Image Map Image Map
Page 2 of 7 FirstFirst 123456 ... LastLast
Results 11 to 20 of 61

Thread: Olivetti M24 and IMD -> No interrupt from FDC

  1. #11
    Join Date
    Oct 2014
    Location
    near frankfurt/m, germany
    Posts
    868

    Default

    Hi Chuck, I think you are sydex company which created most of these wonderfull tools I try to use, right? Don't you believe your own tools?

    etvdisk.jpg
    That is the (only) original diskette which I have. (allo others I have now to play arround are copies of this one, made with anadisk)

    teledisk.jpg
    That is teledisk trying to read the disk. When starting to read, I can tell teledisk to make it over BIOS, then even the LED of the drive goes on, when I select the other option to access direct hardware floppy controller I get error message like this screenshot. I also created a DISKETTE.CFG file with the content "A: BIOS 1 UNIT 0 360K CH"

    anadisk.jpg
    And that is what your anadisk is telling about the same diskette. It also copies the disk workable - even copy of copy of copy still works. Give me your email adress and I give you the dump file which anadisk created.

    Ah, by the way... 22disk... (I created a batch file called etvtool which uses the different tools of 22disk to dir an ETV disk, get and put files from/to it. This works perfectly on my Olivetti D33 (80386) laptop with 1.44 MB drive and ETV 250 diskettes (/oli3 disk type). For M24 / ETV 300 I just changend the disk type parameter to /oli1 and that is displayed correctly and 22disk can read the directory of my ETV 300 diskette. But copy files from/to fails.

    etvtool_dir.jpg

    etvtool_get.jpg

    And that's IMD 1.18, saying it gets no interrupt from floppy controller.

    imd_settings.jpg

    imd_no_irq.jpg
    Last edited by 1ST1; September 16th, 2018 at 11:53 AM.

  2. #12
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    31,774
    Blog Entries
    20

    Default

    The Teledisk thing is because of the lack of FM support on your M24. That's a hardware deficiency. Teledisk starts at the beginning of the disk; if no data is readable on that first track, you'll get exactly that message.

    Anadisk, I believe can skip over tracks and try to figure things out.

    I have no idea why you're having problems with 22Disk on your M24 if you can read the directory--all modules use the same disk I/O routines.

    Not that I can do anything about it, but are you a registered user of any of these products? (Note: Anadisk and Teledisk were sold off in 2000, so I've had nothing to do with them in years.) I still do support 22Disk, but will request a sample of your troublesome disk--and I gave away the 6300 last year, so it may not be possible for me to duplicate your problem. How about putting a 5.25" drive on a more up-to-date system with FM support?

  3. #13
    Join Date
    Oct 2014
    Location
    near frankfurt/m, germany
    Posts
    868

    Default

    No I am not regstered, I just discovered these tools about 16 years after you have closed support most of them, except of 22disk. Until now 22disk worked just fine, no support needed, but I only rarely use it - and until now not on M24.

    Are you sure about no FM support by M24? M24 can emulate Olivetti M20 (Z8001 based computer), but I don't know if this is FM or MFM. M24 is using uPD765 floppy controller which is capable of reading frequency modulation and modified frquency modulation

    http://www.cpcwiki.eu/imgs/f/f3/UPD7...heet_OCRed.pdf

    uPD765.jpg

    m24theoryofoperations.jpg
    The M24 Theory of Operations also describes a circuit on floppy controller logics called "DATA SEPARATOR"

    Data is recorded on to the disk using two different techniques, FM and
    MFM. In order to provide maximum data recording density, the Modified
    Frequency Modulation (MFM) technique is mostly used. This technique
    requires clock bits to be recorded only when two successive data bits are
    missing in the serial data stream. This reduces the total number of
    bits required to record a given amount of information on the disk. The
    fact that clock bits are not recorded with every data bit cell requires
    circuitry to synthesize clock bit timing when clocks are missing and to
    synchronize to clock bits when they are present.

    So Olivetti writes here, there is no limitation of uPD765 in M24, it can support FM and MFM.

