Atari-Coldfire initiative from Switzerland and Austria
According to new impulses from Switzerland and Austria, since the
beginning of december 2008 several developers are contacted personally.
The aim is to find out which skills and how many support-possibilities in
the Atari-Community are in existance. Afterwards can be decided if and
how the project can continue, or if it can get started again together.
Dead or Alive?
In the recent weeks, I (Norman) got a couple of emails requesting
the state of the Atari Coldfire Project. The last official news
entry of this website was published an eternity ago, so I decided
to publish the current state of the project.
The project started with high enthusiasm and a lot of
people were willing to lend a helping hand to let the
project succeed. Since end of 2003 however, there had been only two
persons, Jens and me, contributing development work to the project.
In the last year, things became even more difficult because both,
Jens and me, had been occupied by our jobs more than ever.
Jens had to work like a maniac (in its literally meaning ,-) on
his diploma thesis that turned out to be an extraordinary complex task.
After finishing his studdy, he was pretty much occupied with
searching a job.
I was working a half year at Intel in Hillsboro, Oregon - far
away from Dresden and far away from any Atari-related activities.
After my return to Germany, my job at the operating systems
research group at my university kept me extremely busy - leaving
me almost no spare time to spend on my hobbies. Furthermore, in
February, my son was born - a new high-priority project that
requires a large quantum of my spare time :-)
Ok, enough excuses - let us take a look at the current stage
of the Atari Coldfire Project. The good message is: it is not
dead but still slowly progressing. Jens is still working on
the board and we both are willing to build a functional computer.
The bad news however is, that we consider the project as
a pure hobby and thus, are not aiming to produce, sell and
maintain a product-quality result. We realized that it is
absolutely impossible for two persons to develop, produce,
assemble a custom hardware platform, develop a good-quality TOS-compatible
operating systems environment, ship the product and provide future
support - all that in their spare time using very limited private savings.
Even if we succeed in developing a working computer,
we are not going to take the huge financial risk to produce a large batch
of boards, ship them to customers, and provide warranty and recycling
of our hardware as required by german law.
Our personal motivation for the project is intellectual curiosity,
learning, and the great experience of creating things in joint work.
All these practical and economical things as described above
are just beyond us.
We plan to release our work under
an open licence as I did with my Coldfire-related software developments.
This way, interested developers can take our work as a basis for further
On another note, new Atari Coldfire activity seems to arise. Mark
Duckworth is apparently tired of waiting for new hardware and takes a different
approach to create a Coldfire-based Atari clone. I am happy that he is picking up my efforts in porting MiNT to the Coldfire platform as basis for his work :-) Read more on his blog:
The board develoment is still alive
Although the last update was a really, really long time ago,
the board development did not stop. What are the most important
points since the last update?
Jens modified the layer stack such that we achive a
smaller board size and an easier layout routing. The layout has been
adapted and the last circuits (e.g. USB-ISP1161A1, Xilinx PROM and CPLD,
and a TLV320AIC23 - codec from Texas Instruments) are now integrated.
A technical whitepaper that describes the current prototype will be coming
soon. One of the points that caused some headaches was the initial
operation phase. For this purpose, we bought a new oscilloscope
(Voltcraft 6200), so our equipment should be good enough for the
moment. Actually, we gather experience with the new layer stack
with another smaller testboard because the pcb for our actual
prototype is much more expensive than for the other one. If there are
no additional problems, we plan to produce the prototype as
as soon as possible (probably the start will be in 2nd week '05).
In the meanwhile, Charon started to gain experience in hardware design
and started his own FPGA project. You can visit his www site for more
information on this:
MiNT kernel runs on Coldfire!
The first version of Norman's port of the MiNT kernel to
the Coldfire platform is working! MiNT starts up happily and executes
its internal shell. It is also possible to interact with the internal
shell by typing commands into the serial terminal. Thus, it is possible
to list MiNT's logical directory tree, play with symbolic links etc.
