Home  News  Recent  Forums  Search  Contact


   Register here

 Main menu
   View images
   BBCode test
   Statement of intent
   Terms Of Service
   IRC Channel
   List Content

 In cooperation with

  Amiga Flame
  Amiga Spirit
  View comments
      AmigaOS 4  
mechanicXena/Xorro First Board20110319 22:23

I hope it is still not too soon to again bring the subject of Xena/Xorro
forward. The latest A-eon news has prompted me to do so.

Although I'm not against any suggestions for putting the integrated
Xmos to use, I feel the best/first solution would be one that every
purchaser could use if not for practical purposes, then as a learning
tool. Something to come to grips with the way and how Xena/Xorro/PCIe
is going to work and act to various input and output.

The more esoteric uses can follow only if enough people can see the
possibilities and perhaps test and play with their ideas.

A first board should inspire to ask questions while enabling all to
join in discussions and test others solutions.

And so, simple, practical, easy, usable by all, and of course doable.

???????????????? :-)
Navigate: 1-20 
OldAmiganRe: Xena/Xorro First Board20110320 00:00  #3006

Not sure what you mean here. The news refers to the first 5 boards which are a proof sample. Presumably, if all is well, that is what will be produced for the beta testers.

XMOS have various tools downloadable from their website for PC users to use. I believe some will be made available for X1000 users to download for PPC.

Surely, once the boards are available for general purchase, your other questions will be answered and be available to all.
OldAmiganRe: Xena/Xorro First Board20110320 00:00  #3007
Oops, bump
mechanicRe: Xena/Xorro First Board20110320 00:27  #3008

Although, as TrevorD said, it will be possible to use Xena with a
minimal amount of wiring I am sure not many users would be able to make
a card that would tie Xorro and the inline PCIe slot into a usable
device. Even if everybody made their own card there would be very
little in common amongst them and make it difficult to discuss much
related to learning or tinkering and helping others.

Most people are going to need the knowledge tools before designing
or building a project first. As Xena is also tied into the CPU it
may be helpful if we had a common card to 'play' with.
derfsRe: Xena/Xorro First Board20110320 02:03  #3009

the problem with a common card, is how common do you want it?

how about a standard 25 or 37 pin connector on it, but then you have some pins that need to be powered, what if a differant voltage is needed? people need to make their own cables to do different things.

i say let beta testers get the boards, get the tools for xena on OS4.1, see what these people can do/make, and a-eon can pick a nice idea for when its released properly.
ChrisHRe: Xena/Xorro First Board20110320 17:48  #3028
Just a simple board which connected all the internal Xorro (PCI-e type) connector to an external & easily user-accessible port would make Xena projects 10 or 100 times more likely. If everyone has to design & manufacture their own custom PCB board before they can try anything, then IMHO it's a no-go for most people who just want to try tinkering.

If someone with a "simple board" manages to make something interesting, then they are going to be much more likely to put the effort into getting a custom PCB built for a final "professional" result.

As to what connector they use, it doesn't really matter, as they're still going to have to wire each pin up themselves. All we need is the pins to be made easily accessible (and identifiable, so as to avoid any nasty mistakes, e.g. connector 5v to ground).

Perhaps the board could have two modes: The default mode would be a "safe" mode where all the pins went though some kind of diodes and resistors to prevent nasty accidents (e.g. 5v to gnd) damaging the mobo, and an "advanced" mode (perhaps changed via a physical switch) to remove all that protection in case it caused problems with certain kinds of signals. Or forget the switch, and just manufacrure two different PCBs, one "safe" and one "advanced". Ship the "safe" PCB with the X1000, and allow the "advanced" one to be ordered.
mechanicRe: Xena/Xorro First Board20110320 19:17  #3034

That's more like it. Especially if there were a basic kit(?) to use in
exploring possibilities.

