054
The first working CPU-board!
Two weeks ago my order of three ColdfFire XCF5206e/54MHz arrived.
The CPUs are really cheap: about 15 € (about $14) was the price I ordered them!
Now, the next work to do is to develop a first prototype-board to make a start.
Last week my brother Christian started to draw the layout for the 5206e by hand.
The reason for this is, that my layout software "Ranger2" is not able to route
the small pin pitch (0.65mm) of the 5206e. This job was not simple because the
layout needs to have the different raster sizes (0.65mm, 1/10inch) together in
one drawing. To solve this, I wrote a little program which generates different
raster sizes as line-graphics. The advantage is, that the average error doesn't
multiply with the amount of pins. By the way: the drawing program was DPaintIV!
|The hand drawn layout then was printed to a inkjet film with a high resolution
inkjet printer (HP DeskJet 890C) and then was used as mask for the photographic
process of photopositive board material. After developing the photopositive
surface, the board looked like in the picture on the left.
|The developed board now was put into acid which removes the unused copper of
the board. The remaining copper are the signal and power lines used for this
CPU board. The photo on the left illustrates this process. You can see that the left
part of the board has already a good line structure. The bright parts still are copper
and the dark/transparent parts are the blank epoxy-board.
|This photo is a picture of the finished, undrilled board. Its line-color is
silver-white because I have given it for some hours into a bath of gloss tin ("Glanzzinn")
which protects the raw copper from oxidation.
|Yes! And this picture is a photo of the soldered and
finished board! You also see a lot of "flying resistors" in the air, this is because
I tested the CPU without any peripherials! This is possible because the
ColdFire has 8kB on-chip RAM and allows control over a debug cable. This BDM-cable is connected (the flatband-cable on the left) and enables me to configure the single
CPU and let run programs on it!
Well, this CPU board alone has not much sense. The idea of this board is to have the
fine pitched 5206e on a easy to connect board. This is the "5206e Adaptor V1.0".
The next job for me is to create a motherboard for the CPU-board which has PS/2-RAM,
GoldCap-buffered SRAM (better than FlashROM!) and an I/O-Interface in form of a
simple but fast network connection to the Amiga via parallel port. I will probably add
also harddrive support in form of a stupid IDE-Interface (later I will of course add also
my SCSI-controller to it). When I have finished this development
stage, I have the basis for further software-developments. For the next time, this
computer will probably only have ANSI-graphics, but this is enough (just think on Linux...)
for creating the whole OS and the rest of the system... :)