1

(37 odpowiedzi, napisanych Sprzęt - 16/32bit)

Hi All!

As I got citated I´d like to give you some answers directly :)

-) No MIDI recently at the FireBee. We are still debugging the interface itselve. It seems that the FPGA needs improved timings and some more attention. So the main question is answered here at this point, for MIDI get an Falcon right now.

-) We are planning to get some of the killerapps up and running, but there is no timeline for that. But I think we can give this summer Notator a try. No promise of course as we are all working for free in our spare time.

-) But before taht MIDI has to be debugged, and one of the more GEM-near Sequencers has to work. (Recently no Sequencer is running without problems)

-) The idea about an ST inside the FPGA is an idea and nothing else. Nothing has been done in this direction. It can be done, but nobody tried it. Recently there are 3 "compatibility layers" possible.
1. cf68klib from MicroAPL inside FireTOS
2. Didiers Purce-C patcher, catching some Pure-C specific stuff

Both of them are inside TOS and you don´t have to deal with them. If an application does not work, you can try:

3. Vincents soft CPU 68kemu. Thats a CPU Emulation, while the System itselve stays ColdFire. That one is used by renaming ".PRG" to ".68k" so the application is loaded by the SoftCPU.


None of that helps with all the hardware near Sequencers. For example we had to create a Black&White Screenmode for mordern displays as well. Gladly that one is now working with the latest FireTOS from February. So the MIDI stuff is one agaenda of the project, but it will not happen the upcoming months, as it is still much much work.

Whoever would like to support us (especially FPGA developers) is highly welcome to reach our goarl faster and improve the FireBee functionality.

2

(91 odpowiedzi, napisanych Sprzęt - 16/32bit)

Cyprian napisał/a:

I completely have no idea, let's wait for Cod3r's link and see what it is worth.

Yes, I am very interrested too! Let´s see what he shows us. I attached a picture of the Cofilint (the Japanese board that disappeared) It was done in 2007 and was 250 Euros. You can see that there are very few connectors.

Cyprian napisał/a:
MathiasACP napisał/a:

Cyprian: it is much for eastern european. I am especially sad about russia, where people cannot pay 3months income for a FireBee for example. But show me that western european who earns 2500 a month, ... I live in Vienna, the 7th rich country of the world, and I have to live with 600 Euros per month! My "girlfriend" finished university and has to work for 850 Euros. So have many of my friends.

Wow, I wasn't aware about that. I estimated that based on minimal wage multiplied by two (~medium wage): http://en.wikipedia.org/wiki/List_of_so … nimum_wage

That list is wrong. At least for Austria. The median income is somewhere at 1400,- Euros (but you have to exclude taxes and social insurance so about 1000 Euros is left (as median!). So the list is median income not minimal wage! As you know there are few rich people who kill the statistics! For example Taxi drivers in Vienna, have a minimal income of 480 Euros (for 60 hours work a week! A collective treaty between trade union and  employer's federation) per month!

And the 2nd border is the  poverty line. That is recently at 720 Euros cash per month. And here are 2 Millions below that line, or very very near (out of 8 million inhabitants in Austria).

In fact I wanted to come to Silly Venture, but could not effort it, to travel that 800 kilometers. :-(

Cyprian napisał/a:

I'd just like to see similar solution like Raspberry Pi but based on ColdFire processor (with GPU similar to Mali-400). Small and simplify as possible and at low price (maybe not $35 but rather $100/200).

MathiasACP, there is no any connection between price suggested by me and FireBee.

Thanks for clearing this one!

Cyprian napisał/a:

I'm sure it's worth that price. I'm tracking (and keeping finger crossed) this project from the beginning. Definitely you have done great job.
Actually you can call me FireBee advocacy here :) (I explained here that FireBee isn't close project because of continues development, possibility to upgrade FPGA code; that it has three levels of cpu 'compatibility' - Coldfire, Vincent's emulator and 68K FPGA emulation in the future etc.) And to be honest I'm going to buy (not this year) one FireBee for myself...

I read that you explained several things! Thanks for the kind words as well. But as I said, I am not here for selling computers! ;)
I just like to explain what we are doing. For example we really need some VHDL Guru! ;)

3

(91 odpowiedzi, napisanych Sprzęt - 16/32bit)

It seems I killed all discussion, ... ? Please go on in Polish language, I can follow via translation tools.

4

(91 odpowiedzi, napisanych Sprzęt - 16/32bit)

Candle napisał/a:

could you perhaps explain

Yes, I will try. Even if some of your questions show me that you did not read my huge posting above. I will take this opportunity to write more detailed about your points.