Thanks to the approach of running the complete MiNT kernel within an
execution environment, only two lines of code of the MiNT kernel
had to be changed. The execution environment comprises the
cf68klib and ca. 600 lines of custom code that provides the
functions of TOS that are actually used by the MiNT kernel.
More technical details about Norman's work are described
in his updated article
"My life with the MCF5407 evaluation board".
Subversion repository of Norman's source code
The document about Norman's work with the MCF5407 evaluation board
features a new section about porting the MiNT kernel to the
Furthermore, all source code that is referred within this
document is now publicly available under the terms of the
GNU General Public Licence 2. Section 2 describes how to
checkout the source code from our Subversion repository.
First drawing of our new board
The first drawing of our new board is available at
logbook. It is an illustration of the main components
of the system. Some minor stuff such as USB are not
Logbook of our hardware developments
We added a logbook
of our custom hardware developments to this
website. Jens - our main hardware developer shortly outlines
the progress of his work in chronological order. Currently, the
document describes the progress since december 2003. We will update
the logbook on a regular basis.
Norman's life with the Coldfire Evaluation board
Norman Feske wrote an
article (pdf, html) about
his work with the Coldfire MCF5407
Evaluation board by Motorola. It illustrates some steps on the path
to run MiNT on the Coldfire platform. Originally, this text was meant
as personal notes but probably it may be interesting for other people, too.
Last but not least, it expresses that the Atari Coldfire Project is alive
and steadily progressing.
For keeping track of the development of our operating system more
closely, please visit Norman's private weblog:
ACP Coldfire Developmentboard
Although there was no update of the web-pages for some time, the development of the ACP did not stop. During the Error In Line #3 people could have a first look at the Coldfire board developed by the ACP team.
With this board the programming of the FPGA and the routing can be checked and optimised. Among other things the FPGA is responsible for resetting the Coldfire. After about 50000 clock cycles - the Coldfire builds up his internal clock during this time - the bus activities of the Coldfire start.
Meanwhile there is still some testing done with the Coldfire Evaluation board. Norman Feske tries to initialise several PCI graphic cards. Though there is relatively much documentation, this seems to be harder than thought.
Coldfire Evaluation Board
Two months ago Gerald Kupris, who works at Motorola, contacted us to support
our project. Last week we received a Coldfire4 Evaluation board from him
without any charge! What a great surprise!
Currently, there is a meeting of some ACP members (Matthias Alles, Norman
Feske, jk28, Robert Wetzel) at Norman's place.
We already did some testing, tried out the GCC development tools in real life,
compiled and run uCLinux and tried to access a PCI card. The Evaluation
board is a great opportunity to get practical programming experience with
the Coldfire4 processor. A lot of thanks to Gerald Kupris for trusting in
us and supporting us that great!
ACP at Error In Line #3
The major part of the ACP team will be present at the Error In Line #3
Coding Convention in Dresden. So this event will also be an inofficial
ACP meeting. We hope to present the current state of our development
to the other visitors of the EIL3.
French version of ACP news section
The french version of the ACP news site is now available!
A lot of thanks to Bruno Aubin - who is responsible for the translation -
for his great help!
Fredi back on track
Great news: We got a life sign from Fredi Aschwanden. He told us that
he recovered from his personal problems and is willing to support
our project again. Although his spare time limitations will not allow
him to take over a leading role during the development he will
support us as much as possible!
Article "Behind the scenes of the Atari Coldfire Project"
Norman Feske released an article about the previous development of the
Atari Coldfire Project. Beside interesting insights of the work of the ACP team
it discusses the already mentioned board alternative of Wilhelm Elektronik.
Error In Line #3 announced
The third edition of the Error In Line coding convention was just
announced. It is one of the biggest Atari events of the year 2003.
Traditionally, it is held during the easter weekend.
While the origin of the Error In Line conventions lays in the
demo scene it also attracts all kind of Atari enthusiasts now.
