Atari 1090XL Schematics?

[Dropcheck]

Does anyone have schematics to this box? I'd love to get my hands on it.

[Rybags]

ed - misread title, thought it said 1064.

 

ed2 - Supposedly Curt has schematics, no idea if he's publicised them http://atariage.com/forums/topic/108759-a-reproduction-of-the-atari-1090xl/?do=findComment&comment=1315527

 

Found elsewhere on AtariAge, technical notes with some handy info but no schematics.

1090XL Technical Notes.pdf

Edited May 9, 2016 by Rybags

[sup8pdct]

Not much on it. power supply, couple of buffer chips and 5 slots.

 

James

[Dropcheck]

Rybags,

 

Thanks for the technical doc.

 

That thread is old.

 

I sent Curt a PM to see if he is still working on it or would be willing to hand it off if he is too busy. It'd be nice if we could get a prototype up and working, even if at first there wasn't a case to put it in.

[Rybags]

I think some of this was discussed in that thread and possibly elsewhere.

 

The box as it is is somewhat outdated and in the modern day the form factor would be totally unnecessary.

The weakness of the PBI and ECI is that they don't offer all the signals to make external expansions fully usable.

e.g. missing from one/both are stuff like /Halt, CSync. In fact it might even be desirable to have the AN0-2 bus represented, it could allow mirroring of GTIA functions e.g. for alternate video generation.

[Mathy]

Hello Rybags

 

HALT is pin C on the ECI.

 

The ECI has one unused/reserved pin, the PBI has three (not counting 47 and 48, which are used in the 600XL for +5VDC). Meaning we can get an extra signal out.

 

Sincerely

 

Mathy

Edited May 9, 2016 by Mathy

[Bryan]

I've always thought that the best way to make an expansion box is to put something like a 6507 and a 256-byte dual-port RAM in it. That way things don't have to sync up in real-time and the bus issues go away. The host machine issues a command to the expansion CPU and it returns the data as soon as it can or vice versa. Then it could be built to take any 8-bit ISA cards.

 

A CGA card, AdLib, and Serial/Parallel card might be a nice start.

[tschak909]

like a DS1609 part?

 

-Thom

[Bryan]

like a DS1609 part?

 

-Thom

Yep, build an interface with a 6507, DS1609, a ROM, and some TTL interface stuff and it should work fine. You could also have another SRAM on there for more advanced stuff or co-processing.

[tschak909]

yeah, I'd want a bit more RAM, to leave all of page 0 free.

 

-Thom

[Bryan]

Here's an example of interfacing with simple IO cards:

 

https://web.archive.org/web/20130308031815/http://mycorner.no-ip.org/6502/vic20/isa/index.html

 

Things get more complicated when the card has a memory interface (like video cards).

[Rybags]

Yep, ISA would be fairly simple. Problem is though they're getting rare to the point where you couldn't really do a public project where a given card is expected to be used, maybe early Soundblaster but can't think of any others.

[ricortes]

Where would the project end up? It seems like the concept of plug in card cage found most of its utility in the 80s through ~the year 2000 while everyone was searching for killer apps. I mean at one time things like SCSI cards made sense because you could hang on a CD ROM, flat bed scanner, printer, back up tape, hard drive, et cetera. Now even if you had the ability to add things like that, the market has left them behind. There are also problems is the Atari just being a bottle neck for data transfer. If you developed say a VGA card, typical images would take ~256K of data tossing between the Atari storage and display. Ditto for things like a camera in that it what we have come to expect as far as real time image capture and/or display would be limited by the processing power of the Atari.

 

I'm not against the concept, just need help coming up with a reason for it to be necessary even as a hobby. Kind of like if you came up with a CP/M card, nice achievement and a good flaunt of technical expertise, but not exactly a must have in this day and age. I'd probably buy it, but then I am known for the impractical and useless.

[Dropcheck]

Where would the project end up? It seems like the concept of plug in card cage found most of its utility in the 80s through ~the year 2000 while everyone was searching for killer apps. I mean at one time things like SCSI cards made sense because you could hang on a CD ROM, flat bed scanner, printer, back up tape, hard drive, et cetera. Now even if you had the ability to add things like that, the market has left them behind. There are also problems is the Atari just being a bottle neck for data transfer. If you developed say a VGA card, typical images would take ~256K of data tossing between the Atari storage and display. Ditto for things like a camera in that it what we have come to expect as far as real time image capture and/or display would be limited by the processing power of the Atari.