Candle napisał/a:

why there are so many so low density ram chips used,

Do yo mean why there is so much RAM? If yes, Medusa has the knowledge, that RAM was the biggest problem at Atari Clones in the past (different quality, different layout etc.). Aditionally the requested DDR that fits to the onboard RAM controller of the ColdFire is not available as reseller RAM bar. The size (512 MB + 128MB) was choosen as a good size for the upcomng years, as the needed maximum for recent Atari applications in matter of the price. As we get hints from both sides (not enough, too much RAM) we seem to have it done correctly. ;)

If you mean the 64MB of the chips, that was also for long term delivery, and for the price.


Candle napisał/a:

and if you care - why there is a li-ion cell on it?

It´s not just our PRAM battery, but also our included UPS! ;) No let´s be serious; why not, it is cheap, you do not have to change it every year, you can drive your FireBee for half an hour without external power, and we got all the charging electronics for future mobile possibilities already onboard. A 2nd battery would just have bee a comlication of the entire thing.


Candle napisał/a:

how many layers this board has?

8


Candle napisał/a:

what is smallest pich on BGA component used?

I am not a developer so I do not 100% know. The MCF5474VR266 has 1mm.

Candle napisał/a:

fairly sure this would be 0.8mm and 6 layer pcb could be possible, but using bunch of 64mbytes ram chips is not helping here at all (...) i'm asking because i'm pcb designer myself and been working in the past with bga chips requiring HDI boards (0.4mm) and up to 8 layers

Well, you can always take the schema and improve the (nearly) 1000 parts layout! ;) I have enough trust into Mr Aschwanden that he knows his business. Reduction of further layers would not have been possible. Our conductor paths are 120µm (so we are talking about HDI), the grid is 12µm to 3. Even the swiss EMS asked Medusa to rework some of the conductor paths as they cannot – even if all paths where within the specs – assure that such a fine layout would work 100% with their production possibilities.

Perhaps there are some minor optimizations possible, but as you said a few days agao "aguy who doesn´t know what he wants" is an extremely hard judge, especially when you never had a look at the layout or schema. Designing and layouting an entire computer is not that trivial. I belive you should know that. BTW he did all hardware design and layout within 5 months – beside earning money for living.

Candle napisał/a:

last of my questions - how come it looks like pci card?

Well in the beginning of ACP reloaded (beginning of 2009) Wolfgang Förster (Suska) and Fredi Aschwnaden (Medusa) met, and discussed the possible ways. They agreed that a small series with a specific form factor (e.g. MiniITX) would have limited the PCB to a usage in a very special enclosure. The main problem was the PCI Bus. If we would have included several PCI slots, the bard would have become huge, and the cards would have had an angel of 90° to the board. If we would have included on slot for a riser,  there would be an angel of 180° without the possibility to use standard towers etc.

So they developed the idea to use backplane like known form industrial SBC design. Now we can use the board in several different standard enclosures (desktops, towers, 19" racks, etc.) with a standard PCI-backplane. As well we can use the Computer stand alone like in the mini-cases now, or in a Notebook etc. Aditionally the numer of PCI-slots is totally scalable and can fit to every need.

This solution is not perfect, but it is a solution to be able to use the computer everywhere! It would never have been possible to design one or two further layouts for different usage possibilities. So the PCI form factor is a great idea and gives us the most possible flexibility for using our new TOS machine!

But I have to agree – if one first sees the PCI-design it is strange. It is the biggest problem of the entire project to communicate this issue. Several people also think there is another computer needed and the FireBee is just another Janus card, what is totally wrong of course. I thought as well; "what are they doing!" when I first saw the plans ;) But once one really thinks about it, it becomes better and better. It is the best design for a small series computer for different requests of usage in different housing environments.

5

(91 odpowiedzi, napisanych Sprzęt - 16/32bit)

Hi sqward

sqward napisał/a:

Hi Mathias, I was always very open that I didn't like how Firebee paned out, but again, this is my personal opinion.

I know. And its absolutely ok for me, I respect your opinion. I am just wondring what changed the last decade as you have been part of the ACP once, ... ?

sqward napisał/a:

This, by no means, changes the fact that you've all put a lot of hard work into that project and it's benefiting Atari community in many different ways.

Oh, sorry that should not have been a "fishing for compliments" from my side. I just wanted to clear several question. And one of them is that we do not make anything for profit or for selling or any income, and that we all work for free.


sqward napisał/a:

the fact I don't like the product,

Thats absolutely ok. Just because someone likes Atari he/she doesn´nt have to like the FireBee! ;)
I just like to avaoid theories that a much cheaper PCB would have been possible. Thats not true, kicking all legacy ports would never have leaded to a board below 500 Euros. Cheap boards only happen if you can produce in high volume!

