1090XL remake

[+CharlieChaplin]

On 8/10/2023 at 5:10 PM, reifsnyderb said:

Update on the 4MB 1090 card.

 

Well, the good news is that I got most of it working.  I say most of it because there is still some debugging required where the optional main memory is concerned.  (I made the main memory optional so as to allow the card to be used stand-alone or with another memory card.)

 

Now the bad news....

I couldn't get the card to bank on the Axlon register of $CFFF.  I couldn't understand why it wouldn't work when all the signals looked fine on the schematic.  I didn't put a scope on it, though, as it can be really hard (as in almost impossible) to catch the right signal at the right time.  After some thought, and lots of testing, I started to wonder if $CFFF would never work on an XL as the MMU would also be enabling the ROM and there could be bus contention.  So, as a test, I changed the decoder to decode the address to $D1FE and it works.  I can use $D1FE to bank the memory and have 4MB available.  ($D1FE is actually reserved, by Atari, as a banking address for 1090 cards.)  However, there is nothing that uses $D1FE to bank the memory as the 1090 was never released.  Also, using $D1FE to bank, like the Axlon register banks, breaks Atari's banking design for $D1FE.  (The 1090 1066 card uses $D1FE for banking.)

 

So, in summary, 4MB can be available via the 1090 but using the Axlon register is a no-go.   😞

 

 

Edit to add:  Writing to a ROM address is possible, on an 800 (i.e.  Mosaic banking), so I figured this would work, too.

 

 

Well,

I wonder how @tmp's Sub Cart does it then...

It connects to the PBI or ECI and makes 1MB port-B RAM or 4MB Axlon RAM available with XL-OS Rev. 2 or 3 in use.

 

There is a 4MB Axlon Ramtester and it worked fine, even with XL-OS Rev. 2.

 

TstAxlon.zip

[reifsnyderb]

10 minutes ago, CharlieChaplin said:

 

Well,

I wonder how @tmp's Sub Cart does it then...

It connects to the PBI or ECI and makes 1MB port-B RAM or 4MB Axlon RAM available with XL-OS Rev. 2 or 3 in use.

 

There is a 4MB Axlon Ramtester and it worked fine, even with XL-OS Rev. 2.

 

TstAxlon.zip 33.8 kB · 0 downloads

My understanding is that this is done by holding the Refresh line low on the PBI.  Due to the design of the MMU, holding Refresh low will basically disable the MMU under many situations.  (Take a look at my PBI document as I have the MMU code in the Appendix.)  However, ANTIC normally controls the Refresh line.  On the PBI, the Refresh line is supposed to be an output only and holding the Refresh low, from the PBI, will conflict with ANTIC.  Supposedly, ANTIC works fine with this even though it is basically being shorted out and was never supposed to be used this way.  My objection to controlling the Refresh line, from the PBI, is that I do not like the idea of shorting out a 40 year old chip.  The 1090XL board also buffers the Refresh line so there really isn't a way to short out the Refresh line without re-working the 1090XL board.

 

Atari designated $D1FE for it's memory board and memory banking on the 1090XL.  Banking via a $D1xx address avoids any conflicts with the function of the MMU.

[_The Doctor__]

In previous discussion I suggest the inclusion of a resistor to keep it from being a dead short and limit current keeping it in the correct range.

[reifsnyderb]

21 minutes ago, _The Doctor__ said:

In previous discussion I suggest the inclusion of a resistor to keep it from being a dead short and limit current keeping it in the correct range.

I've tried something like that once and only managed to change the logic level so that it was neither high nor low.  Admittedly, I didn't try too many resistances.  I didn't want the resistance too low, either.  Maybe it would work with more experimentation.  I still don't care for the idea of doing that to an ANTIC chip.

Edited August 22, 2023 by reifsnyderb

[kenames99]

On 8/15/2023 at 7:24 AM, reifsnyderb said:

I just looked up the MIO board.  From the manual, it has:

 

"a parallel printer interface, a serial interface for printer or MODEM, a printer buffer, a bootable RAMDISK, and an SCSI/SASI interface for hard disk drives"

 

A bootable RAM disk and a printer buffer can all be done with memory.  Does anyone still use the old parallel printer interfaces (db25?), serial interfaces, and/or SCSI interfaces?