Gee Whiz...When I works,I works. When I plays, I plays. When I thinks,,,,,
I falls asleep.
AlexCRe: Xena/Xorro First Board20110321 05:11  #3047

IMHO the most convenient approach would be to have a board with a decent-sized breadboard mounted on the inside, pre-wired to both the Xorro and PCIe slots using opto-couplers (if they don't create significant latency) to completely isolate the board electrically, and have a small piece of breadboard sticking out the back of the bracket, that way people can add any component they may want and connect any type of connector that fits breadboards, like these:


The board could then be sold bare or as a full kit with ICs, resistors, capacitors, diodes, jumpers, leds, plugs, etc and Xena code to assemble various functional projects like parallel port, digital joystick ports, ps/2 ports, 2nd sound card, LCD driver, etc.

That way it would be really easy to experiment with Xena and if someone comes up with a new application anybody with the full kit or needed parts could add it to his breadboard.

I think it would open the door for some really cool "standardized" (easily reproduced) hacks :-)
daedalusRe: Xena/Xorro First Board20110321 12:15  #3056
What I'd love to see (and may start work on as a project, if I decide to get an X1000 in the future) is a board which gives the user a Geekport at the back, compatible (or at least similar) to the one featured on BeBoxes back in the day. It had lots of inputs and outputs to keep the tinkerers happy, and suitable for many different purposes. If something like that was available, I'm sure it would suit most people's needs and let them start hacking away at other stuff...

GeekPort on HwB

Last edited by daedalus at 20110321 12:15
mechanicRe: Xena/Xorro First Board20110321 15:30  #3063
@daedalus + ChrisH

Yes. Both great ideas.

Still,,, something is needed for the, uh, non technical future geek.
(If that makes any sense.)

Perhaps incorporating these ideas along with something that could be
used as a discovery tool. Something people could use immediately not
only for specific purposes but also to test ideas and programming.
Something that does not take up very many functions/threads but could
be used to simulate various scenarios before committing to hardware.

I'm sure such a configuration could not be used for 'everything' but
even if it's only used as a learning tool and perhaps as a way to test
portions of larger projects/programs it might be of value.

......thinking..........What can 'amateurs' use and play with that
would be of value to the more technically minded.......?

DeniilRe: Xena/Xorro First Board20110321 16:35  #3064

So the Xorro is the only place the XMOS is wired? And it is a PCI-e type connector?

From a picture I noticed an array of something like an IDE or floppy connector (which I assume the X1000 doesn't have). I hoped that would be the access to Xena.
rigoRe: Xena/Xorro First Board20110321 16:51  #3065

The Xena chip is wired on one Serdes Lane, so this gives it a high-speed connection to the CPU. The "Xorro" slot is not PCI(e) comaptible in any way, although the actual slot looks mechanically similar.

The Xorro slot and a x1 PCI-e slot sit in line at the board location, and the logic is that a "bridgeboard" could be developed so that a card could bridge the Xena I/O lines into a PCI-e busmaster chip which can have access directly to the bus.

In esscence this means the chip can be programmed by the CPU at high-speed, and once running, Xena could theoretically have access to the PCI-e bus, and hence system memory via DMA. This is all dependent on the required components are on the "bridge" card, and where the hardware developers imagination stops. :)

DeniilRe: Xena/Xorro First Board20110321 23:20  #3082

I was thinking about the connector itself. Would have liked an easier to access connector that didn't require an actual card to insert but some pins that would be poked with half an IDE cable or something.

Bridging Xena to the main memory many not necessarily be a good idea. You may go outside the intentions of the chip and just end up with poor performance in that possibly foreign environment. But then again I have no clue :-)
rigoRe: Xena/Xorro First Board20110321 23:40  #3084

It may well be that a "breakout" board could be developed, which gives access to the I/O lines of Xena in a "breadboard" style way. Just because bridging Xorro and PCI-e is possible doesn't mean it is compulsory.

As with most general purpose connections, the limitation is usually the imagination.

mechanicRe: Xena/Xorro First Board20110322 00:59  #3086

'But then again I have no clue :-)'

And the reason for this thread.

I think having Xorro as a card slot is a good choice. The connector
you desire can be put on that card, or even several connectors.

