New memory boards for the P2

  • As a D flip flop has tow outputs and the second one is the first complimentary, the addition of an or gate allows to address the lower, unbanked 16k.


    translation+decode.png


    So if a '138 is fed with "-SysRegion" select signal...



    Four controlled 2k regions can be selected independently: three ROM select lines and a RAM select line. Later I'll deal of how this is tied against SRAM and ROM, but the following is the goal to achieve.



    There's overlap on the MOS memory and also on the ROM region. That's more or less what's wanted: overlap must be controlled, mainly by switches. Most people (including me) doesn't have material to read a TMS27xx, even less writing it. By using an standard 27c64 or 2864 (I think it was this reference) and three select lines, ROM can be selected in 2k chunks, allowing both partial (if > 0 but < 3 system roms failing) and full ROM replacement.

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:


  • I've tied the previously generated control lines. The EPROM is also included in the diagram. All overlapping on the 16k is controlled. At this point the memory mode is not selectable and stuck at 64K mode. Once redesigned the switch logic, I'll start optimizing the diagram, placing references (Ux) at the ICs, specifying power supply pins and this kind of stuff. Still, counting gates and other components, this board should be buildable with 11 ICs.


    I have to put this project on hold again: I have an exam in 19/06 so active development will temporally cease (but I'll follow this and other threads) until later in that same day or until the following one.


    That being said, I'd like to know if you (especially who are experienced with electronics) can find points where this design may fail. After all, It's my first project and, while I've soldered a lot of kits, I've never created a board from scratch.


    Thank you very much!

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:

  • Greetings!


    I've returned, and with a new version of the schematics! :)



    Those modifications should enable/disable RAM sections as well ROM using a switch.


    Switch positions are:


    ROM Switch (SW1 on the schematics):

    SW1 SW2 SW3 Effect
    0 x x MOS ROM (on the new board) disabled
    1 x x MOS ROM (on the new board) enabled
    x 0 x Floppy control ROM (on the new board) disabled
    x 1 x Floppy control ROM (on the new board) enabled
    x x 0 Video and keyboard control ROM (on the new board) disabled
    x x 1 Video and keyboard control ROM (on the new board) enabled



    RAM Switch (SW2 on the schematics):

    SW1 SW2 SW3 Effect
    0 x x RAM 1800h-2000h disabled
    1 x x RAM 1800h-2000h enabled
    x 0 x Memory limited to 8000h-FFFFh, bank switching disabled (aka 32K mode)
    x 1 0 Memory map expanded to use the 8000h-FFFFh range, bank switching disabled (aka 48K mode)
    x 1 1 Memory map expanded to use the 8000h-FFFFh range, bank switching enabled (aka 64K mode)


    Have a nice evening.

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:

  • Hello Jaume,

    I hope the exams are now completed successfully.

    In summertime you have less time to take care of the computers.

    Something in the garden or vacation.

    I have looked at the new plan.


    The USER netmercer and gpospi has a lot of experience with the changes and also the repair of hardware units. He had reported here on this subject - he is possible on vacation.


    I think only from the philophie of the software as a memory card can then be used!

    I would like to know more about the strategy of the real memory parts for this memory card.

    Why should the EPROMs be switched ON or OFF by a mechanical switch?


    So what memory areas can be configured by switches and how can the program run which memory areas (PORT - BANKING) are switched.


    If I the DS2069 memory card is 128 kB. In the reset state only the 48kB-bank 0, of 4000h-FFFFh become active. The MOS (0000h - 17FFh) and RAM of 1800h-1BFFH are also active on the CPU with RESET. And clearly the display controller memory!

    This is a P2 configuration (sks, TA, Dr. HELL) to operate with a 48 kB cp / m.


    For a DS2069 I have a 64 kB cp / m (HELL other PORT banking) with a software RAM DISK of 48kB (from the BANK 1 of 4000h - FFFFh)

    available. I'm looking for the software after the holidays - just as background.


