1090XL remake

[+mytek]

21 hours ago, reifsnyderb said:

I received the 320k SRAM upgrade card for the 1090/1091 yesterday.  So I took the chips off of the prototype and moved them to the new board.  It's a success!

It looks beautiful 👍

[reifsnyderb]

@Bee sent me a 130XE bad memory with the idea that I figure out how to get one of my external memory boards working with it.  This idea is fraught with challenges.

 

1.  For some reason, the 130XE does not have an EXTENB pin on the ECI.  However, it does have the "A" pin reserved and unused.  Why EXTENB was not added, I'll never know.

2.  The current chip shortage still is a major problem as CPLD's are not readily available.

 

So, after exchanging some messages, we agreed to see if I could get his 130XE working with a 1090 board.

 

The theory is that because of the omitted MMU, PORTB needs to be shadowed, 3 PORTB settings need to be back fed to an adapter board, and an external MMU needs to be made.

So, just the other day I received some 16V8B PAL chips and tried programming one for an MMU.  Well, it didn't work and I'll need to troubleshoot it.  

 

There is another way to get full functionality, though.  It requires soldering pin 16 of the MMU to the "A" pin on the ECI.  I setup the adapter board so this could be done as well.  Since I was unsure as to whether there were more problems with the 130XE or whether it's something I am trying I figured I'd try adding the CI (EXTENB) connection from the MMU to the ECI.  It works!  So there is a way to run a 1090 with a 130 XE.   🙂

 

So, here's a few pictures:

 

The 130XE as booted up without the 1090 running.

 

 

The 130XE connected to the 1090.  (I realized after taking the picture that I had some shipping information in the picture and redacted it.)

 

 

The adapter board:

Note the socket for the external MMU as well as the accomodation for 2 cartridges.

 

 

The temporary wire connecting the MMU to pin "A".

 

 

The SimCheck screen of the 130XE running.  Note that there is no ANTIC banking as my card doesn't handle ANTIC banking.  This 1090 board and adapter board, however, both have the capability of handling the /HALT line and could do ANTIC banking with the right board.  (This 1090 board is slightly "upgraded" so as to be able to handle a /HALT signal.)

 

 

Now that I got it working I can start troubleshooting the system to get a 1090 working with an XE without soldering that extra wire.

 

 

 

 

Edited October 9, 2022 by reifsnyderb

[kenames99]

hi reifsnyderb,

  well done! I gotta say, you did a great job in this project. the logic to generate the CI signal in the 1090 would not be difficult. if you do not want to do it let me know and I will present a circuit. of course it will likely involve a 16v8. the 1090XL is supposed to present an "ECI" interface anyway which means this needs to be done. ECI = Enhanced Cartridge Interface. it is in the specs released by Atari.

 

again, great job!

 

Ken

 

[reifsnyderb]

42 minutes ago, kenames99 said:

hi reifsnyderb,

  well done! I gotta say, you did a great job in this project. the logic to generate the CI signal in the 1090 would not be difficult. if you do not want to do it let me know and I will present a circuit. of course it will likely involve a 16v8. the 1090XL is supposed to present an "ECI" interface anyway which means this needs to be done. ECI = Enhanced Cartridge Interface. it is in the specs released by Atari.

 

again, great job!

 

Ken

 

Hello,

 

Thanks!

 

It's possible I am approaching the generation of the CI/EXTENB signal from the wrong direction.  Since the 1090 buffers the EXTENB signal, I figure the signal has to come from "outside" the 1090, so to speak.

 

My current approach is to use a 1090 card (the shadow board) that has 3 flip-flops on it to shadow and store the following PORTB settings:  PB0, PB1, and PB7.  Those signals are then "back feed" to the external MMU (the socket is shown in my pictures), via 3 "available" lines, so that EXTENB can be generated from outside of the 1090.  Since I've got the 130XE working, I plan to use a meter or scope (if I have to) to first determine if the shadow board is working right.  I am unsure as to whether the problem is the shadow board, the external MMU (the 16v8), or something else yet.

 

I'd like to see your idea as to how to get this working as maybe my approach is wrong.

 

Thanks!

 

Brian

 

 

  

[reifsnyderb]

I was able to test out the 1090 board that shadows $D301 today.  Bits 0,1, and 7 of $D301 are appearing at the external MMU socket on the 130XE PBI adapter.  So the problem must be with the programming of the 16V8B.  I am not too surprised as the jed file looked rather suspicious.  But, either way, I needed to isolate the problem.

