1090XL remake

[kenames99]

hi,

  A while ago, before Curt Vendel passed away, we were colloberating on a reproduction of to 1090xl expansion box. I was remaking the board and Curt was going to do the mechanicals parts. I do not know how far he got but I got the board design done. I never had the chance to get it produced. would there be any interest for me to put this on a github repo?

 

thanks

Ken

 

[Gavin1968]

Hey Ken,

Absolutely! I would love to make some of these. I am sure others would as well. Would love to have the CAD files. 

Gavin

 

[leech]

Who else has the capabilities to help finish this up?  Would be pretty sweet to get the 1090XL after all these years...

[kogden]

Yeah I'd certainly be interested in taking a look at this.  What format are the design files in?  KiCAD?

[bfollowell]

Did we have any good examples of any of the add-on cards? I can't even remember what all was planned for it, let alone what made it out into the wild after Atari went under.

 

Was this a project just to remake the 1090XL itself, just to say we/you did, or was the plan to have several of the cards available as well?

 

Great work either way! I'm sorry that Curt didn't get to complete his portion before his passing.

[+DrVenkman]

I seem to remember @Dropcheck was working on a similar project a couple years ago. 

 

As for cases and mechanical parts, if they can be 3D printed, someone like @Mr Robot is a genius and could do amazing work that stylistically respects and builds on the original. 

[+MrFish]

52 minutes ago, bfollowell said:

Did we have any good examples of any of the add-on cards? I can't even remember what all was planned for it, let alone what made it out into the wild after Atari went under.

Atari 1090 XL Expansion Cards

 

 

[danwinslow]

Well. Can't help with the hardware side directly, but I'd sure put some money up.

[_The Doctor__]

Absolutely!

 

The last conversations with with Curt were about the buffered Julia cable and a few other iterations of what could be used to connect to the 1090XL remake, an idea was to use the cables from old TI expansion boxes and maybe track down what sources of that or similar cable could be.

 

Did you discuss the different buffering methods and location of the buffer on the cable? My thoughts were to keep the buffer as close to the Atari as possible and make whatever corrections that might still needed at 1090XL box itself. I don't think the project should be considered ended, let the estate know about the project and those that were involved Ken. Maybe it will still be possible to continue the project. Or perhaps others can pick up the torch.

 

It's a bit different slots with integrated power supply. Still was of the mind an off the shelf supply was going to be used, but doing so would eat up space inside the expansion box. Thanks for showing what was in the works and offering to place it on github.

[Dropcheck]

A lot of people have purposed various designs over the years.  The one common thing has been the massive amount of time, money and expertise that is required to bring the 1090 expansion box to fruition.  Remember this box was never released to the public as a finished product, so there are a lot of grey areas and gotchas that never got fleshed out.  I was working with three other people for a time to create a modified version of the 1090.  But then got bit by the time/money/expertise juggernaut.  It's not that it can't be done, just that you have to have all three items in spades.  Two out of three won't get you there. 

 

You also have to go in very defined steps.  You can't start with an everything including the kitchen sink approach.   Get the basic setup up and working reliably.  And then you have to decide how true to the original specs you are willing to stay.  Is the Atari computer the host or the client? 

 

I wish anyone who attempts to work on this kind of project well.  If you need some background I've collected and published what info and work I was able to accomplish here.

[+x=usr(1536)]

Speaking as someone who missed out on the 1400XL/1450XLD boards, I would absolutely love to have a 1090 PCB and a BOM

3 hours ago, Dropcheck said:

Remember this box was never released to the public as a finished product, so there are a lot of grey areas and gotchas that never got fleshed out.  I was working with three other people for a time to create a modified version of the 1090.

Agreed, and it would make sense to build something that is basically the same spec as the known-closest-to-final Atari revision is.  Anyone who wanted to make expansion cards for it could then work from a common set of known specs that should have a high degree of compatibility with both PBI and ECI (assuming that a 1090 can work with ECI via an adapter) machines.