This however, points out the need for a first common card. It would not
be feasible to produce cards for everyone's desires. Especially since
very few know exactly how their wanted card would interact with the way
Xena is wired into the system. That is why a fairly simple card that
may connect Xena/Xorro/PCIe into some device(s) that could use as many
of those functions as possible, while usable by just about anyone, might
be good.

If this card could allow the novice access to the technology and also
assist the technophiles to create, without completely gobbling up the
whole bandwidth/function/pin count of the combined on-board offerings
then expanding right on card using something like a header could be
possible. Perhaps even some optional in/out flashy LED gizmo could be

First.. a discovery tool. Light on the system, usable out of the box,
of use to all.

I think I need to ask relevant question(S)...maybe..????
mechanicRe: Xena/Xorro First Board20110323 17:42  #3137

''Bridging Xena to the main memory many not necessarily be a good idea''

If a 'first board' did use main memory for some purpose it could prove
very valuable as an example for tinkerers.

After some cogitation I believe a decent test/discover card for the
X1000 would have the following input. The output, except to drive the
following, would be left to the user to implement as everybody may have
different projects in mind. For testing, out could be simulated on the
computer screen or some LEDs.

Variable Detectors:
+ - values
Edge / Incident

Discrete Detectors:
Command / Create
Start / stop
Alert / Notify

Before working on how to implement this, are there any additional
functions to add to the list?

mechanicRe: Xena/Xorro First Board20110326 22:42  #3201

Deniil wrote:

Bridging Xena to the main memory many not necessarily be a good idea. You may go outside the intentions of the chip and just end up with poor performance in that possibly foreign environment.

My argument for a first card to link to the PCIe slot is probably best
reinforced in the XMOS documentation --Programming XC on XMOS Devices--
where it is described using the chip as a UART.

If used as a UART it seems its function would be to move data in and out of
the computer. So the question is;

Where does the data come from to send, and what do you do with the data

Unless your going to send, and expect to receive, the same data over and over
as with some embedded device, there isn't much need for having a computer.

OldAmiganRe: Xena/Xorro First Board20110413 16:23  #3209

Alook at the Xmos site provided some reference designs for PCBs and some ideas

RobRe: Xena/Xorro First Board20110422 06:06  #3210

rigo wrote:
[[he Xorro slot and a x1 PCI-e slot sit in line at the board location, and the logic is that a "bridgeboard" could be developed so that a card could bridge the Xena I/O lines into a PCI-e busmaster chip which can have access directly to the bus.[/quote]

A bridgeboard opens up the possibility of using the X1000 to simulate devices and the environments they are expected to operate in.

LyleHazeRe: Xena/Xorro First Board20110922 03:39  #3259

I look forward to all the possibilities.
Once a "product" is developed, connecting directly to the builtin XMOS chip on a custom circuit board is the way to go, but for an "experimenters" board, I'd get nervous about wiring directly to the builtin XMOS chip.

I'm still getting up to speed on them, but it appears that one of the unique properties of this chip is the ability to join them into X-networks to distribute the load across many chips. Both 2 wire and 5 wire links are available, with speeds approaching 30MByte/second. (reading docs, have not witnessed it yet)

Back to my point, if we want to make an experimenters board (great idea) it might be wise to have an XMOS chip on the board too, as an "interface" between the builtin XMOS and the breakout pins you described. That way, if mistakes are made in wiring your various bits and bobs, the damage should be limited to the experiementers board, and the XMOS on the motherboard should be "safe".

It would also be nice to support 2 or 5 wire links to additional external XMOS chips, to support building XNets of any size.. :)

In any case, it will be fun to explore the possibilities.

mechanicRe: Xena/Xorro First Board20110922 22:40  #3267

LyleHaze wrote:

It would also be nice to support 2 or 5 wire links to additional external XMOS chips, to support building XNets of any size.. :)

In any case, it will be fun to explore the possibilities.


There are 2 Xlinks in Xorro. So adding more Xmos should be no problem.
Building up a sub-system for whatever should be interesting.
Navigate: 1-20 
Snack! forum website engine, Created in 2008 by Björn Hagström