And it is important for me, that I am not here to sell anything. That makes no sense. Frustrated users are not our aim! ;) I like to explain our thoughts and ideas, and our plan to stabilize Atari Szene at all, or even get new users. But I do not have to tell you marketing tricks, like "FireBee is cool for Falcon Games" or something.
In fact we wanted to make a TT clone first (well in 2009) and later realized that people like the Falcon more, and so switched to Falcon design as base for the FireBee.
Only get a FireBee if you like to drive the general "productive" GEM developments foreward, and like to use it for everyday tasks. If someone is RETRO interrested, a FireBee makes no sense! In this case better wait for the cheap Suska.

6

(91 odpowiedzi, napisanych Sprzęt - 16/32bit)

Thanks for the kind words! I am happy that "interrupting" your discussion in a foreign language is ok for you.

About the preorders, we did call for them. A long time ago. And the positive replys were the main reason why the computer got built. But we decided against production of "just the proeordered" machines, as people need the ability to see the product before buying it, and getting the trust. So we produced much more for the crowd that will like one now after the first Alpha Series is shipped.
A real price reduction would not be possible below 1000 boards. Thats for sure. And nobody likes - or better ia able -  to produce that amount of boards.

And wieczor, I absolutely understand the retro interrested people. It was my main reason to explain that they are not our main target group. Bot "parties" can friendly coexist aside each other. And thats good, as both groups can participate from each other.  And of course, for a Retro Atarian, the FireBee is crap! ;) But who knows perhaps a FireBee will be aside a ST or Falcon in future replacing the PCs or Macs, ...

7

(91 odpowiedzi, napisanych Sprzęt - 16/32bit)

Dear Atari Users: As I am the one who is coordinating the new Atari Coldfire Project, I have to answer several questions you asked, as some thoughts you expressed here. I hope it's ok to do this in english.

First of all, what is this project about, for whom is this new computer?
It is the try to get new hardware for the remaining professional users, like from layout fields (Calamus) and from audio engeneering, as for Atari GEM users who still like to use GEM or MiNT for Multitasking on a daily basis. This computer is NOT primarily for Falcon Demos or Falcon games - in no case it is for "Retro-only interrested people". The entire project is a great  association of (former) commercial companies, of clone producers, of Atari users (like me), of Demo coders, or the FreeMiNT developers. It is the try to create an machine that has as much upgrade possiblities as possible - for the next decade! More than 100 people contributed the one or the other part to the project, or the first products FireBee and the mini-case or the new FreeMiNT setup! So there is no "single producer" like you might think. There is an Atari compatible computer that is done by us Atari users for us Atari users with great support from clone producers and (former) commercial developers.
The hardware itselve was mainly done by Medusa Computer Systems. The swiss company that produced the Medusa T40 and the Hades (040 and 060) in the past. As well involved is Wolfgang Förster who is doing the Suska boards. Aditionally Milan Computer helped with TOS and other things. So all clone produsers, who already prooved - for 4 times - that they can produce Atari compatible computers, are involved somehow!

This hardware that we have done is our best opportunity between costs and needed functionality. Of course we can discuss a PPC (a e600 core costs alone $400 vs. the ColdFire for $20), or emulation etc. But we decided to do a ColdFire clone as the compatibility to 68k is high, and some 95% compatibility can be reached one day.

You have to understand, that we are doing in our spare time a computer that has the ability to advance, to develope further and to become a real beast. Of course we are not there at the moment, but with support from more developers it can happen. The hardware is ready.  Further drivers and software need to be implemented, to let the project become a real success. But it has to be done by us together, not by any strange company from anywhere (and nowhere). The FireBee is our own Atari Users - means also your - computer, not just any product.

Please consider that there are many many different users out there, and several of them need the one or the other port. Like Serial, Paralell, etc. Today I had an US-User who has several SCSI devices and likes to wait with FireBee shipping until the SCSI protocoll is implemented inside the FPGA.

Now some direct answers:

Adam: of course the DSP is missing, but it was started, and everybody can continue. The code is GPL and at atariforge. Support is welcome. And yes, 68k Demos and Falcon games make no sense (recently). A ColdFire Demo with cool FPGA Stuff (512KB SRAM) would be nice ;)

evening: how much is a Falcon recently? How much is a CT6x, a NetUSBee plus CTPCI? And not everybody can search and solder it. With the Bee we got a ready systems that can - after a decade - again be presented at a dealer. It is ready for "unpacking and starting" even right now! Just get it and start to work with it. That is one of the main problems why no new users appeared. And Retro is still possible for those who like it, with original machines.

That it's not backward compatible is not true! About half of the tested GEM applications are running right now (without changes/adaptions). As soon as Mikro does his planned cf68klib replacement, and illegal insrtructions handler, the situation will massively improve! And why isn't it interresting? For example someone could implement a Supervidel directly to the FPGA, without any further hardware costs.