[+kheller2]

Doing a prototype 1090 and 64k card is probably the easiest place to start since we have more documentation for that than anything else.

 

[Mathy]

Hello Ken

 

20 hours ago, kenames99 said:

 A while ago, before Curt Vendel passed away, we were colloberating on a reproduction of to 1090xl expansion box.

 

Is this a 1:1 copy of the original?

 

Sincerely

 

Mathy

 

[Mark2008]

So cool.  I know you all are doing it for the love of the hobby, and it is really neat.

 

My opinion is probably completely obvious, but I would not expect expansion cards to be developed.  That's just my opinion, the attach rate" for the 1090 won't be high.   However, if as part of the initial effort to develop the 1090 some cards came with it, those could be quite tempting.

 

Personally, I have external memory expansions, hard drive solutions, parallel and serial ports, even cp/m - and more critically they are all still easy to buy.  But what I don't have is an external 80 column solution.  VBXE (I own 2) is not external and its signal is not supported by a wide range of equipment i.e not even recognized by any of my scart monitors or even 15khz capable vga monitors.  xep80 may be external, but it cannot be bought.  80-column looms large for me as the most likely "must have" expansion.

 

But, if in one box, one beautifully designed case, I could have memory expansion, hard drive, and i/o, I'd be a sucker for the beauty of that as well.

 

Anyway, best of luck to you all, it's quite interesting to follow the progress.

 

 

Edited March 15, 2021 by Mark2008

[danwinslow]

i don't know that I personally would want an EXACT copy. Seems like we could do it better/faster/cheaper with at least some modern tech and still preserve the essential 8 bit nature. Easier to source parts, as well.

[leech]

1 minute ago, danwinslow said:

i don't know that I personally would want an EXACT copy. Seems like we could do it better/faster/cheaper with at least some modern tech and still preserve the essential 8 bit nature. Easier to source parts, as well.

Yeah, I mean the ones I would be interested in is the memory and the 80 column display card!

[kenames99]

2 hours ago, Mathy said:

Hello Ken

 

 

Is this a 1:1 copy of the original?

 

Sincerely

 

Mathy

 

yes, as close as I could get without an actual board in hand.

 

Ken

 

[_The Doctor__]

My thought was that it was going to accept the 1090 boards but have added capability on an additional 1 or 2 slots using an off the shelf power supply.  The other was to continue with a board or ribbon cable across the top to carry added signals for enhanced pcbs. To clarify was that relayed to you?

Edited March 16, 2021 by _The Doctor__

[+x=usr(1536)]

1 hour ago, danwinslow said:

i don't know that I personally would want an EXACT copy. Seems like we could do it better/faster/cheaper with at least some modern tech and still preserve the essential 8 bit nature. Easier to source parts, as well.

 

True, but if the PCB is already designed and can use off-the-shelf components, there's no reason not to.

 

Having said that: I'm not in total disagreement with what you're proposing, and can see some definite advantages to it.  However, there is an opportunity here to recreate something that Atari never released for sale but that was getting close to being ready to market.  This is where I see the expansion cards as being beneficial: if people want to build Z80, 80-column, etc. cards, they can - but it also leaves room for a FujiNet-style approach where modern hardware interfaces with the machine using its stock expansion slots and bus.

 

Note that I'm saying this as someone looking right at both a FujiNet and a Concerto sitting on his desk, so I'm in no way opposed to modern hardware being used with 8-bit machines

[leech]

8 minutes ago, x=usr(1536) said:

 

True, but if the PCB is already designed and can use off-the-shelf components, there's no reason not to.

 

Having said that: I'm not in total disagreement with what you're proposing, and can see some definite advantages to it.  However, there is an opportunity here to recreate something that Atari never released for sale but that was getting close to being ready to market.  This is where I see the expansion cards as being beneficial: if people want to build Z80, 80-column, etc. cards, they can - but it also leaves room for a FujiNet-style approach where modern hardware interfaces with the machine using its stock expansion slots and bus.

