Image Map Image Map
Page 3 of 3 FirstFirst 123
Results 21 to 26 of 26

Thread: M24-IDE: a 16-bit IDE adapter for the Olivetti M24 / AT&T 6300

  1. #21
    Join Date
    Oct 2014
    Location
    near frankfurt/m, germany
    Posts
    1,054

    Default

    Quote Originally Posted by Trixter View Post
    The EEMS board is a 16-bit board (which I'll be installing next week when I get the 41256 chips to populate it

    I have a memory board as well, but it is (as far as I remember) full of sockets for 4164 chips.

  2. #22
    Join Date
    Aug 2006
    Location
    Chicagoland, Illinois, USA
    Posts
    6,546
    Blog Entries
    1

    Default

    Quote Originally Posted by 1ST1 View Post
    I have a memory board as well, but it is (as far as I remember) full of sockets for 4164 chips.
    Yes, the board I'm talking about is an EEMS board (precursor to EMS 4.0) which was essentially an AST Rampage! board adapted for the 16-bit bus of the M24 and clones, as well as the 286-based 6300 plus. Mine is branded the AT&T Enhanced Memory Expansion Board and can supply from 512KB to 2048KB as either extended memory (286) or as paged EMS (8086). I've put my materials about it here: ftp://ftp.oldskool.org/pub/drivers/A...emory%20Board/
    Offering a bounty for:
    - A working Sanyo MBC-775, Olivetti M24, or Logabax 1600
    - Music Construction Set, IBM Music Feature edition (has red sticker on front stating IBM Music Feature)

  3. #23
    Join Date
    Aug 2017
    Location
    New York, Rome, London
    Posts
    131

    Default

    Quote Originally Posted by Trixter View Post
    The EEMS board is a 16-bit board (which I'll be installing next week when I get the 41256 chips to populate it
    Looking forward to seeing that!

    To clarify my comment: I meant that I thought there werenít any expansion boards that did native 16-bit I/O transfers (as opposed to 16-bit memory transfers). For example, the memory expansion board described in the Theory of Operations manual does 16-bit memory transfers but doesnít actually have any I/O registers. It would be interesting to find out what kind of registers the one you have has - I assume it must have some configuration registers to switch memory banks within the access window, and I wonder if those registers are 8-bit or 16-bit.

    In any case, I was wrong: the OEM video card has 16-bit I/O registers, mapped at 0x3C0-0x3DF. How do I know? The schematics at page B-24:

    7EAF26F5-1DB3-4AAF-889E-38BBB9150513.jpeg

    This is also very useful for me to have found out as it confirms that the proper way to drive /16BCH is to decode both the address lines and the I/O read/write strobes (called /XIOR and /XIOW in the schematic).

  4. #24
    Join Date
    Aug 2017
    Location
    New York, Rome, London
    Posts
    131

    Default Yet More Updates

    Ok, itís definitely almost certainly an intermittent hardware problem (aargh!!!). It turns out that it almost always works - if you try enough times, eventually format works and it even boots DOS!

    8B17A693-1019-416A-BA51-DD8DA5FCC9E8.jpeg

    But itís not stable - the next time around it might not boot.

    So I expanded my low-level IDE testing utility to include a test whereby I write a pattern to the buffer internal to the IDE device and then read it back, and compare. The IDE buffer read / write commands ( 0xE4 / 0xE8 ) allow you to transfer 256 words of 16-bits to and from the device (without actually writing anything permanently anywhere, so itís ideal for testing). I do this with IN AX, DX / STOSW and LODSW / OUT DX, AX (first write a full set of 256 words, then read it back).

    Again - this almost always succeeds, but once in a while it fails (I set the testing utility to repeat the write / read / compare test 0x4000 times, and I got a dozen failures or so). Now I need to find a smart way to capture the timings for when it fails...

  5. #25
    Join Date
    Oct 2014
    Location
    near frankfurt/m, germany
    Posts
    1,054

    Default

    Any progress here? Maybe you are only fighting againsgt glitches occoured by the unshielded flat cable?

  6. #26
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    2,403

    Default

    I noticed in the photo of the boot messages you were using a Sandisk device. Prooobably not the problem, but have you tried something else? I ask because I tested a number of different size/brand cards with my XT-CF adapters and Sandisk cards all ended up in the “bad” category. They might appear to work but running disktest.exe’s “mediatest” produces tons of errors. I assumed it was something to do with 8-bit compatibility, but I am running the same BIOS as you are.
    My Retro-computing YouTube Channel (updates... eventually?): Paleozoic PCs

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •