Jump to content
retrocanada76

Designing an Axlon compatible board

Recommended Posts

7 minutes ago, Jeffrey Worley said:

I spent the weekend with a 544k 800 as my personal machine and really liked it.  I did the phi2 mod on it so I could use the Side2 cartridge for mass-storage.  Aside from the fact that the lid needs to stay open to accommodate the tall cartridge, the machine is very compatible and very pleasant.  Spartados loads with a low-memlo due to being able to relocate itself into the axlon memory, and the drivers for the Side's CF card load, which they won't on a stock 400/800 machine.  What was particularly and pleasantly surprising was how much software actually runs on a 48k 800.  Most all games run.  There are exceptions of course, but it is a much more useable Atari than I would have credited it to be without the Incognito installed.  Of course Incognito makes nearly everything run, but a stock 800 with this ram card in it still has a not of miles left on it!  I was impressed!  What a great computer.

 

best,

 

Jeff

Sounds like a winner!

like maybe you should do an interest/order thread.... ;)

 

Share this post


Link to post
Share on other sites
43 minutes ago, _The Doctor__ said:

Sounds like a winner!

like maybe you should do an interest/order thread.... ;)

 

I could do that, but I spoke with the Corey Koltz from 8bitclassics and he indicates that he is going to put these boards in production and sale on his site sometime this year.  In the mean time I suppose I could make more.

 

Is anyone else interested in more of these boards?  If I can get ten orders then I will make a short run of them and have them ready in a few weeks.

 

best,

 

Jeff

Share this post


Link to post
Share on other sites
1 hour ago, Jeffrey Worley said:

I spent the weekend with a 544k 800 as my personal machine and really liked it.  I did the phi2 mod on it so I could use the Side2 cartridge for mass-storage.  Aside from the fact that the lid needs to stay open to accommodate the tall cartridge, the machine is very compatible and very pleasant.  Spartados loads with a low-memlo due to being able to relocate itself into the axlon memory, and the drivers for the Side's CF card load, which they won't on a stock 400/800 machine.  What was particularly and pleasantly surprising was how much software actually runs on a 48k 800.  Most all games run.  There are exceptions of course, but it is a much more useable Atari than I would have credited it to be without the Incognito installed.  Of course Incognito makes nearly everything run, but a stock 800 with this ram card in it still has a not of miles left on it!  I was impressed!  What a great computer.

 

best,

 

Jeff

Second this one, completely...

 

And get ready for a little "shocker": 

 

Booting a SDX Ultimage/SD or MaxFlash (1Mbyte) cart=image, from SIO-attached .ATR with config.sys, autoexec.bat, etc., plus RAMROD in 52K mode, I am able to load a (wrapped) special version of Avery's XEP80 Ultra-driver in the $C000-$CFFF region, with its own memory-manager, and, after everything is said-and-done, I end up with 37,200 Bytes of free SDX base RAM, AND 80-columns active (!)

 

It all goes to show the 800's (unique) architectural flexibility, and how easy it is to expand its functionality, cleanly and neatly, and by keeping the base 800-HW essentially intact (!)... It's virtually plug-and-play, with no need of PBI-bus, and it all remains nicely tucked inside the expansion bay!

 

...just like a pregnant goose! 😜😉💪

Share this post


Link to post
Share on other sites
52 minutes ago, Faicuai said:

Second this one, completely...

 

And get ready for a little "shocker": 

 

Booting a SDX Ultimage/SD or MaxFlash (1Mbyte) cart=image, from SIO-attached .ATR with config.sys, autoexec.bat, etc., plus RAMROD in 52K mode, I am able to load a (wrapped) special version of Avery's XEP80 Ultra-driver in the $C000-$CFFF region, with its own memory-manager, and, after everything is said-and-done, I end up with 37,200 Bytes of free SDX base RAM, AND 80-columns active (!)

 

It all goes to show the 800's (unique) architectural flexibility, and how easy it is to expand its functionality, cleanly and neatly, and by keeping the base 800-HW essentially intact (!)... It's virtually plug-and-play, with no need of PBI-bus, and it all remains nicely tucked inside the expansion bay!

 

...just like a pregnant goose! 😜😉💪

I always wondered what the male card connector on the rear of the 800 was for, always assumed it was for a test rig connection, but there must be some kick butt signals coming out of that thing.  No one has exploited this connector?