 

Note that I'm saying this as someone looking right at both a FujiNet and a Concerto sitting on his desk, so I'm in no way opposed to modern hardware being used with 8-bit machines

I still need to get a Concerto... also an 80 column display for my 800 or 1200xl... or maybe one of the 800xls, can't decide which I want to type on..

[Sugarland]

Couldn't a LOT of money and time and sweat and tears by saved by combining everything into one PBI device? Consolidation is where the market moved by including sound cards, network adapters, drive controllers, USB/serial, wireless, etc onto the motherboard. 

[Dropcheck]

5 hours ago, Sugarland said:

Quote

Couldn't a LOT of money and time and sweat and tears by saved by combining everything into one PBI device?

In essence that is what the 1090 expansion box is, with the added benefit of it being adaptable to changing wants and needs.  Insert upto 5/8 cards of whatever feature set you want in one chassis.  Plug it into your 800XL or 130XE and away you go.  

 

Quote

Consolidation is where the market moved by including sound cards, network adapters, drive controllers, USB/serial, wireless, etc onto the motherboard. 

The downside to moving everything onto one main motherboard is the lack of flexibility.  It might reduce the cost some, but now you have to replace the entire motherboard to get updated/additional features or get stuck with features you don't use or want.  Or heaven forbid have a part fail.  Don't forget that super motherboard with all the bells and whistles needs a new type of case to allow connection to all the new fangled other devices you can connect it to now. 

 

All of these ideas are a trade off in one form or another.  Whether you produce a single main motherboard or an add on box, you really only reduce cost by production in the thousands or millions, not the few hundreds likely to be built. ?

 

[danwinslow]

True enough, but using modern board production and components would be somewhat cheaper. Sourcing ancient parts and getting weird old boards made would be a lot more custom and more expensive per unit even if the run were in the hundreds. I think. Anyway, I kind of agree with Sugarland, I don't have any nostalgic feelings about righting the wrong that Atari never got the 1090 out the door, and so we now need to recreate it. What I want is utility. If it comes in a nice 1090 box, all the better. I want plug in boards for things that now take soldering, basically. Video cards, accelerator boards, etc. Who knows what we could do with a good PBI bus?

 

I agree that it should have a basis in the Atari bus idea, I mean I'm not suggesting we just attach a modern motherboard or anything. Maybe we should combine this with the Atari Backplane Computer being discussed elsewhere.

[+Spancho]

What were the main benefits of a 1090XL design BITD?

In my opinion Atari addresses the following:

 

1) Get more slots (PBI exposed slots)

2) Buffering the signals

3) do already address decoding for the needed PBI ID

4) additional common power supply

5) maybe more I don`t know

 

The points 2 to 4 for sure aimed to make expansion card design cheaper and easier BITD.

By todays technology that is maybe not so importnat anymore.

 

My hope would be that a 1090 could be an enabler to produce more PBI expansion cards which can co-exist and the entry barrier for hobbyists which are not perfect in electronics were lower.

 

In the end it`s a hobby and there is no good reason for doing things than "It´s fun and joy".

Edited March 16, 2021 by Spancho
typo

[+x=usr(1536)]

1 hour ago, danwinslow said:

True enough, but using modern board production and components would be somewhat cheaper. Sourcing ancient parts and getting weird old boards made would be a lot more custom and more expensive per unit even if the run were in the hundreds.

Maybe, but it depends on how exotic you'd need to get in order to build this.

 

For the PCB, there are any number of places that can make them; just send them your gerbers and they'll run it off.  WRT the components, looking at the 1090 PCB sketch in the first post, it appears as though the majority of components consist of resistors, capacitors, possibly a small number of transistors, ICs, and card connectors.  Provided that the ICs are 74-series logic, they should still be available; if not, there (generally) are substitutes.  The card connectors, depending on pitch and width, might be more difficult but there are ways to work around that.  None of the parts involved should be terribly expensive, though the card connectors would probably be the single most-expensive component of the BOM.