[Mathy]

Hello Brian

 

The XE might not have EXTENB, but it does have a signal called $D1xx.  I've been told it can be used instead of EXTENB.

 

Sincerely

 

Mathy

 

[reifsnyderb]

19 minutes ago, Mathy said:

Hello Brian

 

The XE might not have EXTENB, but it does have a signal called $D1xx.  I've been told it can be used instead of EXTENB.

 

Sincerely

 

Mathy

 

Hello,

 

The $D1xx signal comes from a 74LS138 and is only a select signal for addresses within $D1xx.  This signal would be great for decoding any of the $D1xx PIA addresses and should have been present in the XL's so as to make decoding a lot easier.  EXTENB is a completely different signal and comes from the MMU.

 

There is a lot of scattered, bad, and conflicting information out there regarding the PBI/ECI.  For example, I've had bad boards made after reading that EXTENB enabled the PBI device as per Earl Rice's articles only to find out it enables only certain memory locations.  Because of this, I've been putting together a document that brings all of this information together.  It's a work in progress, but here is the current revision:

 

TCGTHAPB.pdf

 

I am planning to add the example assembly language driver code I've found to the document as well...with the caveat that I haven't tested nor confirmed it.  This way anyone developing for the PBI doesn't have to compile everything that is scattered all over the place.  Some of the information I've even had to pull from the atarimuseum website via the wayback machine website.

 

Best Regards,

 

Brian

 

[reifsnyderb]

Making progress.  I removed the wire from the MMU to the ECI.  The ATF16V8B programming has been mostly figured out and the 130XE is now running off of the RAM on the 1090 without any internal modifications.  I just have a little bit more to figure out and it will be possible to upgrade or replace a 130XE's memory with memory on a 1090 without any internal modifications to the 130XE.

 

I've also been thinking more about how to make this work better in the future.  I propose that the 1090 be modified by adding a jumper (or even a DIP switch) on the EXTENB line between the transceiver and card slots.  With an XL you would have the jumper installed as the XL has EXTENB on it's PBI.  On a 130XE, you would remove the jumper and add an XE compatibility card.  This card would only be needed if you were using other cards that needed EXTENB and would be optional.  (Most likely, this would only be memory cards, anyhow.)  

 

This XE compatibility card would be like an improved version of the shadow board I am using.  It would shadow PORTB bits 0, 1, and 7.  It would also have an ATF16V8B (or other CPLD) that would re-create the EXTENB signal and put it on the 1090's card slot bus.  With the proposed jumper, on the 1090 board, removed the EXTENB signal would be isolated to just the 1090 cards.

 

The advantage of adding this jumper to the 1090 board would be that an ATF16V8B (or other CPLD) would no longer need to be added to the 130XE's PBI extender board.  It would now be on a single board in the card slot.  Also, the 3 reserved lines would no longer have to be used to back-feed the shadowed PORTB bits from a board in the 1090 to the XE's PBI extender board.  In short, it would be a lot cleaner of a system with only a single modification to the 1090 board.  Given that Atari made minor revisions to their boards over time I don't think this would be an unacceptable revision.  Also, I sort of ran out of lines for the current extender board and add to add a jumper to it to de-select the +5VDC from the computer to the PBI card edge connector so I could use the power line as an additional signal line.  (I did put a resistor on that line so as to protect the chip in the event I screwed up and applied power to that line.)

 

Best Regards,

 

Brian

 

Edited October 10, 2022 by reifsnyderb

[reifsnyderb]

The new plan for Atari XE EXTENB compatibility is to replace one of the transceivers on the 1090 board with an ATF16V8B programmable logic chip.  One pin would be programmed as a select pin for EXTENB while the other pins would just pass the signals.  A reserved line on the 1090 bus would be connected to the select pin for EXTENB.  If you want to use a memory card, an XE compatibility card would be able to be plugged in and create it's own EXTENB line on the 1090 bus while shutting down only EXTENB on the transceiver (ATF16V8B).  This would allow any 1090 memory card to be used with an XE and would be completely plug and play.