Developer meeting at Oliver Kotschi's place
Last weekend the core of the Atari Coldfire Project met at
Oliver Kotschi's place to discuss the next steps of our
project. Sadly, Jörg Wilhelm - the chief of Wilhelm Electronics -
did not manage to attend to our meeting. Thus, he could not
present his hardware solution. Even so, we held long discussions
via telephone about the details of his hardware. Some important
issues are still left. They will be debated during the next
Another big topic at the meeting was the future of the operating
Please stay patient about more detailed information.
We do not want to release concrete information about technical
data, price and release date until a final agreement is found.
ACP team debates alternatives for the Coldfire board
After Fredi Aschwanden stopped the development of his Coldfire board
we searched for alternative solutions. We got several offers from
hardware developers to help us. Rodolphe Czuba as well as the
company "Wilhelm Electronics" offered us the development of complete
boards at low costs.
A lot of thanks to everyone, who contacted us and offered us help!
After an intensive discussion we decided to go for the board by
"Wilhelm Microelectronics" because it can certainly and quickly be
realised. In fact, the board is already ready and only needs to
be slightly modified to meet our needs.
A new developer meeting is planed to coordinate the next steps and
the further development. There the details about the board's
properties, it's price and our time schedule will be discussed.
myAtari reports about the Atari Coldfire Project
The current issue of the famous myAtari online magazine
features two articles about our project:
These articles were previously be released by the german print magazine
ST-Computer. Visit myAtari:
- an article about the developer meeting in Dresden in march 2002
- a large interview with Oliver Kotschi and a presentation of the
Hardware development discontinued
Due to serious personal problems of Fredi Aschwanden, he
is not able to continue the development of the coldfire-mainboard
at the moment.
Thus, we are currently searching for a skilled hardware developer,
who can continue the work of Fredi.
Received first batch of Deesse PCBs
We received the PCBs of the first 10 Deesse cards from our
manufacturer. You can find a photo of it at our
Production of the first Deesse cards
The PCB-producer of the Deesse DSP cards told us that we will get
the the PCBs in the first week of may. After getting the PCBs,
Elmar Hilgard will solder the first batch of Deesse cards by hand.
New PCI-BIOS version
Markus Fichtenbauer released a new version of the PCI-BIOS. The
current version is 1.13d. It is available inside the 'documentation'
section of this website or at the website of Markus Fichtenbauer.
ACP-website in german
The german version of the ACP-website is online now. You can visit it
New Project Website
This website will cover everything about the Atari Coldfire Project.
The sections are:
- 'About': general information about the project
- 'News': Here you can see the current state of development.
- 'Documentation': As soon as the hardware description and other
documents are available you will find it here.
- 'Links': websites that are more or less related to this project
- 'Contact': Get in touch with the involved people!
Developer Meeting in Dresden
At the weekend 8th-10th of march a developer meeting of our project
took place in Dresden. The topics of our discussions were:
- the design of the motherboard: There is still no perfect solution
for the graphics. Since there are no standard chips to connect
PCI/AGP to Coldfire processors, Fredi Aschwanden has to do that stuff
by himself. The final decision depends also on the available
documentation of the used graphics option.
The rest of the principle layout of the board seems to make no
- the operating system: Thomas Raukamp told us that Milan allows us
to use the latest Milan-TOS version as our base of development. We
will try to run MiNT together with TOS as standard operating
system. Frank Naumann will help us with that.
- Deesse-debugging: the prototype of Deesse had problems with its
codec. Elmar Hilgard will try to solve the problem. Beside that
Deesse runs fine already. Mariusz Buras - the main Deesse software
developer - already developed a very powerful programming interface
for it. Since Deesse will be standard in the new computer it will
be an essential part of the operating system.
- Elmar Hilgard solved the last problems with the Deesse-layout - thanks
to the help of Fredi Aschwanden. The first series of prototypes can now