CF for S-100? - CP/M

This is a discussion on CF for S-100? - CP/M ; Does anyone make a Compact Flash (or SD card) FLASH Drive for S-100 systems? I use CFFA cards (Compact Flash For Apple) ( http://dreher.net/?s=projects/CFforA...pleII/main.php ) for old Apple II equipment to save wear on old drive hardware and to "instant ...

+ Reply to Thread
Results 1 to 11 of 11

Thread: CF for S-100?

  1. CF for S-100?

    Does anyone make a Compact Flash (or SD card) FLASH Drive for S-100 systems?

    I use CFFA cards (Compact Flash For Apple)
    (http://dreher.net/?s=projects/CFforA...pleII/main.php)
    for old Apple II equipment to save wear on old drive hardware and to "instant
    boot" when testing and troubleshooting. The card can be as many volumes of
    whatever size can be supported. Typically a 256 MB CF or larger is used.

    If not available, want one? I think I'll price out an S-100 sized board. The
    components will be minor compared to PCB cost.

    -- Charlie Springer


  2. Re: CF for S-100?

    On Wed, 1 Oct 2008 11:03:11 -0700, Charlie Springer
    wrote:

    >Does anyone make a Compact Flash (or SD card) FLASH Drive for S-100 systems?
    >
    >I use CFFA cards (Compact Flash For Apple)
    >(http://dreher.net/?s=projects/CFforA...pleII/main.php)
    >for old Apple II equipment to save wear on old drive hardware and to "instant
    >boot" when testing and troubleshooting. The card can be as many volumes of
    >whatever size can be supported. Typically a 256 MB CF or larger is used.
    >
    >If not available, want one? I think I'll price out an S-100 sized board. The
    >components will be minor compared to PCB cost.
    >


    It's been done via S100 to IDE with a IDE to CF. HOwever there was
    only one commercial board that could do IDE and I think thats out of
    production for a while.

    A board to do a CF on S100 will be under 11 chips os SSI TTL. It's
    a trivial task as CF does 8bit mode and that matches the bus nicely.
    Code to drive it will look like IDE (check out the GIDE project).

    One note... CP/M not Zrdos or any of the imporved CP/M replacements
    can address a maximum of 8MB per logical disk and there are a max of
    16 logical disks so you never use the whole CF if larger than 128K...
    IF you run any of the improved CP/M replacements then 1GB is the
    ceiling. However the bios vector for ALLOC is really huge at that
    point (4kbytes!). However CP/M applications tend to be small so a
    32MB or 64MB CF is a big "disk" but still practical!


    Allison

    >-- Charlie Springer



  3. Re: CF for S-100?

    On Wed, 1 Oct 2008 20:27:46 -0700, no.spam@no.uce.bellatlantic.net wrote
    (in article ):

    > It's been done via S100 to IDE with a IDE to CF. HOwever there was
    > only one commercial board that could do IDE and I think thats out of
    > production for a while.
    >
    > A board to do a CF on S100 will be under 11 chips os SSI TTL. It's
    > a trivial task as CF does 8bit mode and that matches the bus nicely.
    > Code to drive it will look like IDE (check out the GIDE project).


    CF has an IDE "mode" which makes it very simple as you say. The drivers are a
    question mark. I'm new to CP/M though I did a lot of Z80 work with ZWorld
    controllers and older embedded Z80's. I was really a RISC and 65C02 guy since
    it did about 4 times as much per clock as an 8086 plus only memory mapped
    I/O. I never liked wasting wafer space with special I/O instructions.

    My point is that I can do the hardware easy but would need some folks to
    partner up on the software side. I don't know how many working S-100 systems
    are out there, but my buddy with the CFFA design is on his sixth batch of 200
    cards and everybody loves them.

    I want to do some stuff with this IMSAI, and CF storage would make it a lot
    easier. In fact, with all the real estate on an S-100 card I would put in two
    sockets so that copies are easy to make.

    Another alternative is an MSDOS compatible file system since I can get the
    whole thing with DOSonChip -- it could be switchable. In fact, DOSonChip can
    use SD or mini/microSD and has a real-time time/date clock built in. Since
    the whole thing is about the size of an SD card, thre is plenty of room so
    might as well add some more features like hardware floating point.

    http://www.dosonchip.com/index.html
    and
    http://www.micromegacorp.com/whatsnew.html
    can do a 64 point FFT 100 times a second.