I do still use scsi and printer from that era. somtimes you just gotta.

 

Ken

 

[ijor]

On 8/21/2023 at 10:41 PM, reifsnyderb said:

My objection to controlling the Refresh line, from the PBI, is that I do not like the idea of shorting out a 40 year old chip.

It's even worse for a new ANTIC replacement chip. Old NMOS chips might tolerate an output being externally forced to ground, a CMOS part might not at all.

[_The Doctor__]

current limit and take down to the common denominator logic low value for all parts involved.

[reifsnyderb]

I think the safest solution to using the refresh line as an MMU disable line is to put an open collector output buffer immediately after the refresh pin on ANTIC.  A 4.7k ohm resistor, after the buffer, would be needed to hold the refresh line high.  Of course, that defeats the purpose of just plugging a device into the PBI without modifying the computer.

Edited September 11, 2023 by reifsnyderb

[reifsnyderb]

A couple months ago I started digging into where the RAM was supposed to be supplied from between $D600-$D7xx for PBI use.  On a couple Atari documents, Atari divided this region up into 64 byte sections for each board but didn't specify where the memory came from.  Since I found it hard to believe that Atari expected every single PBI card to have it's own 64 bytes of memory I started suspecting that maybe the computer was to supply this memory.  Of course, most XL's and all XE's do not supply this memory.  In the latest edition of the Atari Hardware Reference Manual, Avery writes that the 1400XL supplied this memory but didn't divide the memory up into 64 byte per PBI device sections as per Atari's documentation.

 

As more of an experiment than anything I decided to add an option to add that memory back from $D600-$D7xx.  This card 320k card has a solder jumper to enable the RAM from $D600-$D7xx.  While this would be fine for any future 109xXL devices, it is incompatible with some internal expansions.  With all the chips on this card, I am thinking about moving some of them onto a PLC.

 

[+Stephen]

4 hours ago, reifsnyderb said:

A couple months ago

Quoting because that's some of the nicest SMD soldering I have seen!  I assume that isn't done by hand?

[reifsnyderb]

1 minute ago, Stephen said:

Quoting because that's some of the nicest SMD soldering I have seen!  I assume that isn't done by hand?

Thanks!  I do it by hand with a soldering iron and hot air soldering tool.  It took a while to come up with a process that seems to work.  First I take a tooth pick and apply the solder paste to the pads.  Next, a soldering iron is used to melt the paste and tin all the pads.  Excess solder is just shook off the soldering iron.  Then, I lay the chips on the pads and use the hot air to solder them in place.  Sometimes the chips want to float on the solder and I have to push them down with a dental pick.  Afterwards, it's not uncommon to have to hunt down a bad solder joint, with a magnifying glass, and re-flow it with the soldering iron.  On some occasions, I have to add more solder.  Afterwards, the board is usually quite a mess with all the flux.  So, I have a tupperware container with 91% alcohol and soak the board before scrubbing it with a tooth brush.

 

In that picture, after I looked at the picture on the computer I noticed that near R80 there's a little dirt.  It's residue from the alcohol bath and can be removed with a fingernail.  I figured I'd just use the picture anyhow.  🙂

 

[reifsnyderb]

On 2/21/2023 at 10:46 PM, reifsnyderb said:

1090 Compy-Shop compatible 192k card...

 

I decided to add the /halt line to the PBI on my 600XL.  Afterwards, I installed the 192k Compy-Shop compatible test card and it is somewhat working.  I say somewhat because Simcheck 1.5 shows the memory properly.  SimTest 0.8 runs the memory tests on all banks and doesn't fine any errors.  SpartaDOS X doesn't load the drivers and displays:

 

Config error (132):  DEVICE SPARTA

Config error (132):  DEVICE SIO

Config error (132):  DEVICE ATARIDOS

Config error (132):  DEVICE SIDE3

 

Base RAM is working fine.

 

The SIDE3 cartridge loads fine, displays programs, and runs programs.