Edited by Jeffrey Worley

Share this post


Link to post
Share on other sites

No, there's only one signal there that's not on some other slot. It's called EXSEL (different from the /EXSEL signal on slots 0 and 3). It disables the motherboard's main address decoder which, in turn, disables all RAM, ROM, and I/O registers (except ANTIC). 

 

I presume it's there so the tester can insert its own RAM and ROM into the address space. Interesting that the 400's test connector doesn't have that signal.

Edited by ClausB
  • Like 2

Share this post


Link to post
Share on other sites
5 hours ago, ClausB said:

No, there's only one signal there that's not on some other slot. It's called EXSEL (different from the /EXSEL signal on slots 0 and 3). It disables the motherboard's main address decoder which, in turn, disables all RAM, ROM, and I/O registers (except ANTIC). 

 

I presume it's there so the tester can insert its own RAM and ROM into the address space. Interesting that the 400's test connector doesn't have that signal.

This is beautiful!

 

Isn't this what we need to run a 200-400 Mhz 6502-compatible CPU board with its one memory space, and stream, at local-bus speed, pre-cooked GFX-data to be displayed immediately by ANTIC and GTIA? Does the local 6502 remains active with EXSELin use?

 

The idea is to let the high-speed CPU board have its own graphics output, but still being able to rely on local physical interfaces like Keyboard, SIO, PIA-ports, etc..

Share this post


Link to post
Share on other sites
On 7/6/2020 at 12:01 PM, Jeffrey Worley said:

I could do that, but I spoke with the Corey Koltz from 8bitclassics and he indicates that he is going to put these boards in production and sale on his site sometime this year.  In the mean time I suppose I could make more.

 

Is anyone else interested in more of these boards?  If I can get ten orders then I will make a short run of them and have them ready in a few weeks.

 

best,

 

Jeff

Boards and parts are ordered. I should have them in within a couple of weeks.

Share this post


Link to post
Share on other sites
On 7/6/2020 at 12:01 PM, Jeffrey Worley said:

I could do that, but I spoke with the Corey Koltz from 8bitclassics and he indicates that he is going to put these boards in production and sale on his site sometime this year.  In the mean time I suppose I could make more.

 

Is anyone else interested in more of these boards?  If I can get ten orders then I will make a short run of them and have them ready in a few weeks.

 

best,

 

Jeff

Boards and parts are ordered. I should have them in within a couple of weeks.

 

Ok, so I can write up a site description, v1.2 of the board as is can be used to replace all three 16K boards as long as it sits in slot 2 and has the pin attaching to s6/pin20? If not, it just acts as a 16K board plus expanded memory is used as a ram disk?

 

What are the mods on the back of the board for? The resistor/capacitor combo.

 

It can be used in the Atari 400 as is to replace the 16K. No mod has been found to make it work as a 48K board?

 

Thanks for the help, looking forward to making it available.

Share this post


Link to post
Share on other sites
16 hours ago, 8bitclassics said:

Boards and parts are ordered. I should have them in within a couple of weeks.

 

Ok, so I can write up a site description, v1.2 of the board as is can be used to replace all three 16K boards as long as it sits in slot 2 and has the pin attaching to s6/pin20? If not, it just acts as a 16K board plus expanded memory is used as a ram disk?

 

What are the mods on the back of the board for? The resistor/capacitor combo.

 

It can be used in the Atari 400 as is to replace the 16K. No mod has been found to make it work as a 48K board?

 

Thanks for the help, looking forward to making it available.

I replied to this and for some reason it did not post, so if a duplicate appears.... well, things happen.

 

The board works great in the 400 as a 16k board.  Only uses 5volts.  

 

In the 800, it must be accompanied by at least one stock 16k board (or another of this very board acting as a 16k board) in slot one.  A 48k base memory machine requires a 16k board in both slots 1 and 3 and this board in slot 2.  To get 544k out of the machine, this board in slot 2 needs a wire attached to the personality board.  One could use a test-clip to do this connection instead of soldering it as I've done.

 

The back of the board has a resistor and a small capacitor as a final bodge to complete it.  The resistor is a pull-up and is documented in this thread, pretty early on.  The two components are required for proper function.  IIRC, the resistor is 3.3k ohm and the capacitor is a .1uf.

 

One could populate all three slots with this card and wire just the one in slot 2 to the personality board.  This would give you a very low-power configuration that would be very nice on the power supply, as opposed to the stock Atari cards, which use 5volts, -5 volts, and 12 volts (and -12 volts?).  It is a drop-in replacement for the stock card.