 

I'm not against the concept, just need help coming up with a reason for it to be necessary even as a hobby. Kind of like if you came up with a CP/M card, nice achievement and a good flaunt of technical expertise, but not exactly a must have in this day and age. I'd probably buy it, but then I am known for the impractical and useless.

 

 

I wish I could say I have something practical and useful in mind. At this point I would just like to get a basic box up and running. Then I get to figure out how to justify it.

 

Maybe the only realistic practical/useful thing is moving some of the current internal mods to the external box. It's getting crowded under the hood.

[flashjazzcat]

IMO, that would be the most useful thing: something to house all the naked PCBs hanging off the expansion bus. That would require a breakout board with a number of male edge connectors sticking up at right angles, or 50 way pin headers for short ribbon cables.

 

Problem there is that most modern ECI/PBI gadgets are all-in-one devices with RAM/HDD/RTC/cart pass-thru or any mixture thereof and and bristling with buttons and switches which need to remain accessible somehow. So I think - realistically - you'd have to hope for new boards specifically designed to be housed in the 1090, and then - assuming that happened - you'd have to think of some way of making freezer switches and HDD partition swap buttons accessible without ruining the aesthetics.

Edited May 11, 2016 by flashjazzcat

[Bryan]

you'd have to think of some way of making freezer switches and HDD partition swap buttons accessible without ruining the aesthetics.

 

I suppose the housing could have a display and buttons and each card that wanted to could be configurable through a common screen. You would just have a protocol to integrate them all into a common menu system.

[+SoulBuster]

To make buttons accessible, the box would need to be designed so that when a card is inserted, one end of the card is exposed on the front with all the buttons needed for operation of that card. Very similar to cards of today's desktops that expose ports on the rear of the computer.

[Dropcheck]

 

I suppose the housing could have a display and buttons and each card that wanted to could be configurable through a common screen. You would just have a protocol to integrate them all into a common menu system.

 

That would be great. Except that would make the box a smart box and quadruple the cost. I'm thinking at the very least a big cpld or even a Pi 2 or 3 with a touch screen.

[Bryan]

 

That would be great. Except that would make the box a smart box and quadruple the cost. I'm thinking at the very least a big cpld or even a Pi 2 or 3 with a touch screen.

Nah, a cheap $2-$3 micro, an LCD module and an I2C bus shared among the cards would do it.

[Dropcheck]

Nah, a cheap $2-$3 micro, an LCD module and an I2C bus shared among the cards would do it.

 

 

Would you have a schematic example?

[Bryan]

Would you have a schematic example?

This is the first thing that came up in a Google search:

 

http://maxembedded.com/2011/06/lcd-interfacing-with-avr/

[ricortes]

You don't need to much HP to drive one of those LCD's. Here's an example of a schematic that could be interfaced to an Atari using one chip<74ls00> and the existing $D100 for the PBI or $D500 line from the cart.

http://6502.org/mini-projects/optrexlcd/lcd6502.gif

 

*BUT* getting everyone to agree on what should be done is like trying to herd kittens! For instance, I would like it if every card followed the ~IBM type standard of including the code for itself in ROM. People making the cards would probably just as soon include a handler loaded from disk so they don't have to add ROM to a cost sensitive project. I mention this not as a criticism but a tip of the hat to people that tackle it.

 

I personally would like to see one of those cell phone displays that are finding their ways into Arduino projects used, but I know it even goes against what I have already said. The ability to address 100's of k of display memory with a 6507 would slow the system to a crawl. Adding an AVR to the mix will ruin some people's idea of purity.

 

Heck, I am probably in no matter what it ends up being.

[Bryan]

I'm starting to think of a box with an 8088. They're cheap ($3-5 on ebay), interfacing them to ISA is a no brainer, and you could possibly make the box capable of running MS-DOS. That would make it kinda like the CP/M add-ons that existed in the day.

[Mathy]

Hello guys (m/f)

 

Why ISA? The 1090 has its own semi-standard. And with modern electronics it should be easier to make something much smaller than the average ISA card. And less power consuming. And more reliable. And maybe even easier to find.

 

The new cards inserted into the 1090 would probably be no bigger in physical size than 5 x 10 cm (about 2 by 4 inches). It should be possible to combine cards that have their own drivers on board with those that don't. And some of those could have an extra port (I2C?) for exchanging data that doesn't have to be shared with the Atari. This internal 1090 bus (the bus not connected to the Atari) could be connected to for instance a small display and/or a couple of switches/push buttons.

 

Sincerely

 

Mathy

[Dropcheck]

Collecting info on PBI and 1090XL

 

So far:

 

PBI Specs

 

1090XL Technical Notes