Atari 800 RAM Selection

[fibrewire]

As promised, the Mosaic 64K board with metallic ATASCII stickers - but no manual! Funny thing is, it looks identical to an Atari 16K board.

 

Yes, I paid $125 for a Mosaic 64K box... Go ahead - rub it in

[ClausB]

It shouldn't take me long to make a schematic of that, once again starting from Jerzy's 16K!

[fibrewire]

I contacted the guy who sold me the Mosaic card. Hopefully he still has his father's Atari 800 in the closet with the REAL Mosaic 64K card in it!

[kool kitty89]

I know this is a rather old post, but I missed it originally. (or forgot to respond to it)

 

you also need to take into account the type of people they were expecting to sell to... they were expecting to sell the 600XL to someone who wanted a computer, they were expecting to sell to someone who wanted a game console but could be later expanded into a full computer, which often times never happened, they would just buy a new machine...

 

add-ons werent made because of the fracturing of the market or line, it was because they wanted to make a whole new product. why would they sell you a $50 add-on with a 40% margin, when they could sell you a whole new product for $200 with the same (or even better) margin? why would atari want to sell you a 256k upgrade to your 800XL, when they could sell you a whole new 130XE? same with keyboard/mass storage for the 7800, why sell those things when you could be talked into buying a 800XL or even a 130XE? even tho the graphics and sound were more limited relatively, it was the ability to sell bigger profit items... simple business, ask any sales man, they want you to sell the big profit margin items over small margin, even if its something you dont want/need...

Yes, selling all-new systems could be more profitable, but only if the products remained competitive enough to allow that . . . and you'd potentially risk fracturing your market by NOT providing add-ons to older users (who may simply stick to the older systems instead), not to mention 3rd parties taking up the slack and profiting from such add-ons instead. (so not just competition from other computer makers, but other peripheral/upgrade manufacturers too)

 

You could totally omit any expansion-friendly features from a machine to cater to that business model (and block 3rd party upgrades), but that would potentially make an even bigger hole against competition with such expandability (and flexible,open-box expandability was one of the biggest hardware features that led to the Apple II and IBM-compatibles success -albeit there were other factors too).

On top of that, if you offer expandability, it adds to the value of the system, and potentially makes it more cost-competitive to competing platforms, be it due to their having such expandability, or cases like Commodore's vertical integration, which offered a major advantage in manufacturing cost as well. (which could only be competed with by genuinely more cost effective hardware, lower margins, buying up manufacturing firms to become vertically integrated themselves, or promoting other features -like expandability- which added value aside from out-of-the-box support -albeit there's other things, like peripheral support -and Atari's much faster disk drives- and software support . . . and software pricing)

 

 

Albeit, even worse than lacking expansion (for added value/competitive edge, and for general upgrades to improve transitions to new standards) is lacking backwards expandability and totally jumping to a new platform with no legacy support for established software. (that's arguably unimportant for a dedicated game console line -and many non-compatible consoles have been very successful, but for a computer line it's generally critical for maintained marketshare, so an efficient evolution of hardware -even without much expandability- is very important . . . unless the preceding machine was a total flop or had been pupular for a time but had long since declined prior to the launch of a new system, and thus starting fresh is more attractive -or there's legal issues that prevent continuation of the older system, etc)

 

Obviously, even with heavy provisions and support for expandability and compatible/evolutionary development of hardware, there's still inevitably points of incompatibility (both especially forewards compatibility -where add-ons become generally impractical), and that obviously happened with PCs too. (new generations of CPUs and expansion cards needing totally new motherboards) And you eventually lose some legacy support as well, or have it become generally much less accessible. (both in terms of OS and base hardware compatibility) But those changes are so gradual that there's no shock to the market with totally new systems needed for every upgrade, and most/all old software (and peripherals) being incompatible.