 

 

 

  

[reifsnyderb]

I figured out the last bit of the external MMU chip and this 1090 configuration can completely replace and/or upgrade the memory of an Atari 130XE.  🙂

 

[reifsnyderb]

If an ATF22V10C is used instead of a PAL20L10 on an 80 column 1090 board, here is the WinCupl code:

 

Name     1090 80 Column Board U4;
PartNo   ATF22V10C ;
Date     10/28/2022 ;
Revision 01 ;
Designer Unknown ;
Company  Atari ;
Assembly None ;
Location U4 ;
Device   g22v10 ;

 

/* Re-created by Brian E. Reifsnyder from JED file downloaded from atarimuseum.com */

 

/* *************** INPUT PINS *********************/
PIN   1 = A8                      ; /* Address Lines                   */ 
PIN   2 = A9                      ; /*                                        */ 
PIN   3 = A10                     ; /*                                       */ 
PIN   4 = A11                     ; /*                                       */ 
PIN   5 = A12                     ; /*                                       */ 
PIN   6 = A13                     ; /*                                       */ 
PIN   7 = A14                     ; /*                                       */ 
PIN   8 = A15                     ; /*                                       */ 
PIN   9 = CARDSEL             ; /* Card Select  (Always set low.)  */ 
PIN  10 = xxFx                   ; /*                                 */ 
PIN  11 = xxxF                   ; /*                                 */ 
PIN  13 = MNR                   ; /*                                 */ 
PIN  15 = RW                     ; /* Buffered Read/Write             */ 
PIN  16 = Phi2                    ; /* Buffered Phi2 Clock             */ 
PIN  21 = DSEL                   ; /* Device Selected (if high)       */ 
/* *************** OUTPUT PINS *********************/
PIN  14 = MPD                     ; /* Math Pack Disable               */ 
PIN  17 = LDBUFFSEL               ; /* Data Buffer Select Logic        */ 
PIN  18 = DBUFFSEL                ; /* Data Buffer Select              */ 
PIN  19 = EXTSEL                  ; /*                                 */ 
PIN  20 = D1Fx                    ; /*                                 */ 
PIN  22 = D1FFw                   ; /* D1FF Write                      */ 
PIN  23 = D1FFr                   ; /* D1FF Read                       */ 

/* Equations:  */

!MPD = DSEL;
MPD.oe = DSEL;

!LDBUFFSEL = EXTSEL #
       Phi2 & D1Fx #
       !MNR;
LDBUFFSEL.oe = !CARDSEL;     /*  Originally, was set to VCC.  As !CARDSEL is always low, this works. */

!DBUFFSEL = !LDBUFFSEL;
DBUFFSEL.oe = !LDBUFFSEL;

!EXTSEL = A11 & A12 & !A13 & A14 & A15 & !CARDSEL & DSEL;    /* EXTSEL if $D8xx through $DFFF & DSEL */
EXTSEL.oe = !CARDSEL;        /* See note for LDBUFFSEL.oe   */

!D1Fx = A8 & !A9 & !A10 & !A11 & A12 & !A13 & A14 & A15 & !CARDSEL & !xxFx;
D1Fx.oe = !CARDSEL;          /* See note for LDBUFFSEL.oe   */

!D1FFw = A8 & !A9 & !A10 & !A11 & A12 & !A13 & A14 & A15 & !CARDSEL & !xxFx & !xxxF & !RW & Phi2;
D1FFw.oe = !CARDSEL;         /* See note for LDBUFFSEL.oe   */

!D1FFr = A8 & !A9 & !A10 & !A11 & A12 & !A13 & A14 & A15 & !CARDSEL & !xxFx & !xxxF & RW & Phi2;
D1FFr.oe = A8 & !A9 & !A10 & !A11 & A12 & !A13 & A14 & A15 & !CARDSEL & !xxFx & !xxxF & RW & Phi2;
 

 

 

 

File is attached here:

 

 

1090 80 COLUMN BOARD.PLD

 

 

 

Edited October 29, 2022 by reifsnyderb

[reifsnyderb]

It looks like I posted too soon.  I've got an 80 column board I am trying to bring to life and the previous code I just posted doesn't work...even though the logic looks good.  Obviously, I missed something.  @kenames99 sent me a .jed file that works.  I am planning to take a look at them, compare, and figure out what went wrong.  

 

I still have another problem to workout on the board.  But it's close to working.

[+kheller2]

Anyone want to take on the Floppy card?  Schematics are available which is interesting given the known 1090s don't have a place to stick a mech.

 

 

[reifsnyderb]

5 minutes ago, kheller2 said:

Anyone want to take on the Floppy card?  Schematics are available which is interesting given the known 1090s don't have a place to stick a mech.

You do!   😎

 

I think something better would be an SD Card reader that works something like an S-Drive....but for the 1090.

