1090XL remake

[+DavidMil2]

10 hours ago, reifsnyderb said:

It was for a remote power switch and remote LED.  A switch could go between pins 1 and 2.  A power LED, and resistor would go between pins 3 and 4.

 

 

I added it as an idea and removed it in newer versions.

 

 

So the connector could be used for anything.  I was thinking about cutting the traces on 1, 2, and 3 and running 5 volts on pin 3, +12 volts

on pin 2 and either a 12-volt ground or a -12-volts on pin 1.  Mainly for an internal fan, but possibly for any internal device one might want

to add to this box.  I left an extra inch above the expansion boards for cables or whatever one might add.

 

DavidMil2 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[+DavidMil2]

10 hours ago, kenames99 said:

hi David,

  I am looking at the picture you posted just above and see it has too many ICs installed! only 1 of those on the right side of that photo should be installed. the other is an optional IC if you prefer a different driver IC. please remove one or the other. the same is true of the other ICs on the other side of the board, there is optional , well, options.  use either the 74LS244 drivers or the 74LS541 drivers but not both! I have attached a photo of my working 1090XL for reference.

 

Ken

 

PS: hope everything returns to normal there in a reasonable timeframe.

I think that reifsnyderb pointer that out to me too.  I'll have to remove all three chips before I do anything with the expansion box.

But it looks cool with all the chips installed!

 

David

[Mathy]

Hello guys

 

Maybe someone should put the "either ... or ..." in the files you need to recreate the new boards.  So that the text appears on any new board that is made in the future.  Just to save us all a lot of frustration.

 

Sincerely

 

Mathy

 

[reifsnyderb]

Here's the latest schematic.  While most information was already on the schematic, I added some additional notes about the transceiver chips.

 

1090.pdf

 

[kenames99]

thanks Brian!

[reifsnyderb]

I made a video about the 1090XL and it's cards.  Sorry about the poor picture quality.

 

 

 

[+scorpio_ny]

Hi @reifsnyderb,

That was fantastic!  It is great to see all these cards working together!

[+Allan]

Fantastic, @reifsnyderb. Looking forward to getting one of these with some cards. I would love to see a printer buffer card. Maybe even add it to another card. Maybe someday we could have a USB card that could use some early USB devices, like 3.5 floppy drives, early USB cameras, USB controllers, etc.

[+kheller2]

Great video.   The legacy CPM card looks huge compared to the modern boards!  

[reifsnyderb]

1 hour ago, Allan said:

Fantastic, @reifsnyderb. Looking forward to getting one of these with some cards. I would love to see a printer buffer card. Maybe even add it to another card. Maybe someday we could have a USB card that could use some early USB devices, like 3.5 floppy drives, early USB cameras, USB controllers, etc.

Every now and then I look for a USB host solution...that is reasonable for the application.  It appears there has been some experimentation with a Raspberry Pi Pico.  I don't think it would be too hard to add some USB support once a USB host device is found as there should be libraries that could be leveraged.

 

[reifsnyderb]

12 minutes ago, kheller2 said:

Great video.   The legacy CPM card looks huge compared to the modern boards!  

Thanks!  The CPM card is a newer design but it's modeled to look like a legacy board.   🙂    It should be possible to use an ez80, with a bunch of surface mount chips, to build a smaller and more modern board. 

[venom4728a]

Awesome Video!

 

I am interested in the serial/parallel card, and also the fujinet card.

 

IDE/cf/sd and floppy drive cards would be cool too.

[reifsnyderb]

30 minutes ago, venom4728a said:

Awesome Video!

 

I am interested in the serial/parallel card, and also the fujinet card.

 

IDE/cf/sd and floppy drive cards would be cool too.

Thanks!

 

There will soon be two serial/parallel boards.  One is a remake and the other is a prototype that is heavily influenced by the re-make.  Right now, the re-make board isn't working as planned.  However, the pictures of what we believe to be a working original board had bodge wires.  In addition, we suspect that all older chips, with slower transition times, may be required as there could be timing issues.  We don't really know.  @kenames99 is currently assembling a re-make board as well to see if he can get it working.

 

After some frustration, I decided that the re-make also had some other issues.  In particular, it uses a RIOT chip and an out-of-production eprom chip.  Also, the timing may be hokey.  So, I re-did the board with components that are available new, today.  I added the 74LS123 circuit so as to properly time the circuitry.  The RIOT chip was replaced by a 6520 and a AS6C1008 chip.  (Yes, most of the RAM is wasted, but it's still cheaper than tracking down a working RIOT chip.)  The EPROM was replaced by a modern flash ROM.  In addition, I was able to add an extra 2k bank to the ROM.  Much of the glue logic was replaced by the big ATF2500C PLD.  (This PLD is really cool and is practically a CPLD in that it has buried logic and buried latches.   🙂  )  Also, I was able to add a jumper so two of these cards could be installed.  (The available documentation mentions adding the capability to have more than one parallel/serial card.)  I did have to remove the second serial port.  However, the re-make second serial port would have required a ribbon cable, connector, and mount for the card slot.  So, it would have used up a card slot.  Adding the capability to have more than one card resolves this problem and adds a second printer port.