 

There's also cases of generally inefficient evolution of older hardware that leaves it largely tacked-on for compatibility only . . . which is OK when the older hardware is pretty simple/cheap to embed (especially if only some components are tacked-on with others inherently compatible or partially useful in the new system), but when that old hardware still makes up a considerable chunk of overall cost and doesn't mesh usefully with the new hardware/modes then you've got more problems. (rather like the C128 having the VIC-II and 6502 basically useless in most C128 modes, and a new -though more limited in some respects- VDC and CPU added on top of that)

 

Then there's cases where the evolution of the hardware simply lags too much to remain competitive or marketable (more of an R&D and management issue than anything else -especially when the original system was relatively simple and clean, with open-ended evolution prospects . . . like the ST -the Amiga was obviously far more complex to build onto . . . though what CBM ended up doing was a particularly bad example even in that context -both in terms of hardware design/features and the marketing/management issues alongside that)

And there's even more problems if you have complex custom hardware and lose many (or all) of the original engineers from the project . . . or far worse if there isn't detailed documentation of all that custom hardware (to allow new engineers to move forward with the design -if less efficiently than the original designers).

[fibrewire]

As a follow-up, i finally found the Axlon 48K board. Hopefully it is NOT an upgrade that requires a plate with contacts to be smooshed against the bottom of the motherboard.

 

Crossing fingers that Axlon 48K + Axlon 128K + Bit 3 work together.

[fibrewire]

Bump - manual and better pics of Axlon 48K soon.

[awace]

What board are you thinking of putting those chips onto? The most common board, of course, is the Atari 16K. My original article describes how to modify it for 64K x 1 DRAMs (ignore the 400-specific parts). You could make a 32K or 48K board out of that with minimal additional circuitry.

 

The 64K x 4-bit DRAMs (4464) are partly pin-compatible with the 64K x 1 chips (4164) but they require more than 7 bits refresh address, which is all the older ANTIC in the 800 provides. So it would be a complex mod to use those.

 

The 32K x 8 SRAM is really a good choice. They're not at all compatible with DRAMs so you would not use them on a RAM board, but they are nearly pin-compatible with Atari ROMs. It would be not too hard to add one directly to the 10K OS ROM board. I have a few of those chips from a dead '486 PC and have thought a bit about using them this way. If there's interest I'll post a design.

 

 

 

 

If I send you a 16k board could you make it into a 48k for the 400 or i could send the whole 400 and a 32k board I would like 64k but if i can only use 48k of the 64k that would be ok too as most of the software i have to run on the 400 is 48k

Ill even give a 400 away to anyone who makes the other one into 48k or more..I have some 32k boards and some 16k left from my other 800

[awace]

As promised, the Mosaic 64K board with metallic ATASCII stickers - but no manual! Funny thing is, it looks identical to an Atari 16K board.

 

Yes, I paid $125 for a Mosaic 64K box... Go ahead - rub it in

 

you know you got jipped that is a 16k board put it into the 800 and check it I bet you got the BOX for the 64k pcb and the 16k board they replaced it with think its been 20 years I would bet thats just a 16k board in the box I hope you didnt get gouged on ebay although you probably did whell a 16k board for $125 thats a exepensive box ..

PS if it is 64k does it work and could you mod a 400 to be 64k or a 800 to be 64k cause I have some .atr files i got off german sites i want to convert to REAL floppys and run on my 800 but it needs more than 48k to run some files most run w 48k but some want 384k 360 atari emulation or they lock up.

[awace]

I contacted the guy who sold me the Mosaic card. Hopefully he still has his father's Atari 800 in the closet with the REAL Mosaic 64K card in it!

 

See I told you I didnt even get that far reading and knew you bought a 16k in a new box cause those Atari pcbs are so tell tale and they have the REV F witch in turn might just be worth $125 cause REV F is RARE anyway its a late model!! mabie you have a new rare PCB..!!

[ryanr256]

This was in an 800 I acquired a while ago. Is this a standard 800 memory board? It looks different than an uncased Intec 16K board.

 

Anyone know what the smaller board with switch is for? The 800 will boot with the switch in one position, but not in the other. The board is in the last memory slot in the 800.

 

I've attached pics of what I have.

 

Thanks,

 

-Bob

Edited August 30, 2012 by ryanr256

[BillC]

I think I have one of the Mosaic 64K boards. It came in an 800 I acquired a while ago. This one has a smaller board with a switch attached. The 800 will boot with the switch in one position, but will not boot with it in the other. The board was in the last memory slot in the 800.

 

First, can someone confirm this is a Mosaic board? It looks like the one pictured above.

Secondly, does anyone have any idea what the small board with switch does? My guess is that it disables the 64K board, but since the 800 doesn't boot with it in one position, I don't know for sure.

Third, did any manual or documents turn up for the Mosaic 64K board?

 

I've attached pics of what I have.