[reifsnyderb]

Progress is being made on a hybrid 80 column card that uses as many newer, still manufactured, electronics as possible.

 

This card is around 60% the size of the original.  It is a 2 layer board.  While it could have been smaller if it was a 4 layer board, I kept with a 2 layer board due to the $35 engineering fee on a 4 layer board, with any side greater then 100mm, as I didn't see any way to make the board smaller than 100mm.  Some current issues to fix on a newer board are to fix the diode directions on the silk screen, add bigger pads for the SMD diodes, add a bigger heat sink area for the voltage regulator, fix a screw-up on the pads for the crystal, fix a screw-up on the connections to the composite jack, and add ground planes in the empty areas of the board.  All of those connections have been hacked by cutting traces and adding jumpers on this board.  The jumper, for the crystal, is evident on the bottom right of the board in the attached picture.

 

@kenames99 has suggested on thing to try is to replace a silicon diode with a germanium diode.  I'll know once I get them in.  The card is quite close to working.  It's quite possible the current diode is on the "hairy edge" of working and that is why there is the streaking on the screen.  If I type in a few characters, all synchronization is lost.

 

 

Edited October 30, 2022 by reifsnyderb

[reifsnyderb]

Admittedly, my previous success in getting the 130XE to work with a 1090 board was a hack.  So, the next step is to create a 1090 board that works as natively as possible with both XL's and XE's.  Show below are pictures of the new board, the old board, and both together.  The new board is shown with only the +5VDC circuitry as there is only one board I could find, the serial parallel board, that would have used the 12VDC circuitry.  For most people it would be an unnecessary expense.  Also, I've noticed that the 8 diodes on the data lines are shown as being omitted in pictures of an actual 1090 board so I omitted them as well.  (While I am not great with electronics, I could never figure out why Atari had those diodes unless is for some sort of static discharge protection???)  Just right of top center is a footprint for a connector that could be installed so as to put the power switch and LED at another location.  (i.e.  The front of the case?)  The connector used is keeping with the timeframe of when the 1090 was developed. 

 

The biggest difference, however, is that one of the transceivers have been replaced with an ATF16V8B programmable logic chip.  This is to obtain compatibility with the 130XE (and 800XE) for memory expansions.  Originally, with a transceiver in this position, the /EXTENB signal would be unconnected and therefore floating when connected to a 130XE.  This would cause memory expansions to fail.  So, there are 2 options to get a working memory expansion with a 130XE and the 1090.  The first is to solder a jumper wire from the MMU to pin A on the ECI.  This is the cheap but more technical way to solve the problem.  With this board, a compatibility card will be able to be installed that sends a signal to the ATF16V8B so as to isolate the /EXTENB on the 1090 bus and allow the compatibility card to recreate the /EXTENB signal without interference from the transceiver.  All that be needed is the compatibility card and the XE ECI to PBI adapter board.  Basically, it will be plug and play like Atari liked to do things as is evidenced by their other technologies such as the SIO as PIA specifications.  Also, given that Atari's MMU appears to be some sort of programmable logic chip, it would not be unreasonable to use a similar technology in the 1090 so as to be compatible with the 130XE and 800XE.

 

The new board works fine with my 320k memory module and I hope to have a new compatibility card soon.

 

 

 

 

 

 

 

[kenames99]

>Admittedly, my previous success in getting the 130XE to work with a 1090 board was a hack.  So, the next step is to create a 1090 board that works as natively as possible with both XL's and XE's. 

>Show below are pictures of the new board, the old board, and both together.  The new board is shown with only the +5VDC

 

How so? an extender board (eci to pbi) does that along with your compatibility board. And I do not really see a need for anything but the unregulated +10v supply. each board supplies it's own +5V regulator anyway. the supplied =5v from the 1090XL main board is supposed to be a reference only.

 

>circuitry as there is only one board I could find, the serial parallel board, that would have used the 12VDC circuitry.  For most people it would be an unnecessary expense.  Also, I've noticed that the 8 diodes on the data lines are shown as being omitted in pictures of an actual 1090 board so I omitted them as well.

 

That's quite OK, they are only clamping diodes to protect from reverse spikes but the resistors do that (to an extent).

 

>Just right of top center is a footprint for a connector that could be installed so as to put the power switch and LED at another location.  (i.e.  The front of the case?)  The connector used is keeping with the timeframe of when the 1090 was developed. 

