Was anything ever made for the 1090 expansion box?

[DavidMil]

Was anything ever made for the 1090XL expansion box? I had part of one a long time (everything except the brown

plastic front). I gave it away along with a dead 830 modem about 15 years ago. I wish I could remember who I gave

it to...

 

DavidMil

[_The Doctor__]

http://www.cyberroach.com/cyromag/14/DSCN3932.jpg

 

notice the boards are populated?

 

There were 5 boards... only 3 are ever talked about...

 

while internal ideas for 8 slots are talked about, I have on ever seen a number of 5 slot boxes and only 1 box with 7 slots...that's it.. yes 7 not 8.

 

CP/M Z*80, 80 col / graphics, dual personality memory, 1090 shared memory(1090 devices only) , I/O card, network card.

 

some of the cards had a header at the top for other purposes but I don't see them.... I am certain if industrious people look they will find it all, some of it here on Atari Age, and the rest floating around cyber space. search for 1060... Atari CP/M 2.2 etc.

 

you will probably find a list of ideas planned or otherwise... I'm missing a card... it might have been a developer card or something to with the 1400/1450

[_The Doctor__]

I'd be surprised if Albert or Curt don't have all of the paper as well as the devices themselves...

[Level42]

There's some serious cash on that table..... ;D

[DavidMil]

I guess I'll have to do some research. As well as see if I can track down my old one.

 

Thanks Doc and Level42,

David

[+slx]

There's some serious cash on that table..... ;D

 

It certainly should not be left unguarded if Atari collectors are present....

[+DrVenkman]

Wasn't Dropcheck working on a sort of re-imagined 1090XL expansion device for the PBI, along with sounding out folks for ideas for new expansion boards?

[Gunstar]

Wasn't Dropcheck working on a sort of re-imagined 1090XL expansion device for the PBI, along with sounding out folks for ideas for new expansion boards?

That's what I heard and she damn well better be still! I installed the PBI in my 1200XL in anticipation! Otherwise I'll have to settle for an IDE 2.0 or find a BB/Multi-IO!

 

The last news of it on the site, from July 2017 says:

 

On the 1090XLR front, I am happy to report that Roland Scholz, who developed the prototype PCBridge some years ago has successfully developed the first internal expansion card for the 1090XLR project. The card is a combination dual serial/single parallel port I/O card. Both the 1090XLR and card are still in prototyping mode. tf_hh is also working on a version for the 1090XLR of his SysCheck device. Progress should go much quicker once both have returned from their much deserved vacations this month.

Edited January 12, 2018 by Gunstar

[Dropcheck]

That's what I heard and she damn well better be still! I installed the PBI in my 1200XL in anticipation! Otherwise I'll have to settle for an IDE 2.0 or find a BB/Multi-IO!

 

The last news of it on the site, from July 2017 says:

 

On the 1090XLR front, I am happy to report that Roland Scholz, who developed the prototype PCBridge some years ago has successfully developed the first internal expansion card for the 1090XLR project. The card is a combination dual serial/single parallel port I/O card. Both the 1090XLR and card are still in prototyping mode. tf_hh is also working on a version for the 1090XLR of his SysCheck device. Progress should go much quicker once both have returned from their much deserved vacations this month.

 

I am..... I am!

 

tf_hh ran into some issues with his redesign for the SysCheck. Then life interrupted and the last I heard from him was that he was planning on returning to it this month. I ran into similar issues in life too. The 1088XEL came along and I temporarily set the 1090XLR work aside for that. So yes, it's been a long time since there was any update on the project.

 

There is a working prototype of the 1090XLR. But housing the finished product is the next big hurdle. I have been collecting information and possible design specs over the last few months even if there's been no updates. It's no easy task to find or design a case that will be acceptable for most and fit all PBI equiped machines and be electrically safe and be economically feasible to do all on a shoe string development budget.

 

It looks like there will be a major design branching of the 1090XLR project over the next few months. One for the original 1090XLR project and the second towards a combined 1088XEL and 1090XLR motherboard. It just makes sense to combine the Atari and the expansion option in the same case. More details will follow as they are finalized.