    One potential pitfall. If CP/M thrashes disks the way MSDOS does, the FLASH
    could wear out from too many writes to directory tables and such. And wasn't
    MSDOS an extension of a West Seattle rip-off of DRDOS? A CF physical block is
    good for over 100,000 writes, but not by much. Reads are unlimited.

    Hey, this sounds like fun. Any reason not to make an S-100 card that is about
    two inches high?

    What else should it have? A header for a USB daughter card (USB stack in a
    PIC or somthing so it looks like simple serial)?

    -- Charlie Springer


  4. Re: CF for S-100?

    Charlie Springer wrote:
    (snip)

    > One potential pitfall. If CP/M thrashes disks the way MSDOS does, the FLASH
    > could wear out from too many writes to directory tables and such. And wasn't
    > MSDOS an extension of a West Seattle rip-off of DRDOS? A CF physical block is
    > good for over 100,000 writes, but not by much. Reads are unlimited.


    I thought CF, or at least the IDE interface to it, spread the
    writes out to avoid that problem.

    -- glen


  5. Re: CF for S-100?

    On Wed, 1 Oct 2008 22:45:10 -0700, Charlie Springer
    wrote:

    >On Wed, 1 Oct 2008 20:27:46 -0700, no.spam@no.uce.bellatlantic.net wrote
    >(in article ):
    >
    >> It's been done via S100 to IDE with a IDE to CF. HOwever there was
    >> only one commercial board that could do IDE and I think thats out of
    >> production for a while.
    >>
    >> A board to do a CF on S100 will be under 11 chips os SSI TTL. It's
    >> a trivial task as CF does 8bit mode and that matches the bus nicely.
    >> Code to drive it will look like IDE (check out the GIDE project).

    >
    >CF has an IDE "mode" which makes it very simple as you say. The drivers are a
    >question mark. I'm new to CP/M though I did a lot of Z80 work with ZWorld
    >controllers and older embedded Z80's. I was really a RISC and 65C02 guy since
    >it did about 4 times as much per clock as an 8086 plus only memory mapped
    >I/O. I never liked wasting wafer space with special I/O instructions.
    >
    >My point is that I can do the hardware easy but would need some folks to
    >partner up on the software side. I don't know how many working S-100 systems
    >are out there, but my buddy with the CFFA design is on his sixth batch of 200
    >cards and everybody loves them.
    >
    >I want to do some stuff with this IMSAI, and CF storage would make it a lot
    >easier. In fact, with all the real estate on an S-100 card I would put in two
    >sockets so that copies are easy to make.
    >
    >Another alternative is an MSDOS compatible file system since I can get the
    >whole thing with DOSonChip -- it could be switchable. In fact, DOSonChip can
    >use SD or mini/microSD and has a real-time time/date clock built in. Since
    >the whole thing is about the size of an SD card, thre is plenty of room so
    >might as well add some more features like hardware floating point.


    Not viable for CP/M as CP/M has it's own on disk format. If the ODS
    is dos fat then CP/M would have to have eitehr some really weird
    interface or it's internal filesystem would have to change.

    >
    >http://www.dosonchip.com/index.html
    >and
    >http://www.micromegacorp.com/whatsnew.html
    >can do a 64 point FFT 100 times a second.
    >
    >One potential pitfall. If CP/M thrashes disks the way MSDOS does, the FLASH
    >could wear out from too many writes to directory tables and such. And wasn't
    >MSDOS an extension of a West Seattle rip-off of DRDOS? A CF physical block is
    >good for over 100,000 writes, but not by much. Reads are unlimited.


    Cpm doen not write as much as dos. It does a lot of reads mostly
    reading the directory . The difference is DOS fat has the allocation
    map on disk and CP/M keep if in ram after reading the extent bsed
    directory.

    >
    >Hey, this sounds like fun. Any reason not to make an S-100 card that is about
    >two inches high?


    Shorter is cheaper. Don't for get S100 uses unregulated voltaged on
    the bus and everything should be buffered but 2" is tall enough for
    all that.

    >
    >What else should it have? A header for a USB daughter card (USB stack in a
    >PIC or somthing so it looks like simple serial)?