 

Best,

 

Jeff

Share this post


Link to post
Share on other sites

I got the boards in and assembled the first one except the capicitor. My .1uf are Axial leads so they won't fit on the board properly. Recommendation? These radial leads will fit the 2.5mm spacing. 

 

Mouser

 

I will probably use the axial leads on the back to keep it pressed up against the board.

20201102_174501 (1).jpg

20201102_185051.jpg

  • Like 1

Share this post


Link to post
Share on other sites
On 7/6/2020 at 11:56 AM, Jeffrey Worley said:

What was particularly and pleasantly surprising was how much software actually runs on a 48k 800.

 

It's not really that surprising: software developers knew well the common denominator for the 8-bit line of computers. Ultimately, it was more than a minor annoyance for those expecting to see the extra memory in their (stock) 64 KB and 128 KB machines exploited to some advantage more often and brought up to standards more commensurate with that of their peer/rival machines.

 

Share this post


Link to post
Share on other sites
11 hours ago, 8bitclassics said:

I got the boards in and assembled the first one except the capicitor. My .1uf are Axial leads so they won't fit on the board properly. Recommendation? These radial leads will fit the 2.5mm spacing. 

 

Mouser

 

I will probably use the axial leads on the back to keep it pressed up against the board.

20201102_174501 (1).jpg

20201102_185051.jpg

https://www.amazon.com/gp/product/B08B3VCK42/ref=ppx_yo_dt_b_search_asin_image?ie=UTF8&psc=1

 

These are the ones I bought for mine.

 

Best,

 

Jeff

Share this post


Link to post
Share on other sites
5 hours ago, MrFish said:

 

It's not really that surprising: software developers knew well the common denominator for the 8-bit line of computers. Ultimately, it was more than a minor annoyance for those expecting to see the extra memory in their (stock) 64 KB and 128 KB machines exploited to some advantage more often and brought up to standards more commensurate with that of their peer/rival machines.

 

Much of the good system-software, as opposed to games, uses the extra memory (of XL/XE) in one way or other.  These don't run on the 400/800.  However, when you have a dos like SDX, that uses Axlon ram to move itself into, that makes a big difference in compatibility with such software.

 

best,

 

Jeff

  • Like 2

Share this post


Link to post
Share on other sites
4 hours ago, Jeffrey Worley said:

Much of the good system-software, as opposed to games, uses the extra memory (of XL/XE) in one way or other.  These don't run on the 400/800.  However, when you have a dos like SDX, that uses Axlon ram to move itself into, that makes a big difference in compatibility with such software.

 

SDX is an amazing piece of ware.

 

Share this post


Link to post
Share on other sites
9 minutes ago, MrFish said:

 

SDX is an amazing piece of ware.

 

I first saw the cart in an advertisement that came with my RTIME8 cart, last page of the manual or something.  I was only 16 or 17 and had little in the way of income, but I sent them the 79.00? pre-order right then and there.  I got one of the first carts off the line.  When I finally got the thing in hand, some months later, I was astonished to see how well-done the manual was.  Last thing I expected.  I read the thing cover-to-cover, probably several times.  SDX was better than I could have DREAMED when I read the advert.  And it is still going strong.  Fat16 support!, all kinds of groovy stuff.  God Bless ICD and the coders who worked there, and God bless the folks at the Spartados Upgrade Project (Drac030?).

 

Best,

 

Jeff

  • Like 1

Share this post


Link to post
Share on other sites
4 hours ago, Jeffrey Worley said:

Much of the good system-software, as opposed to games, uses the extra memory (of XL/XE) in one way or other.  These don't run on the 400/800.  However, when you have a dos like SDX, that uses Axlon ram to move itself into, that makes a big difference in compatibility with such software.

 

best,

 

Jeff

+1, Back in the day, we built the 256K Atari upgrade for my 800XL ("we" is my Dad did it)....and what did I use it for, RamDisk, and it worked great.  The BBS bogged on picking up posts from the 1050 drive, so I ran the BBS all day on ramdisk and backed it up once a day.

 

I didn't know it then, but what I would have liked from Atari, it to upgrade the graphics of the machine.  But they just never did. If they had, and games had depended upon it, they would have driven an upgrade cycle, possibly....who knows, also it is possible the market would have hated it, we didn't like our purchases being obsoleted back then.