 

 

I am wondering if SpartaDOS X doesn't like having 192k Compy-Shop cards....

(I never noticed a problem on the 130XE.)

 

Edit to add:  I just ran XRAM Test 0.22.0 and it's not showing bank A7...which is really odd.

 

 

 

 

Tonight, I re-visited this card and discovered that U47, pin 3, was connected to ground as opposed to +5vdc.  This would mess up XE banking!  So, I put the chips back on the Compy-Shop compatible card (with ANTIC banking), hacked in the fix, and it worked!  I can now use the 1090 card to, essentially, turn my 600XL into a 130XE.  🙂   I don't know how many times I checked that logic and didn't find it.  This evening, I found it in under 5 minutes then spent at least 15 minutes studying the logic to ensure that it was truly wrong.

 

I don't have /halt latched, though.  Maybe I need to do this as the 800 latches /halt for when ANTIC is handling the memory.  The memory tests work and SDX loads.

 

After a brief test, I then re-enabled the extra bank, on PB-6, and the memory tests worked.  However, SDX fails in this case.  I suspect that since Compy-Shop also used pin 7 that this could be causing a problem.  But, I am not certain.  Presently, I don't know what to test with XE banking to see if it works and to see if I really need to latch /halt or not.  I don't know if the Sally chip latches /halt or if it just acts on /halt.  Either way, this is promising as it's one step closer to a working 1090 card with ANTIC banking.   🙂

 

Edit to add:  I just tried Video Blitz and it doesn't run like the video. (https://www.youtube.com/watch?v=gfo3yN3TPDw&ab_channel=boschreurs )   I guess /halt needs latched.  So, this is a step closer anyhow.

 

 

 

Edited October 29, 2023 by reifsnyderb

[Markk]

Hello

I've been following this thread for over a year.
I also had the circuit boards made from the first ones
Layouts that “kenames99” released here. Thank you for that.
Since then I have been following the many prototypes and memory cards etc. from "reifsnyderb" with interest.
Now my question is, will any of this be released to the general public or can you buy it at some point?
It would be a shame if you had to do all the work
comes to nothing.
(Translated from German with google)

greetings Mark

[reifsnyderb]

3 minutes ago, Markk said:

Hello

I've been following this thread for over a year.
I also had the circuit boards made from the first ones
Layouts that “kenames99” released here. Thank you for that.
Since then I have been following the many prototypes and memory cards etc. from "reifsnyderb" with interest.
Now my question is, will any of this be released to the general public or can you buy it at some point?
It would be a shame if you had to do all the work
comes to nothing.
(Translated from German with google)

greetings Mark

I wanted to have boards out a couple months ago and wanted a final "killer application" for everything.  However, the 4MB Axlon card didn't work as planned.  So I decided to turn it into a 4MB RAM Disk card that has a lot of extra features.  That 2 week project turned into 2 months and had some directions that didn't work out.  For example, I was going to add a printer spooler partition, got the code working, and found out that the VBLANK printer spoolers are mostly a carnival trick of sorts.  At the moment, I am at the 3rd revision of the 4MB boards, need to finish assembly of a prototoype board, and migrate the code into the flash ROM.  It's coming along...but a lot slowly than I had hoped.

 

Once that is out, I'll start making boards.  Eventually, I'll release all the sources.

[Markk]

Hello “reifsnyderb”
Thank you for the answer. Of course you should
Everything works before it is made available to the public. I can understand that. I'm looking forward to it, and
Thank you again for your great work on this project.

greetings Mark

[Mathy]

Hello Mark

 

4 hours ago, Markk said:

(Translated from German with google)

 

"Deepl" seems to be better at translating.  BTW you do know the ABBUC website and forum?