 

So, here's what I came up with:

 

 

I'll be ordering a batch of test boards soon.....

 

 

 

 

 

[Mathy]

Hello Allan, guys

11 hours ago, Allan said:

 I would love to see a printer buffer card. Maybe even add it to another card.

 

Why not add it to the serial/parallel card?  For "parallel out" only.  "Parallel in" would be nice for old centronics webcams.

 

Sincerely

 

Mathy

 

 

 

 

[LessNick]

12 hours ago, reifsnyderb said:

I made a video about the 1090XL and it's cards.  Sorry about the poor picture quality.

Interesting video, but why such low quality? Only 360p? o_O

[Forrest]

The video would be much better if you could actually read what’s on the monitor.

[reifsnyderb]

3 hours ago, Mathy said:

Hello Allan, guys

 

Why not add it to the serial/parallel card?  For "parallel out" only.  "Parallel in" would be nice for old centronics webcams.

 

Sincerely

 

Mathy

 

 

 

 

The ROM shows the board was programmed for parallel out, only.  However, the design is such that parallel in would be possible if the ROM were to be updated.  By using a RIOT or PIA chip, parallel input could be accomplished.  (I decided to use a PIA chip to retain this capability.)

 

The parallel/serial card firmware specifications include a list of possible improvements.  They are listed as follows:

 

9.3  Possible Improvements

 

These would be subject to ROM and RAM space restrictions on technical feasibility, and to time and budget restrictions on available development resources.

 

1.  Restore 45.45 and 56.875 baud using processor to generate serial data clock under timer interrupt control.

 

2.  Modify serial port device drivers to respond to P2 device I.D. for serial printers.

 

3.  Add a new printer (P) Handler, with new features:

  a.  Concurrent Output, using default 40 byte buffer in main memory or user supplied buffer in IOCB.

  b.  Reporting of additional printer port status lines:  Paper End, ACK, Select.

  c.  Control of additional printer port control lines:  INIT, Auto Feed, Select.

  d.  Printer Port input (TBD)

  e.  Redirect printer data to serial port.

4.  Modify Serial Handler so that a second SPC device could implement R3 and R4.

 

While I doubt #1, above, is important, #2 should be pretty easy.  Adding #3b, #3c, #3d, #3e should be pretty easy.  Adding #4 is also planned as I added that jumper so two cards could be used together.  Also, I added another 2k ROM bank so as to ensure there is plenty of room for the above features.

 

#1, above, was supported by the 850 interface.

 

The first thing to add, to the firmware, is CIO support.

 

A printer buffer would be a nice feature to add.  I'll double-check the schematic, but I don't see a good solution at this time.

 

 

 

 

 

 

 

 

 

 

 

Edited July 23 by reifsnyderb

[_The Doctor__]

Those baud rates are useful for communications with Baudot, teletypes, and for RTTY. They also worked with the paper tape reader. Ham radio and VCF attendees prolly would enjoy them still being supported imho.

[reifsnyderb]

31 minutes ago, _The Doctor__ said:

Those baud rates are useful for communications with Baudot, teletypes, and for RTTY. They also worked with the paper tape reader. Ham radio and VCF attendees prolly would enjoy them still being supported imho.

Would this be done by using POKEY timers?

[reifsnyderb]

For a while, I've been messing around with a 1090XL board that would upgrade the OS.  This is possible, of course, but any software that changes PORTB bit 0 will result in a system crash.  It doesn't matter if the OS is running off of RAM or still on the ROM.  Recently, I decided to modify the firmware board by adding a chip that monitors all the D1FF bits to determine if another card is enabled.  If no card is enabled, it can assert the math pack disable signal and swap out the built-in floating point package with Fast Math F+.  Testing shows this usually works quite well.  If, instead of replacing the OS I use the OS to add more features, that works too.  So, I put together an OS Extension Card.

 

This card allows for the following:

 

a. A port of Matthias Reichl's HIAS High Speed SIO that works with compatible devices. (i.e. Happy Enhancement)

b. Fast Math F+ floating point math pack to, on the average, double the speed of floating point calculations that use the built-in floating point routines.

c. Left screen margin set to 0

d. Reduced key repeat delay

e. Reduced key repeat rate

f. Option key function is inverted, at boot time, on even Device ID numbers.

g. With PBI RAM installed, has the following additional features when Device ID is set to any number except 1 and 4.

  i. Attract mode is turned off

  ii. Additional keyboard commands

  iii. CONTROL/SHIFT/DELETE cold starts the computer

  iv. SHIFT/HELP toggles international character set

  v. CONTROL/HELP sets left screen margin to 0

  vi. CONTROL/SHIFT/INVERSE starts WozMon (Device ID must be set on 2 or 3.)

  vii. Atari Fuji Screen on entry to self test mode. (Press HELP to continue.) (Not available with Device ID's 1 and 4.)

 

I have new boards ordered that will include an additional jumper so as to allow Fast Math F+ to be disabled as well.  All features can be disabled by booting with the Select key held down.

 

Of course, if anyone has any other options they'd like to see added, please feel free to let me know.