    USB theres an idea... there are issue with driver for the various
    devices but for basic known one like keyboard, memory shoudl be fairly
    sane.

    Allison

    >
    >-- Charlie Springer



  6. Re: CF for S-100?

    "Charlie Springer" wrote:

    > One potential pitfall. If CP/M thrashes disks the way MSDOS does, the

    FLASH
    > could wear out from too many writes to directory tables and such. And

    wasn't
    > MSDOS an extension of a West Seattle rip-off of DRDOS? A CF physical block

    is
    > good for over 100,000 writes, but not by much. Reads are unlimited.


    Unlike you, I am not an electronician, but a friend of mine also noticed
    this technical limitation of CF cards.

    Now, all your talk is about CP/M 2.2. Me, I am a fan of CP/M Plus, and CP/M
    Plus reads the directory of a disk, then stores it in RAM, as long as you
    don't change disks. If you do, it recopies the directory on the disk. (Under
    CP/M 2.2, something called RESET does the same thing. You need to type
    Control-C to "mount " a disk.)

    So, why not store the directory of a CF card in a RAM area?

    This way, searching for a file would be super-fast (the reason why CP/M Plus
    does it this way), and the CF card would last longer.

    (Historically, there is nothing, except folklore, obliging you to put the
    directory in the first tracks. In fact, at least one version of CP/M 2.2 had
    its directory IN THE MIDDLE of the floppy. According to the article, it was
    speeding access to files.)

    Yours Sincerely,
    Mr. Emmanuel Roche, France




  7. Re: CF for S-100?


    "Charlie Springer" wrote in message
    news:0001HW.C509096F00184D28F02845B0@news.nw.centu rytel.net...
    > Does anyone make a Compact Flash (or SD card) FLASH Drive for S-100

    systems?
    >

    Hi I have been using a Compact Flash in the form of a Disk-on-Module (DOM)
    with my Interak. It is available from these people www.milesie.co.uk

    this has dynamic wear levelling already built-in - so hopefully will last
    for quite a few years.

    I am at present trying to make use of a Vdrive2 which can use a USB stick
    via RS232 as storage. I haven't got very far yet, it seems to sign-on OK
    but I can't communicate or reset it without switching off the computer. I am
    assuming it could be a handshaking problem.

    The Vdrive may be more useful for S100 cards than CF.
    See one here http://www.vinculum.com/prd_vdrive1.html



  8. Re: CF for S-100?

    > Hi *I have been using a Compact Flash in the form of a Disk-on-Module (DOM)
    > with my Interak. *It is available from these peoplewww.milesie.co.uk
    >
    > this has dynamic wear levelling already built-in - so hopefully will last
    > for quite a few years.


    I'm too - you can simply replace an old Harddisk with an IDE-SSD, look
    at:

    http://www.kc85.susowa.homeftp.net/i...d=42&Itemid=52

    This works with the GIDE without modification, tried this year in
    march. With our customized selfmade BIOS I can boot CP/M from SSD and
    use the CP/M-Filesystem with/at the SSD.

    > I am at present trying to make use of a Vdrive2 which can use a USB stick
    > via RS232 as storage. *I haven't got very far yet, it seems to sign-on OK
    > but I can't communicate or reset it without switching off the computer. Iam
    > assuming it could be a handshaking problem.
    >
    > The Vdrive may be more useful for S100 cards than CF.
    > See one herehttp://www.vinculum.com/prd_vdrive1.html


    This is done with VDIP1/2 and a common USB-Stick this year also:

    http://www.kc85.susowa.homeftp.net/i...=72&Itemid=131

    The VDIP and CP/M "speak" through a special driver and are connected
    with the help of a Z80-PIO.

    The information is in german only but , if someone has time to
    translate the articles, I can put it on my website.

  9. Re: CF for S-100?

    On Oct 2, 12:45*am, Charlie Springer wrote:
    > On Wed, 1 Oct 2008 20:27:46 -0700, no.s...@no.uce.bellatlantic.net wrote
    > (in article ):
    >
    > > It's been done via S100 to IDE with a IDE to CF. *HOwever there was
    > > only one commercial board that could do IDE and I think thats out of
    > > production for a while.