 

Sincerely

 

Mathy

 

 

[kenames99]

hi,

  I have spent a few hours working on the 1090XL and the 80 column video card and RealDOS. the reults are a fully working video 80 column output, even in basic. the photo is done with RealDOS ver. 1.0 Build 33, and reifsnyderb's newest OS rom (6.2 early beta). the stock os rom has a bug that sets the 80 column back to 40. even tho this is now complete, that OS bug is a problem for people that do not want to change their hardware. the os bug happens only with some print routines used thruout RealDOS and it's support files but this is still not the way I want thing to work. I will be sorting thru RealDOS and all the support fil;es to fix this as soon as I can. I should also mention that one of the SDX update project's members has a SDX driver for the 80 CVC also. this is also good news to hear. I would also like to thank phaeron for all his work with Altirra and the debugger in it that made my life so much easier.

 

I will be releasing the files to run this soon. all are considered beta at this time.

 

Ken

 

[Geister]

Brian has sold me the 1090, the 80 column card and the 320K ram card.  I'm still in "Getting to know you" mode but I'm very happy at the moment.

 

The 80 column card works great with Basic.

Edited January 22 by Geister
fixing horrible, horrible spelling and grammar.

[kenames99]

Freaking awesome! but you didn't show up the actual hardware.

[Geister]

See the link 

 

Edited January 22 by Geister

[Geister]

Please throw money at Brian so that we may have more of this!

[Geister]

I don't know what happened to the video I uploaded, it seems to have died.

 

I've got to say that I'm surprised that there isn't more enthusiasm about the existence of a functional 1090 box.  

 

If you basically feel MEH about the 1090, why?  I do understand that there is a lot of functionality available for the 8-bit computers, but much of it requires major hack-age of the interior of your beloved Atari and some of it seems down-right flakey in behavior (ie. the Rapidus). 

 

The 1090 by itself may not catch your enthusiasm, but the promise of current and future drop in upgrades should be of some value.

 

Feel free to weight in.  Have you already invested in any updates you feel you'll need?  Do you not see any value in the 1090?  Are you waiting for the "Killer App" to or don't you think there is one?

 

I was excited for the 1090 back in the day because I figured it would make possible things that I felt the Atari needed to succeed like an 80 column card.  I'm excited for the 1090 now because I felt that the solutions we did receive along the way were half-assed like the XEP 80.  I'd like to see better solutions come along with the existence of an expansion system.  

 

Or maybe its just something for which the time came and went...and this is just a niche for hobbyists to play in.

[rkindig]

I'm very excited by all this!  I think this is a killer way to upgrade the Atari 8-bit machines.  I've been trying to get Brian to sell me some of this stuff, but I understand that it's still under development.

 

I want the RAM upgrade, FujiNet, 80-column and probably other possible cards that haven't yet been thought of!

 

Randy

[+mytek]

The 1090 really would have been far better executed inside the same enclosure as the computer. This is how the Apple II and the IBM PC approached this. Basically the whole idea of having something attached via a ribbon cable which needs to be kept extremely short, is not a very practical way to go. The whole thing just takes up far too much desk space in my opinion, and seems more akin to a developers system, than a nice compact desktop computer (which is what you'll still have with the internal upgrade path).

 

Personally I'd really like to see this morph into a single enclosure and mimic the idea of the S100 back plane. This would have a spot for the computer to plug in as a card (might be best to redesign and condense the Atari motherboard for this method), A USB Mouse/Keyboard/general purpose USB I/O card, RAM card, VBXE/80 column card (preferably with HDMI output), SIO/RS232/Parallel/FujiNet card, and then have the entire thing powered off of a standard internal ATX PSU (no need for all the independent heat radiating linear regulators). And make this fit into a standard ATX case so that an inexpensive enclosure of the users choice can be utilized.

[Geister]

You make a good point but the time for that would have been when Atari was designing the 1450XLD and had done that instead of the internal disk drive, speech chip and modem.  While I loved the 1450XLD at the time, looking at in the rear-view mirror, it was a train wreck.  At the time, nobody knew what a PC should look like and Atari found itself with the 600XL and the 800XL and the 1090 was supposed to support the entire XL line (excluding the 1200XL).  If Atari had decided to follow your idea we might have got the expansion chassis/computer we deserved, but they were not thinking Apple II, they were thinking multiple computers with a single expansion chassis to serve them all (So more like a TI 99/4A and expansion chassis.).

 

So what can we do now?