Share this post


Link to post
Share on other sites
10 minutes ago, Dmitry said:

+1, Back in the day, we built the 256K Atari upgrade for my 800XL ("we" is my Dad did it)....and what did I use it for, RamDisk, and it worked great.  The BBS bogged on picking up posts from the 1050 drive, so I ran the BBS all day on ramdisk and backed it up once a day.

 

I didn't know it then, but what I would have liked from Atari, it to upgrade the graphics of the machine.  But they just never did. If they had, and games had depended upon it, they would have driven an upgrade cycle, possibly....who knows, also it is possible the market would have hated it, we didn't like our purchases being obsoleted back then.

The biggest difficulties with upgrading the Atari 8-bit in meaningful ways are upshots of the things that made it so special and powerful.  The machine is tied to NTSC/PAL/SECAM for graphics output and for system timing.  The whole machine is wrapped around the TV standards of the 1980's.  To add more sprites or colors or whatever would have been easy enough, but to add PIXELS, add RESOLUTION, which we all wanted and which would have been the definition of 'upgrade' to most of us, was impossible.  The screens didn't exist to use higher resolutions, and those that did, like VGA monitors, were expensive and not really compatible with the colorclock timing of the Atari.  The original design, in this sense, was a very nice, big boat you built in your basement.....  The same applies to system speed.  The system's clock is dependent on the color clock, being a multiple of it, and the graphics hardware can't push faster than color clock without the addition of a slew of expensive parts, so they didn't do it.  Even the sweet16 and, lately, Antonia, operate the system bus at the stock clock rate, no practical way around it short of replacing the entire system with an emulator.

 

Best,

 

Jeff

Edited by Jeffrey Worley

Share this post


Link to post
Share on other sites
11 minutes ago, Jeffrey Worley said:

I first saw the cart in an advertisement that came with my RTIME8 cart, last page of the manual or something.  I was only 16 or 17 and had little in the way of income, but I sent them the 79.00? pre-order right then and there.  I got one of the first carts off the line.  When I finally got the thing in hand, some months later, I was astonished to see how well-done the manual was.  Last thing I expected.  I read the thing cover-to-cover, probably several times.  SDX was better than I could have DREAMED when I read the advert.  And it is still going strong.  Fat16 support!, all kinds of groovy stuff.  God Bless ICD and the coders who worked there, and God bless the folks at the Spartados Upgrade Project (Drac030?).

 

The main idea behind my website was to showcase OSS and ICD products. The concept won't be complete until I finish my hardware-based site, though.

 

Share this post


Link to post
Share on other sites

I have 2 of these, one gives me an error when I run scan512.com, but says it passes the test at the end. Is it a bad scram or ???. It gives the error during bank 00, possibly in the unique test. Error msg is vague where and when it is happening.

20210923_185615.jpg

Share this post


Link to post
Share on other sites
On 9/23/2021 at 7:54 PM, Fuji-Man said:

I have 2 of these, one gives me an error when I run scan512.com, but says it passes the test at the end. Is it a bad scram or ???. It gives the error during bank 00, possibly in the unique test. Error msg is vague where and when it is happening.

20210923_185615.jpg

I'm still having this issue, is there another RAM test I can use to verify these boards?

Share this post


Link to post
Share on other sites
38 minutes ago, Fuji-Man said:

I'm still having this issue, is there another RAM test I can use to verify these boards?

Here:

 

tstaxlon.xex

 

Launch from SIO, or from SIDE  or AVG loader.

Share this post


Link to post
Share on other sites
On 10/15/2021 at 9:09 PM, Fuji-Man said:

I'm still having this issue, is there another RAM test I can use to verify these boards?

Were you able to fix this?  I'm having the same issue, but it's not always the same memory location or bank.

- It's usually at 7FFF but sometime 7FFE

- I've had the error in various banks but it changes from day to day.  

- It starts to pass as I let the computer warmup

- The tstaxlon.xex test program from Faicuai always passes.

- I wrote a program to test 7FFE-7FFF in all banks continuously, but it never fails.

- I've tested my hacked 512k Ramscan and the original 128K Ramscan, and they both show failures.

 

Share this post


Link to post
Share on other sites
On 1/31/2022 at 7:40 PM, tep392 said:

Were you able to fix this?  I'm having the same issue, but it's not always the same memory location or bank.

