• Casting call! WD336RT drive PCBs!

    From Louis Ohland@21:1/5 to Louis Ohland on Fri Jun 17 06:29:11 2022
    Hunting down the elusive WD336RT, used in the 8550 and 55SX [I
    think...]. I want to see tool p0rn of the controller chips on the
    WD336RT PCB. My expectation is that the 8550-ST506 controller uses a
    chipset that incorporates more functionality into the ASICs, but still
    does the same as the WD1006V-MC1 controller.

    Perhaps the WD336RT was identified as some form of ESDI, but it isn't.

    I'm not sure, could an ESDI controller drive a ST506/ST412? I think it's possumble.

    Note, the WD336RT is undocumented, my SWAG is that it's a OEM for IBM.
    They might have had stock left over from the 8550 systems... who can tell?

    http://vtda.org/books/Computing/Hardware/pc_engineers_vol3_misc.pdf

    page 94

    page 29, Abbreviations
    Type "M" MFM, with ST506/ST412 usually 17 sectors per track

    6128287 30MB, 615/4/17
    64F4146 30MB, 615/4/17

    On 6/16/2022 07:26, Louis Ohland wrote:

    Drive A Cable 27F4912
    Drive B Cable 34F0000

    ???

    30MB Hard Disk Drive (DBA) 64F4146

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to Larry Fosdick on Fri Jun 17 06:43:39 2022
    https://groups.google.com/g/comp.sys.tandy/c/c8brrPNjZEs/m/XlCAr8P3BjAJ

    Bill Vermillion
    Mar 13, 2003, 11:25:35 PM
    In article <WtPba.68256$[email protected]>,
    Larry Fosdick <[email protected]> wrote:

    I am contemplating replacing the 5 MB disk in my TRS-80 HD
    with a 31 MB disk. I looked at the information on Ira's site
    concerning this, and came up with a question.

    Are the ST506 and ST412 specs different and incompatible?

    Compatible to a point.

    I've seen these two specs used together, separately, and in a
    few documents, interchangeably. The bubble repair document on
    Ira's site uses ST506, but the Seagate ST-4038 is specified
    as ST412 MFM, not RLL). The drive seems to meet all the other
    requirements for the old (large) controller, in that it has 5
    heads and 733 cylinders.

    Can anyone shed light on this?

    Sure can.

    All the early drives were pretty dumb and the formatting and setup
    utilities did all the work. In modern drives the drives take care
    of this.

    In the ST506 you are limited to 70MB and heads 0 thru 7.
    You also set the value for the track on which write-precompensation
    occurs.

    Write-pre comp is when you modify the signal to take care of
    bit-shifting that occurs on the inner tracks were one bit can
    affect an adjacent bit. Since identical magnetic forces repel
    the bits want to shift away - and the pre-comp change the
    timing/spacing so this doesn't occur. That's also something you
    see in coated media that is relatively soft - which was the norm
    for the old drives.

    In the ST506 there is a line that is toggled when it is time
    to use write-precompensation.

    In the ST412 drives they handle write-precompensation internally.
    The line that was used for write-precomp now becomes a head-select
    line.

    Normally you select heads 0 thru 7, but on the ST412 when you
    toggle the old write-precomp - now head select, when you select 0
    it is head 8, and 1 is head 9 and so on until you get 15 heads.

    So you can used ST412 drives on an ST506 interface >>IF<< you only
    format the first 8 tracks - heads 0 thru 7.

    We used to put the ST4096 Seagate in the Model 16's to get 70MB.
    The ST0496 is a 77MB drive with 9 heads. But if you tried to
    format 9 heads using the standard controller - which used
    the WD1010 controller chip that only knew about 8 heads - once
    you went to head 8 [the ninth head] and since that was
    write-precomp, you actually wrote over track 0. That is not what
    you want to do.

    To sum up:
    ST506/ST412 is seen quite often. Pure ST605 interfaces were
    used only in the early days. They are virtually identical except
    for the write-precomp line becoming head select. You can use
    ST412 drives in ST506 but DO NOT format for more than 8 heads.
    You can also set write-precomp to the maxium number of cylinders
    when using an ST412 on an ST506 controller - because the ST412
    handles write-precomp intenally.

    You used to have to know all this stuff if you used other than the
    drives speced by the factory. The old 16's had a Tandem drive that
    would fail in strange ways and would lose on head or the access to
    one. I'd take the standard ST225 - the 1/2 height 5.25" and put
    those in the 16 - or if I just needed to load the data from the
    Bermoulli backups so I could xfer the data from the 16, I'd just
    route the cable out the back and have the 225 on the table for the
    time it took to transfer.

    Bill
    --
    Bill Vermillion - bv @ wjv . com

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to All on Fri Jun 17 07:10:40 2022
  • From Louis Ohland@21:1/5 to All on Fri Jun 17 07:51:23 2022
    The ACB4070 boards I have talk RLL not MFM, according to the manual....
    You can easily tell because the MFM varieties have a crystal which is a harmonic of 5 MHz while the RLL boards use a harmonic of 7.5 MHz.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to All on Fri Jun 17 07:29:48 2022
    http://minuszerodegrees.net/hdd/st506_st412_interface.htm

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to All on Fri Jun 17 08:15:29 2022
    https://archive.org/details/pc_engineers_vol3_misc/page/n21/mode/2up?q=%22western+digital%22+mfm&view=theater

    RLL has other advantages. A drive producing 20 Mb under MFM would give
    30 Mb using RLL, with the original 20 Mb squeezed down, occupying only
    two thirds of the disk, so the heads don't have to move so far to reach
    the same data, effectively reducing the average seek time. RLL drives
    were actually from MFM production lines, but tested more rigorously.
    Those that failed were marked up as MFM and sold accordingly, which is
    why some drives in the tables below are marked up as both, i.e. M/R.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)