 

Thanks,

 

-Bob

Definitely not a Mosaic board, it is an Atari 16KB board that someone has modified. It's not a memory upgrade since the RAM chips are still only 16Kb, the circuit could be for write protection so that the computer can run copy protected cartridge in RAM

[ryanr256]

I think I have one of the Mosaic 64K boards. It came in an 800 I acquired a while ago. This one has a smaller board with a switch attached. The 800 will boot with the switch in one position, but will not boot with it in the other. The board was in the last memory slot in the 800.

 

First, can someone confirm this is a Mosaic board? It looks like the one pictured above.

Secondly, does anyone have any idea what the small board with switch does? My guess is that it disables the 64K board, but since the 800 doesn't boot with it in one position, I don't know for sure.

Third, did any manual or documents turn up for the Mosaic 64K board?

 

I've attached pics of what I have.

 

Thanks,

 

-Bob

Definitely not a Mosaic board, it is an Atari 16KB board that someone has modified. It's not a memory upgrade since the RAM chips are still only 16Kb, the circuit could be for write protection so that the computer can run copy protected cartridge in RAM

 

Yeah, I saw fibrewire's post and got excited. Then, after I posted my questions, I read the rest of the thread and found out I have one of the "rare" Rev. F cards. So I edited my original post so I didn't look like a fool.

 

Now that you bring it up, I bet it's part of "The Pill". That also came with the 800. I'll check it out and see if I can dump a cart and then use the switch to run it.

 

Thanks for the quick reply.

 

-Bob

Edited August 30, 2012 by ryanr256

[Rybags]

You could just try the switch with a Basic cart inserted. Basic should still work in either position and have 8K less free Ram assuming the switch protects the entire 16K.

[awace]

What kind of RAM for the Atari is this? I've never heard of it.

 

eBay Auction -- Item Number: 390284769754

 

ClausB: I'll post when I get home.

 

its a 32k board as i am using one now inside my atari 400 NO modifications needed i just swapped out the boards and viola a 32k atari 400!

ALSO I have done the BEST type mod before on a a400 i got off ebay and it worked i lost the info on what to do but if i get a nother one it will come w instructions the hardest part of the 400 upgrade is the taking it apart..

[ClausB]

From original post:

One of many clever design aspects of the Atari 800 is how it addresses its RAM. At a time when adding RAM to a microcomputer involved arranging jumper blocks or switches in strange patterns laid out in cryptic instructions, Atari managed to make the task consumer-proof. And they did it without today's fancy Plug and Pray hardware and software schemes.

 

According to Vendel & Goldberg, a software guy, Alan Miller, was responsible for the RAM selection design:

[fibrewire]

Thanks for your RAM select info on the 8K/16K boards.

 

The Mosaic boards were pretty unique. There was an auction on eBay one time for the interconnect cable and documentation, but missed out on that one. I'm not 100% on this because information is sparse, as opposed to the Axlon stuff that I physically own, but here's my understanding of how they work. I base this on information from emulators and the Mosaic Superdrive floppy I have.

 

- There can be up to three 64K Mosaic boards installed. This gives up to 144K (192K-48K) extended RAM. One card enables 4 banks, two have 20 banks, three have 36 banks.

- Banks are 4K, visible at $C000-$CFFF.

- RAM is enabled through an address latch (not data). Writes to $FFC0-$FFE3 enable the bank ($FFC0=bank 0, $FFC1=bank 1, etc), writes to $FFXX not in that range disable the memory.

 

I, too, would like to see what they physically look like.

 

The blacked out part says "Designed for inter-board communication in Atari 800" - unfortunately im still on a quest to get one of these boards, and it looks as if inter-board communication was discontinued at some point.

 

I would prefer interconnected boards as opposed to cutting traces on the 800.

[fibrewire]

After running a few tests with the Axlon RamDisk 128K and Axlon RamPower 32K, I noticed a few issues.

 

What works:

* RamPower alone in Slot 1 shows up as 16K

* 16K in Slot 1 and RamDisk in Slot 3 shows up as 16K

 

What doesn't work

* RamPower alone in Slot 2 or 3 - computer won't boot

* RamDisk alone in any slot - computer won't boot

 

Then to top things off, I believe Axon software identifies the OS ROM slot as Slot 1 for a total of 4 expansion slots.