- It's usually at 7FFF but sometime 7FFE

- I've had the error in various banks but it changes from day to day.  

- It starts to pass as I let the computer warmup

- The tstaxlon.xex test program from Faicuai always passes.

- I wrote a program to test 7FFE-7FFF in all banks continuously, but it never fails.

- I've tested my hacked 512k Ramscan and the original 128K Ramscan, and they both show failures.

 

I still haven't solved this, but thru experimentation, I have figure out some details about when it happens. The RAMSCAN test starts by filling a bank with zero's. It verifies the zero immediately after writing.  It then roll's a bit through every memory address.  Before rolling the bit, it checks that the value is still zero from the previous check.  This is where it fails and displays the error message.  The locations at 7FFE and 7FFF are getting corrupted when lower addresses are being written.  After the roll test, an FF is written to the address to indicate it passed.  I've found that the corruption is happening when this FF is being written. Rolling one bit doesn't cause it to happen.

 

- The corruption at 7FFE and 7FFF is always bits being set, never cleared. 

- The corruption happens with any of the banks, but it happens randomly.

- The values that appears at these addresses varies.  I've seen 40, 54,80, A0,C0, D0, FF, FC,FD, FF

- It happens when writing FF to certain memory address.  Specifically, 77EB, 77FB, 7BEB, 7FEB, 7FFB.  There may be others, but these are the only ones I have seen so far.

- The corruption stops as the system warms up.  

- I have tried several different Alliance brand SRAM's, and they all do the same thing.  They are all 55ns.

 

I think an important clue is that the corruption always happens at 7FFE and 7FFF when writing to certain addresses in the 7000-7FFF range and LSB is always EB or FB.  

 

Anyone have any suggestions on what I should be checking?

  • Like 1

Share this post


Link to post
Share on other sites
On 2/13/2022 at 1:19 PM, tep392 said:

I still haven't solved this, but thru experimentation, I have figure out some details about when it happens. The RAMSCAN test starts by filling a bank with zero's. It verifies the zero immediately after writing.  It then roll's a bit through every memory address.  Before rolling the bit, it checks that the value is still zero from the previous check.  This is where it fails and displays the error message.  The locations at 7FFE and 7FFF are getting corrupted when lower addresses are being written.  After the roll test, an FF is written to the address to indicate it passed.  I've found that the corruption is happening when this FF is being written. Rolling one bit doesn't cause it to happen.

 

- The corruption at 7FFE and 7FFF is always bits being set, never cleared. 

- The corruption happens with any of the banks, but it happens randomly.

- The values that appears at these addresses varies.  I've seen 40, 54,80, A0,C0, D0, FF, FC,FD, FF

- It happens when writing FF to certain memory address.  Specifically, 77EB, 77FB, 7BEB, 7FEB, 7FFB.  There may be others, but these are the only ones I have seen so far.

- The corruption stops as the system warms up.  

- I have tried several different Alliance brand SRAM's, and they all do the same thing.  They are all 55ns.

 

I think an important clue is that the corruption always happens at 7FFE and 7FFF when writing to certain addresses in the 7000-7FFF range and LSB is always EB or FB.  

 

Anyone have any suggestions on what I should be checking?

I want to follow up on this issue. I tested a few new SRAM's of the same type, Alliance 55ns, but it still was failing the RAM test when machine is cold. I purchased a Hitachi 70ns to try and it's working fine, cold or warm. I don't know if it is an issue with the Alliance SRAM's or some borderline timing issue, but it's working now, so I'm done messing with it.

Share this post


Link to post
Share on other sites
On 10/6/2017 at 4:15 AM, tep392 said:

Here's a copy of Warpspeed DOS XL 2.42 with some drivers I've updated to support the 512K Axlon clone and the RAMSCAN memory test program which I have also updated for the 512K board.

 

DOSXL242AX512.ATR 90.02 kB · 97 downloads

 

edit: These are the custom files I added.

WSAXL256.COM - Creates single ramdrive, D4: with 931 DD sectors. Uses banks $01-$0F
 
WSAXL512.COM - Creates two ramdrive's, D3: and D4: with 931 DD sector each. Uses banks $01-$0F and $11-$1F. Bank $10 is available for other uses.
 
SCAN512.COM - This is the Axlon RAM test program modified to test 512K.

Hi tep392,

I know that my question is outdated but have I any luck for a "WSAXL128.COM" file?

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...