 

yea, I see that. kinda uncertain about that since there is already a switch and hole for it in back of the case. *shrug*

 

>The biggest difference, however, is that one of the transceivers have been replaced with an ATF16V8B programmable logic chip.  This is to obtain compatibility with the 130XE (and 800XE) for memory expansions.  Originally, with a transceiver in this position, the /EXTENB signal would be unconnected and therefore floating

>when connected to a 130XE.  This would cause memory expansions to fail.  So, there are 2 options to get a working memory expansion with a 130XE and the 1090.  The first is to solder a jumper wire from the MMU to pin A on the ECI.  This is the cheap but more technical way to solve the problem.  With this board, a

>compatibility card will be able to be installed that sends a signal to the ATF16V8B so as to isolate the /EXTENB on the 1090 bus and allow the compatibility card to recreate the /EXTENB signal without interference from the transceiver.  All that be needed is the compatibility card and the XE ECI to PBI adapter board. 

>Basically, it will be plug and play like Atari liked to do things as is evidenced by their other technologies such as the SIO as PIA specifications.  Also, given that Atari's MMU appears to be some sort of programmable logic chip, it would not be unreasonable to use a similar technology in the 1090 so as to be compatible

>with the 130XE and 800XE.

 

Yea this all sounds good. thought we talked earlier about it but if not, sounds good.

My next thought is maybe an updated main board with these fixes in place, a smaller/less expensive pc board, and compatible with xl and xe out of the box. I also have an idea for 400 and 800 to use this also.

 

if I think of anything else, I will post again.

 

Ken

 

[reifsnyderb]

5 minutes ago, kenames99 said:

>Admittedly, my previous success in getting the 130XE to work with a 1090 board was a hack.  So, the next step is to create a 1090 board that works as natively as possible with both XL's and XE's. 

>Show below are pictures of the new board, the old board, and both together.  The new board is shown with only the +5VDC

 

How so? an extender board (eci to pbi) does that along with your compatibility board. And I do not really see a need for anything but the unregulated +10v supply. each board supplies it's own +5V regulator anyway. the supplied =5v from the 1090XL main board is supposed to be a reference only.

Sometimes I forget to include things while typing.  This board is what I think of as the "next step".  I should have said the "next step was to create a 1090 board...."

 

My mistake on the voltages as the +5vdc is for the transceivers and programable logic device.  I forgot to mention that the unregulated +10vdc goes to each board as is usual.  This 1090 board is 100% compatible with all 1090 boards.

 

There is a slight bus difference, on the 1090 board, in that I had to use the RDY line for the XE compatibility card line.  But a search of all 1090 blueprints showed that the RDY line was never used anyhow.  So it won't affect anything.  I already had to use the 1090's reserved lines for /HALT, RD4, and RD5.  (RD4 and RD5 are needed for the MMU on the XE compatibility card.)

 

If somebody were to want the +12vdc and -12vdc circuitry, everything is already in place to add it.  Most of the component values are silkscreened on the board.

 

5 minutes ago, kenames99 said:

 

>circuitry as there is only one board I could find, the serial parallel board, that would have used the 12VDC circuitry.  For most people it would be an unnecessary expense.  Also, I've noticed that the 8 diodes on the data lines are shown as being omitted in pictures of an actual 1090 board so I omitted them as well.

 

That's quite OK, they are only clamping diodes to protect from reverse spikes but the resistors do that (to an extent).

 

>Just right of top center is a footprint for a connector that could be installed so as to put the power switch and LED at another location.  (i.e.  The front of the case?)  The connector used is keeping with the timeframe of when the 1090 was developed. 

 

yea, I see that. kinda uncertain about that since there is already a switch and hole for it in back of the case. *shrug*

If you think that the extra connector should be removed, it's no big deal to me.

 

5 minutes ago, kenames99 said:

 

>The biggest difference, however, is that one of the transceivers have been replaced with an ATF16V8B programmable logic chip.  This is to obtain compatibility with the 130XE (and 800XE) for memory expansions.  Originally, with a transceiver in this position, the /EXTENB signal would be unconnected and therefore floating

>when connected to a 130XE.  This would cause memory expansions to fail.  So, there are 2 options to get a working memory expansion with a 130XE and the 1090.  The first is to solder a jumper wire from the MMU to pin A on the ECI.  This is the cheap but more technical way to solve the problem.  With this board, a

>compatibility card will be able to be installed that sends a signal to the ATF16V8B so as to isolate the /EXTENB on the 1090 bus and allow the compatibility card to recreate the /EXTENB signal without interference from the transceiver.  All that be needed is the compatibility card and the XE ECI to PBI adapter board. 