    I already considered to use another system, but no decison yet, I would prefer that M24 as I already spent a lot of work with it for making it ready for file transfers. I even plan to add an external 720 kB drive in the future, internally there is no space left, and I don't want to cut a hole in the front panel as some peoples did. The M4 modulo P100i would be an option, or the M24-II (Pentium 133), but 1st I would need to get another working 360 kB drive.
    Last edited by 1ST1; September 16th, 2018 at 12:29 PM.

  4. #14
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    31,774
    Blog Entries
    20

    Default

    No, but what does the data separator look like? The 765 does not include a data separator of its own.

    For example, the IBM 5150 and 5160 and 5170 floppy controllers use the 765, but lack support for FM. There was little reason for a manufacturer to include the capability, as in 1982, MFM was the dominant modulation method for floppies.

    To verify one way or the other, try Dave Dunfield's testFDC utility. That will give you an accurate indication of what the M24 can do.

    Methinks you're trying to teach your grandmother to suck eggs.

  5. #15
    Join Date
    Oct 2014
    Location
    near frankfurt/m, germany
    Posts
    868

    Default

    testfdc.jpg

    We have now how many answers in the thread? And still no explanation for the no interrupt thing, even not from grandma.

    You can have a look into the M24 Theory of Operation, there is a block diagram of the data separator which has been build in the M24.

  6. #16
    Join Date
    May 2009
    Location
    Connecticut
    Posts
    4,303
    Blog Entries
    1

    Default

    Dave Dunfield has released source code for Image Disk including TESTFDC. Straight forward code. Error message is in the waitirq function. Run it in a debugger and see what is triggering the error.

    I suspect a problem with the M24 having separating the clock signal into 4 different ones passing through a buffer (S241) while the IBM PC style designs don't. But it is late and I could easily be wrong.

  7. #17
    Join Date
    Oct 2014
    Location
    near frankfurt/m, germany
    Posts
    868

    Default

    I have no experience with that kind of analysis.

  8. #18
    Join Date
    May 2009
    Location
    Connecticut
    Posts
    4,303
    Blog Entries
    1

    Default

    Unfortunately, the M24 floppy subsystem works in unusual ways. It will take someone with one to figure out what changes need to be made to get specialized floppy utilities working.

    Strangely, the only source I have seen documenting anything about what the M24 floppy does differently is with an old version of Minix. I am not sure that having the floppy drive take 3 times as long to startup would cause IMD problems.
    Last edited by krebizfan; September 17th, 2018 at 06:39 AM.

  9. #19
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    3,210

    Default

    The Imagedisk errors are indicative of a floppy controller that is not 100% IBM PC compatible. I don't know how compatible the M24/AT&T 6300 is or if there were model variations. Perhaps someone else with one of these could confirm if IMD works or not?

    A couple of thoughts: I've seen the "no interrupt" IMD error occur when there is a problem reading the floppy index pulse. There isn't anything funny going on with this disk's index hole is there? Can you format a blank disk in this machine? (index detection is required for formatting).

    Also, if this machine is really loaded up, perhaps there is some hardware conflict?

    I hear what you are saying about no more room for additional machines. If space is a concern, you might consider a Kryoflux or SuperCard pro device attached to a newer computer.

  10. #20
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    31,774
    Blog Entries
    20

    Default

    The "no interrupt" may be an artifact of the implementation. Usually this means a hung 765--the PC implementation disables a lot of functionality (e.g. unit select, polling, drive-ready status, etc.) so things have to be timed by software, which is not an optimal situation. So a timeout must be followed by resetting the 765, then performing the SPECIFY operation and recalibration to track 0.

    In addition, the 765 is a rather slow chip (it's really part of the 8080 family), so some operations have to be delayed for several microseconds--and you'll see BIOS code to that effect when issuing commands.

    In the case of delay loops, I can tell you that I used the refresh status to give me a 50 usec. interval, or, if it's not present, a software loop, which, most likely like Dave's is calibrated to a 4.77MHz 8088 timing.

    I've also seen "no interrupt" timeouts where the subject disk has a bizarre physical interleave, such that it can take more than one revolution, but less that 2 complete revs to locate a sector.

    I'll have a look at the M24 schematics and see if anything jumps out at me. But you're still not going to use the M24 to image these disks with an FM boot track, no matter what else is uncovered for reasons that I've explained elsewhere.
    Last edited by Chuck(G); September 17th, 2018 at 08:25 AM.

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
  •