[_The Doctor__]

I'm still in favor of a flat pack case... it will save so much time money and aggravation, 8 to 12 screws.. no problem

 

Unpack attach the 4 sides pop the cards in the slots throw on the top, off to the races!

Edited January 13, 2018 by _The Doctor__

[Kyle22]

I like the idea of an XEL with slots.

[_The Doctor__]

he he he heh did he say slots?

[Kyle22]

Uhhh... Yeah.. Huh huh huh...

 

Shut up Butthead, I said it first, dill hole.

 

[+Spaced Cowboy]

So I did something similar to this a long time ago when I was going through many a prototype for my fish-tank controller. I was using an AVR chip, not the XL, but the principle remains the same - I wanted a standard bus to plug in different modules without having to get the entire PCB re-made every time. Fish tank controllers need to take in loads of sensor inputs,make decisions, and tell relays to turn on, or alter a PWM signal, or control an analogue profile, they're surprisingly complicated.

 

Anyway, faced with having 6 or so input modules, and a dozen or so output modules, making a bus architecture that I could plug modules into and making them support auto-discovery of modules so I could integrate them into the main UI seemed like the way to go.

 

Electrically (obviously not adhering to the spec in any way), I chose to use a PCIe x4 socket as my standard interface - this meant I had 64 lines to play with, the connectors themselves are relatively cheap for high-density board-to-board connectors (about $2 each in quantity-10), and (crucially when you're making lots of prototypes) the cards themselves don't need anything other than copper laid down in a pattern along an edge. In addition, they're keyed, so you can't plug the card in the wrong way around.

 

As for the case, well, shapeways et al are good for the final product, but I have a 3D printer (and a CNC machine) so I never investigated the commercial side too much. OpenSCAD or Tinkercad (depending on your preference for creating 3D being algorithmic or visual) are good options for creating STL files, and after that, it's just a matter of paying money or doing it yourself. I decided that 3" (tall) x 5" (long) would be fine in the day and age of SMT, with a 1" separation between cards. I put in some guide-slots, and figured I could buy low-profile (which are 3" tall) expansion card brackets (you can get pre-punched ones for common ports) as support. Basically, it's a standard expansion chassis.

 

In terms of how it worked, I used the VME bus as a model. I had 64 lines, so there was always an AVR that managed the bus as the bus arbiter, and a bus-master (here, the Atari or any card) would request the bus using a dedicated line per slot, the bus-grant was encoded using standard binary (4 bits for 16 devices) because that could be shared, and I used a further 4 bits tied high/low to encode each slot's ID.

 

That way, the device knew its slot-ID, could signal on its dedicated line, and wait for the bus-grant lines to be the same as its slot-ID. Once the card has finished with the bus, it lowers its private bus-request line which is a signal to the arbiter that the bus can transition to idle again. This scheme used a total of 16+4+4 or 24 lines for bus management, leaving me with 40 lines for actual signals, which was plenty.

 

Whether this sort of bus management would work on the Atari 8-bits, I'm not sure - it's been a long time (we're talking geological time here) since I looked at the PBI driver code, but if you're planning on having multiple cards, you'll need some sort of arbitration going on, and the above is pretty simple.

 

It also lends itself to bus mastering, so if card A wants to talk to card B, the bus arbiter can dissociate the bus from the host (the Atari) and they can get on with the bus transfers with the host none-the-wiser. If the host (or any other card) wants to use the bus in the mean time, the arbiter will say the bus is busy and the host/card will wait/poll.

 

One other thing worth mentioning is that if you don't want a large ribbon cable going from computer to cage, the PBI only runs at less than 2 MHz. It's easy to get a SPI signal running at 50MHz or so using a low-cost Arm chip (STM32 will do it and it's 5v tolerant) and it's easy to buffer that signal at each end to make it a differential signal in-between. That will cut your wire count down dramatically to ~8 wires, including VCC and GND. That's a lot more convenient for running to an external cage, not to mention more robust signaling due to the differential wiring. You could also use the STM32 as the bus arbiter.

 

I'm gonna stop here because apparently I have to go and get the BBQ for the evening meal, but hopefully there's something useful in the above...

Edited January 13, 2018 by Spaced Cowboy