 

But even though the Ramdisk is in the last slot, memtest and ramscan still function correctly.

 

I do have an Axlon RamPower 48K, but how to get it to work in Slot 1 without cutting traces?

 

I"ll get some pics posted of the Axlon 48K board and IC module, along with a scanned manual soon.

[fibrewire]

I finally finished the schematic for the Atari 400 48K board:

 

Was this for the 32K board from the Atari 48K board?

 

Also I am still unsure about the 128K RamDisk making 48K total usable to the system. I put two 16K modules in and the RamDisk and still only get 32K. Am I doing something wrong?

[carmel_andrews]

What i wouldn't mind knowing is the following

 

Why didn't Atari consider plug in memory modules for the XL series (or was the XL series designed as a closed system from the word go)

 

Weird thing is Atari did the reverse when it came to the ST/STE

 

Lastly would it have been possible for Atari to implement an XL/XE like bank switch for the 800/400 and get more memory per bank switch (16k as opposed to 4k)

[ClausB]

The Axlon 128K RamDisk board was designed to go in slot 2 with 16K boards in slots 1 and 3.

 

A small mod to the 32K board would allow it to go in slot 1 with 128K in slot 2, leaving slot 3 free. Let me work on that and post it later.

 

Atari did not produce any bank switched RAM at all until the 130XE, long after Axlon and others made some for the 800 (and even after homebrewers like me did it for the XL).

[Faicuai]

(...)

 

Lastly would it have been possible for Atari to implement an XL/XE like bank switch for the 800/400 and get more memory per bank switch (16k as opposed to 4k)

 

(...)

 

 

I am not sure what the answer would be (what type of HW surgery is required, either at memory-slot or backplane/MoBo levels). That I would really like to know, as well.

 

However, the question on its own suggests why I would approach (carefully) any older banking scheme NOT conforming to XL/XE 16K size/method:

1. Besides any possible performance / throughput consideration, you also have compatibility and ubiquity variables involved, as I cannot recall to what extent AXLON's banking scheme was actually widely exploited and what mainstream applications or games actually use it.
    
   
2. Assuming the risk of repeating what maybe already known here, the solution to your question does exist, and it is called INCOGNITO, which will handle XL-type banking, and much, much more. It leaves slots 1 and 2 dormant / passive (albeit ZERO physical changes), but slot #3 supposedly remains fully operational. Of course, Incognito does not necessarily fit in the historical context of this particular discussion, but it is still a very elegant solution that somewhat integrates the past, present and future of this platform.
   

[fibrewire]

The Axlon 128K RamDisk board was designed to go in slot 2 with 16K boards in slots 1 and 3.

 

A small mod to the 32K board would allow it to go in slot 1 with 128K in slot 2, leaving slot 3 free. Let me work on that and post it later.

 

Atari did not produce any bank switched RAM at all until the 130XE, long after Axlon and others made some for the 800 (and even after homebrewers like me did it for the XL).

 

Put 128K in slot 2 and 16K in slot 1 & 3 and success! 128K board functions exactly like it should.

 

looking forward to 32k mod.

 

Thank you Claus!

Edited July 30, 2013 by fibrewire

[ClausB]

To modify the Axlon 32K board for slot 1 (with Axlon 128K in slot 2):

 

- Cut trace between edge pin M and edge pin 18, while leaving 18 connected to the circuitry.

 

- Connect edge pin M to edge pin S. (That is, connect traces that go to M and S, so that you don't get solder on the edge pins themselves.)

 

EDIT: Hey, wait, I answered this two years ago, with pictures and all:

http://atariage.com/forums/topic/181770-160k-the-hard-way/?do=findComment&comment=2289324

Edited July 30, 2013 by ClausB

[ClausB]

Lastly would it have been possible for Atari to implement an XL/XE like bank switch for the 800/400 and get more memory per bank switch (16k as opposed to 4k)

 

I am not sure what the answer would be (what type of HW surgery is required, either at memory-slot or backplane/MoBo levels). That I would really like to know, as well.

 

The Axlon 128K RAMdisk did use the same 16K bank region as the XE, just the bank register was different. I suppose you could modify the Axlon by removing its register chip and wiring it directly to some PORTB bits so it acts more like an XE, but what's the point? I think most RAMdisk software already supports the Axlon, no?

[fibrewire]

It's been a hectic past two years, thank you again Claus!