 

Here's a picture of the current board:

 

 

 

 

 

 

 

Edited July 25 by reifsnyderb

[reifsnyderb]

Legacy 1090XL Parallel/Serial card KiCad files:

 

1090 Parallel Serial Board R9.zip

 

Note that this card is currently unproven.

[dely]

Can't wait for finished product!

 

Great job @reifsnyderb. Thanks!

[reifsnyderb]

On 7/22/2024 at 11:51 PM, reifsnyderb said:

Thanks!

 

There will soon be two serial/parallel boards.  One is a remake and the other is a prototype that is heavily influenced by the re-make.  Right now, the re-make board isn't working as planned.  However, the pictures of what we believe to be a working original board had bodge wires.  In addition, we suspect that all older chips, with slower transition times, may be required as there could be timing issues.  We don't really know.  @kenames99 is currently assembling a re-make board as well to see if he can get it working.

 

After some frustration, I decided that the re-make also had some other issues.  In particular, it uses a RIOT chip and an out-of-production eprom chip.  Also, the timing may be hokey.  So, I re-did the board with components that are available new, today.  I added the 74LS123 circuit so as to properly time the circuitry.  The RIOT chip was replaced by a 6520 and a AS6C1008 chip.  (Yes, most of the RAM is wasted, but it's still cheaper than tracking down a working RIOT chip.)  The EPROM was replaced by a modern flash ROM.  In addition, I was able to add an extra 2k bank to the ROM.  Much of the glue logic was replaced by the big ATF2500C PLD.  (This PLD is really cool and is practically a CPLD in that it has buried logic and buried latches.   🙂  )  Also, I was able to add a jumper so two of these cards could be installed.  (The available documentation mentions adding the capability to have more than one parallel/serial card.)  I did have to remove the second serial port.  However, the re-make second serial port would have required a ribbon cable, connector, and mount for the card slot.  So, it would have used up a card slot.  Adding the capability to have more than one card resolves this problem and adds a second printer port.

 

So, here's what I came up with:

 

 

I'll be ordering a batch of test boards soon.....

 

 

 

 

 

Update on new parallel/serial card....

 

I decided to use an ATF2500C for U10.  So far, this has been a disaster.  Even stripping the board of chips and programming a single pin to either be high or low results in an output of around 1.4 volts or so on the test I/O pin.  I've been using wincupl, XgPro, and a T56 programmer.  All of these should work fine with the ATF2500C.  If I don't set the "encryption" fuse, I can even read the chip.  Lots of time and money wasted attempting to use this stupid chip.  I've also tried no less than 3 chips and same result.  (Yes, this would indicate that maybe the T56 isn't up to the job.  However, it does read the chip after I've programmed it, and the program is correct.)

🤬🤬🤬🤬🤬

 

Needless to say, I am not impressed by these ATF2500C chips.  They suck.  It doesn't help that there is practically zero information about programming them, the wincupl examples offer no clues as to what is wrong, and the datasheet is more of a sales document than anything else.  Oh, yeah, they also cost slightly over $10 each.  🤬

 

In the past, I have been successful in using ATF1502 and ATF1504 chips.  I am now contemplating making a PLCC-44 adapter to try to get this board working with an ATF1502.  Of course, I'll have to setup my JTAG toolchain again.  As a permanent solution, integrating a PLC-44 chip, on this 2 layer board, won't be a picnic, either.

🤬🤬🤬🤬🤬

 

 

 

 

[reifsnyderb]

Serial Parallel Card Update....

 

Because an ATF1502 has so many power and JTAG pins, I had to install a decoder chip as well.  (The decoder chip is U15 and would be an ATF22V10C.)  I haven't installed the JTAG interface yet, but that would go between U14 and U15.  Most likely, those components could be optional since a socketed ATF1502 could be programmed on a smaller programming board...

 

[+DavidMil2]

On 7/25/2024 at 11:44 AM, reifsnyderb said:

For a while, I've been messing around with a 1090XL board that would upgrade the OS.  This is possible, of course, but any software that changes PORTB bit 0 will result in a system crash.  It doesn't matter if the OS is running off of RAM or still on the ROM.  Recently, I decided to modify the firmware board by adding a chip that monitors all the D1FF bits to determine if another card is enabled.  If no card is enabled, it can assert the math pack disable signal and swap out the built-in floating point package with Fast Math F+.  Testing shows this usually works quite well.  If, instead of replacing the OS I use the OS to add more features, that works too.  So, I put together an OS Extension Card.

 

This card allows for the following:

 

a. A port of Matthias Reichl's HIAS High Speed SIO that works with compatible devices. (i.e. Happy Enhancement)

b. Fast Math F+ floating point math pack to, on the average, double the speed of floating point calculations that use the built-in floating point routines.

c. Left screen margin set to 0

 

 

Of course, if anyone has any other options they'd like to see added, please feel free to let me know.

 

Here's a picture of the current board:

 

 

 

Can you put an option on the card so that every time someone (namely me) presses Uppercase G and ESC at the same time a pot of gold appears???  This would be a very handy asset!

DavidMil2