>Basically, it will be plug and play like Atari liked to do things as is evidenced by their other technologies such as the SIO as PIA specifications.  Also, given that Atari's MMU appears to be some sort of programmable logic chip, it would not be unreasonable to use a similar technology in the 1090 so as to be compatible

>with the 130XE and 800XE.

 

Yea this all sounds good. thought we talked earlier about it but if not, sounds good.

Yeah, we did discuss this before.  It just took me a while to get the board made for testing.

5 minutes ago, kenames99 said:

My next thought is maybe an updated main board with these fixes in place, a smaller/less expensive pc board, and compatible with xl and xe out of the box. I also have an idea for 400 and 800 to use this also.

The only way I can think of to be compatible with the XL and XE "out of the box" is to put the XE compatibility circuitry on the 1090 board itself.  My thought is that it would be an unnecessary expense for those with XL's and those with XE's that don't mind soldering the extra jumper in place.

 

I am interested in seeing what you have in mind for a 400 or 800 as I don't know of a good way to add some sort of parallel interface to them.

 

Best Regards,

 

Brian

 

[+kheller2]

One of the original concepts/thoughts was that an expansion box was SIO compatible with the 400/800.  I'm thinking maybe some sort of SIO card but in hindsight, I'm thinking Atari just was going to go the cheap route and make it just another, yet smart, SIO device.

[+ZuluGula]

12 hours ago, reifsnyderb said:

I am interested in seeing what you have in mind for a 400 or 800 as I don't know of a good way to add some sort of parallel interface to them.

 

[reifsnyderb]

5 hours ago, kheller2 said:

One of the original concepts/thoughts was that an expansion box was SIO compatible with the 400/800.  I'm thinking maybe some sort of SIO card but in hindsight, I'm thinking Atari just was going to go the cheap route and make it just another, yet smart, SIO device.

It might be possible to make an SIO card that would fit in a 1090 and be able to interface some, but not all, cards with a 400/800.  It think it would be possible to make cards to daisy chain a couple 1090's as well.

[reifsnyderb]

5 hours ago, ZuluGula said:

 

Those card edge connectors are, unfortunately, pretty inaccessible.  My thought is that the SIO port connection would be the most feasible....but speed would be sacrificed.

[reifsnyderb]

The new 1090 XE Compatibility Board came in today.  With the exception of 1 screw-up on my part, it all works.  To make a long story short, I needed 2 extra macro-cells and had to move the /MPD input pin on the special MMU (specially programmed ATF16V8B) chip.  This has been fixed so as to be part of an order for new boards.

 

This board really is now only needed if an Atari 800XE or 130XE is connected to this XE compatible 1090 board, does not have the MMU CI pin jumpered internally to the ECI Pin A, and is using a 1090 memory card.

 

 

 

Here is the 130XE connected to the 1090 board via the 1090 adapter board.  Installed in the 1090 board is a 320k memory board and the 1090 XE Compatibility Board.

 

 

Closeup of the XE compatible 1090 board with the 320k RAM card (back) and compatibility card (front).

 

 

Below is the old hack I used, on the left, and the new design, on the right.  The old design requires that an external, special, MMU be installed on the adapter board (top left).  This makes the adapter board more expensive and complicated.  The old compatibility board (bottom left) only has the chips needed to decode PORTB and store bits 0,1, and 7 so as to provide an /EXTENB signal.  

 

The new adapters boards are on the right.  The Adapter board (top right) is simpler and the compatibility board (bottom right) has 2 ATF16V8B chips that are programmed as an special MMU and as a PORTB decoder.  

 

I think that removing the chip and socket, from the adapter board, makes for a cleaner adapter.

 

 

Finally, here's the confirmation that the 320k board is being seen.  Memory tests are successful.  It would be possible, of course, to design a 1090 card with ANTIC banking as well.  This XE compatible 1090 board supports the /HALT signal and would also allow for an XL to have the /HALT signal added to the PBI.

 

 

 

Edited December 20, 2022 by reifsnyderb

[reifsnyderb]

Here's the latest copy of "The Comprehensive Guide To The Atari Parallel Bus"

 

TCGTHAPB.pdf

[+Allan]

Fantastic work on this!

 

Question:

 

Since the Compatibility board only adds the ability to use 1090 memory expansion boards with the 130XE and 800XE (please correct me if I am wrong here) and it takes up an expansion slot, would it be possible to add some additional functionality to this board?