    >
    > > A board to do a CF on S100 will be under 11 chips os SSI TTL. *It's
    > > a trivial task as CF does 8bit mode and that matches the bus nicely.
    > > Code to drive it will look like IDE (check out the GIDE project).

    >
    > CF has an IDE "mode" which makes it very simple as you say. The drivers are a
    > question mark. I'm new to CP/M though I did a lot of Z80 work with ZWorld
    > controllers and older embedded Z80's. I was really a RISC and 65C02 guy since
    > it did about 4 times as much per clock as an 8086 plus only memory mapped
    > I/O. I never liked wasting wafer space with special I/O instructions.
    >
    > My point is that I can do the hardware easy but would need some folks to
    > partner up on the software side. I don't know how many working S-100 systems
    > are out there, but my buddy with the CFFA design is on his sixth batch of200
    > cards and everybody loves them.
    >
    > I want to do some stuff with this IMSAI, and CF storage would make it a lot
    > easier. In fact, with all the real estate on an S-100 card I would put intwo
    > sockets so that copies are easy to make.
    >
    > Another alternative is an MSDOS compatible file system since I can get the
    > whole thing with DOSonChip -- it could be switchable. In fact, DOSonChip can
    > use SD or mini/microSD and has a real-time time/date clock built in. Since
    > the whole thing is about the size of an SD card, thre is plenty of room so
    > might as well add some more features like hardware floating point.
    >
    > http://www.dosonchip.com/index.html
    > andhttp://www.micromegacorp.com/whatsnew.html
    > can do a 64 point FFT 100 times a second.
    >
    > One potential pitfall. If CP/M thrashes disks the way MSDOS does, the FLASH
    > could wear out from too many writes to directory tables and such. And wasn't
    > MSDOS an extension of a West Seattle rip-off of DRDOS? A CF physical block is
    > good for over 100,000 writes, but not by much. Reads are unlimited.
    >
    > Hey, this sounds like fun. Any reason not to make an S-100 card that is about
    > two inches high?
    >
    > What else should it have? A header for a USB daughter card (USB stack in a
    > PIC or somthing so it looks like simple serial)?
    >
    > -- Charlie Springer


    I can think of a few reasons why S-100 cards should remain full
    height... One is easy access to cable headers; two would be easy
    card removal; and three is that it would nice to combine the IDE with
    a hobby prototyping space, one copper pad per hole style as these
    boards are impossible to find now.

    Having made an IDE card + software for my IMSAI already, I suggest
    that you first experience making the CBIOS for it and just getting the
    thing working off floppy... so you can have a feel for what's involved
    for interfacing with drives. Mine does CF with a regular IDE -> CF
    adapter, but I prefer vintage IDE hard drives for the sights and
    sounds aspect.

    JS

  10. Re: CF for S-100?

    On Thu, 2 Oct 2008 12:03:26 -0700, MdntTrain wrote
    (in article
    ):

    > I can think of a few reasons why S-100 cards should remain full
    > height... One is easy access to cable headers; two would be easy
    > card removal; and three is that it would nice to combine the IDE with
    > a hobby prototyping space, one copper pad per hole style as these
    > boards are impossible to find now.
    >
    > Having made an IDE card + software for my IMSAI already, I suggest
    > that you first experience making the CBIOS for it and just getting the
    > thing working off floppy... so you can have a feel for what's involved
    > for interfacing with drives. Mine does CF with a regular IDE -> CF
    > adapter, but I prefer vintage IDE hard drives for the sights and
    > sounds aspect.


    Good advice. I like the prototyping area. I have a half dozen 20 (40?) meg
    HP Kittyhawk IDE drives that can mount on a card if you want to try one. I
    think there are still CF "double think" format hard drives, but probably not
    for long. I think I'll look around and snap up a couple.

    -- Charlie Springer


  11. Re: CF for S-100?


    >This is done with VDIP1/2 and a common USB-Stick this year also:


    >The VDIP and CP/M "speak" through a special driver and are connected
    >with the help of a Z80-PIO.


    >The information is in german only but , if someone has time to
    >translate the articles, I can put it on my website.


    Hi Good to know that someone has got this working with CP/M. The website
    looks really interesting but unfortunately, as I do not speak German, could
    not get much information from it. There are translation sites on the
    internet - which may/may not help.

    Alan




+ Reply to Thread