Cyprian: yes Cod3r is interresting, and he has much knowledge. About ColdFire boards, I am unsure. The FireBee was from the beginning on the cheapest ColdFIre board at all. Freescale took 850 for the Light Evaluationboards (without FPGA, without 20% of the connectors the FireBee got, without that much RAM, and with buggy PCI-Bus, and they produce in high volumen, and the own factories. We just do small volume series, ... ). There may be a cheap board, (like the dissapeared Cofilint from Japan), but with complete different things onboard.

sqward: I am stunned, you know Fredi Aschwanden personally, did he look like doing ridiculous things? The entire "Atari Ports" like ACSI, SCSI, Floppy, Printer,  Game-Port, MIDI, RJ 12 (yes not RJ11!), etc. etc. are only 50 Euros of the 600! And these ports assure continuation to existing Ataris. We had so many questions from professional Atari users, like mixing desk automation, cutting plotter users, etc. all of them needed these old ports. I had contact to 15 users who only bought the FireBee because of the Atari Ports. And we had to redesign the mini-case last years, as people demanded paralellports even in the mini-case!
Aditionally these ports are mainly FPGA pinouts. So an SCSI can become a Falcon port, or whatever you like!

jury: you are right, except the fact, that nobody is an IKBD fanatic at Medusa ;) We simply had IKBD running at the Suska, and no ready USB driver in the beginning 3 years ago. Now people often use original Atari keyboards, and are very happy.

Saulot: it was not a stupid idea with the "legacy interfaces" as said above, they are cheap. The price for the FireBee is mainly because of the low volumes. If we could do 3000 at once, it would get better, ... Even if we would have cut all legacy ports, put the RAM down to 128MB exclude 2 layers, etc. the price would be at 500 Euros. Thats it if you do small series. If someone can do cheaper, everything is free all schemas, even the layout!

About the background interview, look here for example:  http://forum.atari-home.de/index.php?topic=2536.0
Wolfgang Förster and Fredi Aschwanden are not some homebrewers! They have companies and know their business of hardware production.

Cyprian: it is much for eastern european. I am especially sad about russia, where people cannot pay 3months income for a FireBee for example. But show me that western european who earns 2500 a month, ... I live in Vienna, the 7th rich country of the world, and I have to live with 600 Euros per month! My "girlfriend" finished university and has to work for 850 Euros. So have many of my friends.
About your next sentence, the automatic translations gives me: "I think the price was € 200 bucks to optimal, and 100 € Expected" I do not understand this. ?

Candle: thanks for the "junk" ;) about "non-profit" I am throwing in 30 hours a week for  3 and 1/2 years now - for nothing except being able to buy a new clone (all developers bought the FireBee theirselves!). So do many many other Atari developers. The 600 Euros are the price for just the hardware in small quantities. Get the shema, and get better quotes, and than run 250 units from somewher in Asia (if you be able to get it for 100,- cheaper) and see how to deal with that 125.000 Euros! ;)
If a blitter is trivial, please make one for us! (But Wolfgang Förster already got one and is recently implementing it, so perhaps wait, and improve it later). Yes 8 layers. Please show us how to layout, if these Atariclone-producers all have no idea about anything, ...

The PCI card form factor of the FireBee is to fit to every possible enclosure; like mini case, DT or Tower or 19" rack with Backplane, Notebook etc. Just consider if the PCI would have been onboard (e.g. MiniITX) we would have just standard possibilities as the additional PCI cards would resoult in an 90° angel.

evening: about the compatibility: we are at "Alpha Edition" in our spare time, what do you expect, to have all our aims ready?

AS ... : I am a Hades user. A Hades is a damn cool and stable GEM machine. Nothing more nothing less. No Falcon games of course, but that was never the aim!


The Suska: Is very interresting and a replacement with timing exact implementations. Wolfgang is working on the 68030 and those who demand a new Falcon for games and Demos will than have it. As well Wolfgang is planning a cheap version of the Suska without most ports, just for putting games to SD-Card and play or use classic Atari applications. Than we got all fields done.

Cheap Suska for Retrogaming, Demos etc.
Suska for 1:1 Falcon replacement
FireBee for GEM and serious every day computing
Aranym as Open Source Emulator, crosscompiling etc.
Hatari for Falcon Emulation

The Atari Szene was never so good the last decade!

I hope the automatic translation did not scramble too many of your postings.

thanks for reding

Mathias

Edit: just some typos I read ;)

8

(17 odpowiedzi, napisanych Miejsca w sieci)

artik-wroc napisał/a:

Ruszyły zapisy na Coldari :)
http://www.medusacomputer.com/coldari.html

I´d like to inform you that we decidet to go on as Atari Coldfire Project.
It´s a free project without any commercial interrests. Fredi Aschwanden of Medusacomputer is the main hardwaredesigner, but there are 22 persons involved at the moment.

Look at our english news at http://acp.atari.org/
We today also published a simplyfied schematic of our new computer.
... ;)

If anybody is interrested in translating our news to polish, please contact me.

Regards

Mathias