    Greeting

    Helwie44

  • Hello Helmut, nice to see you!


    This summer will be full of computers. I've a Commodore computer coming and maybe a third P2 :S.

    My philosophy about the board is a restorative one, it is not meant as an upgrade (well, for 48K P2 sure, it will be an upgrade). Video memory is not addressed in this project.


    ROM is switched on/off with a switch because is meant to coexist with the original ones. In case of an original ROM failure, the only thing to do to restore the computer to a functional state would be activating the appropiate switch. This way the TMS ROMs are left out of the equation the same way as the 4116 would. Also, if they were not deactivated, conflict between the original ROMs and the new one may arise.


    Thank you very much,

    Jaume

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:

  • Interesting idea to create a board which may replace several other potential faulty components. Unfortunately I am not so deep into hardware design and the memory layout of the P2. Overall I am not sure if having e.g. the floppy ROM "backup" on the board is very useful. If the floppy controller is broken, just having a working ROM will not really help. If the floppy controller is generally working, I would replace a potentially faulty ROM directly on the controller.

  • Interesting idea to create a board which may replace several other potential faulty components. Unfortunately I am not so deep into hardware design and the memory layout of the P2. Overall I am not sure if having e.g. the floppy ROM "backup" on the board is very useful. If the floppy controller is broken, just having a working ROM will not really help. If the floppy controller is generally working, I would replace a potentially faulty ROM directly on the controller.

    Hello gpospi! Nice to see you too.


    Well, this part would only replace the rom containing the controller routines. If the floppy controller is failing unfortunately another solution should be sought (aka. another board). This is meant to solve memory issues, both random and read-only. Unless you floppy controller ROM is failing it won't help to solve your floppy issues. Unfortunately, this is outside the scope of this project, but sure I'll take a look and taking notes and ideas for possible future projects.


    I'm sorry about delaying this so much. My inexperience weights a lot...

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:

  • After 3 months in transit and a complaint later my memories have arrived this same morning. They were expected to arrive a month and a half ago.


    I think that it is not necessary to say that once I received them rushed its construction...



    Then proceeded to its testing.




    FFEF -> 48KB recognized!


    That being said, I introduced a small error in the design of the board: about 1-1.5mm should be removed in order to place the bus connector. Otherwise everything is fine.



    Regards

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:

  • BTW: Could you meanwhile verify if the 64 KB mode also works? It seems that the 64 KB RAM card in my P3 is now faulty and it would be interesting to test if the machine runs with the Elephant board. However I also suspect an error in the CPU board, hence potential problems with the Elephant board might not automatically be caused by the board itself.

  • Btw, what is the purpose of the U18 socket? Is it to use alternative RAM chips?

    It's U1 A/B :). The same ic, different packages. But I have yet to see a narrow 128K x 8 sram.:/

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol:

  • BTW: Could you meanwhile verify if the 64 KB mode also works? It seems that the 64 KB RAM card in my P3 is now faulty and it would be interesting to test if the machine runs with the Elephant board. However I also suspect an error in the CPU board, hence potential problems with the Elephant board might not automatically be caused by the board itself.

    I have a newer design that has to surpass the 64KB limit. The problem is that either the original trigger and its interactions with the other boards are still not documented. A few months ago I started to trace the board using continuity tests but it is too complicated to do it in a reasonable amount of time. It is also imprecise as broken traces affect the outcome. Without understanding those, any banking design won't be compatible.


    What I do certainly know is that decode is partial and in case of the P3 with two memory boards configuration the data is outputted to the data bus but it is inputted from slot C. To know more I began to disassembly the P3 CPU board (as I understood you needed more information about it) but the desoldering pump broke at mid-job. Fortunately today I have received the replacement and may continue as soon as possible.

    When I tried to list all retro systems I have at home, the "The message is too long, must be under 500 characters" error appears! :lol: