Conference hydra::amiga_v1

    Hey there ,
    
    While I'm still very busy in several job related things , I couldn't
    put all the HW Projects to this Notesfile as I would do , and it
    looks like this Situation will continue....
    
    But , I've got the C't Driver SW and I will put it on TAPE:: as
    soon as possible ( should be within the next week.. ).
    This SW contains also the Sources and A68k which is PD to build
    or enhance the Driver SW.
    It will also allow you to run this Interface with Seagate ST11R
    and ST11M HD-controllers because OMTIS aren't always available
    anymore.
    I have all the stuff together , but as mentioned before no time
    to assemble...
    Some Amiga friends around here and also in Switzerland are successfully
    running this Cheap-HD-Project.
    So watch out on Tape::Upload for CheapHD.Zoo
    
    regards , Ingo
    

    
    
    Cheers Ingo,
    
    		I've also found a load of source material regarding
    Hard Disk Interface Software..Driver Formatter..Interleave tester
    etc..
    
    	It's all in HDRRIV.ARC on TAPE:: in the arc directory I think.
    
    	There is also another Hardware Interface There in the Description.
    Your option still has the edge though..
    
    		Best wishes
    
    			Pete..
    
    

re: -.1

>    	It's all in HDRRIV.ARC on TAPE:: in the arc directory I think.
    
re: -.2
	CHEAPHD.ZOO

	Does anyone have an exact location for these files ? I've
searched through TAPE NORSE and WJG and can't find either of them.

					Alan T.	
    
    

They're on WJG::AMIGA: now.

Couldn't find cheaphd.arc.

John

    The HDDRIV.ARC  is another HD interface and software.
    
    Not to be confused with Ingo's offering, I merely suggested it as
    a reference to what could be made .
    
    	Ingo said that he would upload it within the next two weeks.
    So keep looking...
    
    
    			Pete
    

    
    Cheaphd.zoo is now on tape::upload
    
    		Good Luck
    
    PS I didn't put it there
    
    		Pete

    HMMMMMMMM,
    
    has anyone of the German Amiga-Users the time to translate the .-1
    into English.
    I would like to do it, but i have NO time and i promised Dave Wecker
    to translate a LARGE text for him.

    SO, YOU UNKNOWN GERMAN AMIGA-NOTES-READER WITH SOME TIME, THIS IS
    YOUR CHANCE TO ENTER THE  A C T I V E  AMIGA-COMMUNITY   8-)
    
    Ralf
    ====
    

    Below you will find the requested translation of the German "READ.ME"
    file. I hope this will help you in using the software.
    
-----------------------------------------------------------------------------   
    
     Dear reader,

On this disk you can find the programs, which you need for the installation 
and operation of a harddisk with an OMTI-controler on the Amiga 500 (or 
Amiga 1000). The interfaceprint you can get from eMedia, Hannover.

The disk contains a driver for OMTI- and Seagate-controler (OMTI 5520 and 5527,
ST 11R and ST 11M). There is only one driver, called "harddisk.device". 
The data of the connected harddisk (sectors per track, number of tracks and 
heads) are stored at track zero, when the disk is physical formated. The driver
calculates this data when the drive is initialized.

In the 'c'- directory you can find the BASIC-program "wformat.bas", this 
program can be used for the physical formating of the harddisk. It asks for
all the important parameters for formatting. Besides that it will write the
already mentioned diskparameters on track zero. If you don't want to do a
physical format, but only want to change the diskparameters, you can use
the program "wformat-test.bas" instead. This program doesn't FORMAT, it only
asks for the important diskparameters and writes them to track zero.

 The drivers are tested by the editors. Errors are not known uptill now, however
no garantuee can given for the program to be errorfree.

 The disk is organised as follows:

                          c't-Treiber
                               |
         +---------------------+--------------------+
         |                     |                    |
  fertige_Treiber      Programm-Texte         Public_Domain
 (Running driver)

In the directory "fertige_Treiber" you find both the harddisk-driver and a
park-program. The park-program however can only be used for a ST238 disk from
Seagate. The C-source-text can be found in the "Programm-Texte" directory.
The harddiskdriver should be copied to the "DEVS:"-directory.

In the directory "Programm-Texte" you can find in the sub-directories the
sourcelistings. The directory "beispiel_mountlists" contains two Mountlists.
One of these lists should be copied as "mountlist" in the "DEVS:" directory
(see article). 

NOTE of translator: The file CHEAPHD.ZOO doesn't contains the sources and the
mountlists. 

"Public_Domain" contains the assembler and linker, which can be used for 
making the drivers. For assembling the programs you need however the
INCLUDE-files (which can be found in the System-documentation of 'Verlag 
Addision-Wesley) and the "amiga.lib" library. Both are missing because
these are not Public Domain!.

Unfortunatly some bugs occured in the first versions of the software, however
these are solved. Please read also the tips "Erganzungen + Berichtigungen"
in the 6/88 and 8/88 editions and the article "Floppy, nein danke" in the
c't 2/89 edition.

All further information you should get out the documentation from your
harddisk and controler. When still some open questions remain, we are
available at the normal Hotline channel. Tel. nr. 05 11/ 5 47 47 -10

    I have just completed the construction using a seagate ST11m. No
    problems in doing the initial format (wformat.bas) Mounted the drive
    and then tried unsuccesfully to format under Amigados and FFS. got a
    message....."unable to find Handler". What's missing? Harddisk.device
    is in the devs directory. Fastfilesystem is in the L: dir.
    
    I'm using WB1.3 and ARP1.3
    
    Is there  <<another>> driver?? Any suggestions???
    glenn

I had the same error for a different drive.  

The solution was to format the drive using the OLD filesystem first, then
again with the FFS. This means edit the mountlist for the old  file system,
formatting the whole drive, edit the mountlist againg for FFS and format again.


John

Re: .10, .11

Don't bother trying changing the mountlist to use the old file system
and formatting the drive, instead try the INHIBIT option on FORMAT:

	FORMAT DRIVE dh0: name "hard" FFS INHIBIT

That can avoid certain bizarre problems where trash on the drive
misleads the filesystem.

I don't guarantee that this will solve the problem, but it is easy
to try.

    RE: 11 & 12
    
    Thanx for the advice but no luck. The sequence of events was:
    
    1. install harddisk.device in the devs: directory
    2. create the following mountlist
    	
    	DH0:	Device = harddisk.device
    		Unit = 1
    		Flags = 0
    		Surfaces = 8
    		BlocksPerTrack = 17
    		Reserved = 2
    		Interleave = 4
    		Lowcyl = 0
    		HighCyl = 510
    		Buffers = 30
    		BuffMemType = 5
    	#
    
    This could be all wrong, but it's as described in 2328  and I dont have
    anything else to go on. ( the disk is a Quantum 540 - 31Meg)
    
    3. Mount DH0:
    
    4. Binddrivers
    
    5. Format DRIVE DH0: NAME "HARDON" NOFFS
    
    THEN I GET.......
    
    "Format failed
     Cannot find Handler"
    
    Note that the Harddisk.device is only 3200 bytes long (seems awfully
    short for a disk driver(?) Are Drivers and Handlers synonymous?? do I
    need anything in th expansion directory?
    
    Are there any additional ideas out there?
    
    
    Glenn.

    	DH0:	Device = harddisk.device
    		Unit = 1
    		Flags = 0
    		Surfaces = 8
    		BlocksPerTrack = 17
    		Reserved = 2
    		Interleave = 4
    		Lowcyl = 0
    		HighCyl = 510
    		Buffers = 30
    		BuffMemType = 5
    	#      

You need LowCyl=2 if you have reserved=2

You need dostype in ther as well.  I'm used to at least 2 mountlist entries: 
1 for PREP and one (or more) for the drive partitions. You then need to PREP
the drive first.


Ex:

RES0: Device = hddisk.device
      Unit   = 1
      Flags  = 0
      Surfaces  = 8
      BlocksPerTrack = 17
      Reserved = 2
      LowCyl = 0  ;  HighCyl = 1
      Buffers = 1
      BufMemType = 0
#

FH0:  Device = hddisk.device
      Unit = 1
      Flags = 0
      Surfaces = 8
      BlocksPerTrack = 17
      FileSystem = L:FastFileSystem
      Reserved = 2
      LowCyl = 2  ;  HighCyl = 510
      Buffers = 30
      GlobVec = -1
      Mount = 1
      BufMemType = 0
      Stacksize = 5000
      DosType = 0x444F5301
#

Then mount and PREP RES0:, mount and FORMAT FH0:.

If this doesn't work. Delete the FileSystem line and change DosType to
0x444F5300 in the FH0: entry. Format it NOFFS, change it back the way
it is and format it FFS.

John

    re: .14
    
    John, thanks for the help but it didn't work. I've tried every possible
    permutation of mountlist parameters, even purposely creating mismatches
    to see the result,i.e Dev=harddisk.device and naming the driver
    something else. results are predictable. I either cant get the thing to
    mount or I get a visit from the Guru.
    
    I'm starting to think that my solution is in one of the following areas:
    
    1.	The original s/w lived in a 1.2 environment. Maybe I've got an
    interaction between the s/w and WB1.3.
    
    2.	I'm using ARP 1.3. Possibility of interaction prob??
    
    3.  The driver is corrupt.
    
    I lean towards scenario #3 as I guess that I'm the only one around DEC
    who's built this thing (or at least I'm the only one who'll admit to
    it)
    
    Ingo, If you are listening..... Would you please confirm with your
    friends (who have succesfully built the project) that they are using
    the S/W distribution that you uploaded to TAPE::? Whoever did the
    english translation of the readme file observed that there was not a
    mountlist in the distribution. Was this an oversight or intended? I
    would hate to think that the mountlist expected something really weird
    as a device name.......
      
    Glenn

    re: .14 (p.s.)
    
    I forgot the following....
    
<You need LowCyl=2 if you have reserved=2>
    
    John, not sure about this. An article on Amigaworld (jan89) says that
    the "reserved" parameter specifies the number of blocks used for
    booting. Do I detect that you meant Cylinders? My reading suggests that
    even with reserved=2 LowCyl should/could still be 0.
    
<You need dostype in ther as well.  I'm used to at least 2 mountlist entries: 
1 for PREP and one (or more) for the drive partitions. You then need to PREP
the drive first.>
    
    Sorry, but I dont have "prep".
    
    Glenn.
    

> <You need LowCyl=2 if you have reserved=2>
    
>    John, not sure about this. An article on Amigaworld (jan89) says that


AmigaWhirl strikes again. That's cylinders, although I am not very familiar
with this HD project, you may only need them (Reserved =2 and LowCyl=2) for
PREPable systems.  <CB>? Ingo? I think in any case you should make them
consistant (both zero or both 2).


<digression follows>
I have seen so many errors in Amiga World publications that I don't even
subscribe to it anymore. Besides, it's one of those magazines you can read
cover to cover *at the newsstand* :-) No offense intended! It's just that
about the only more hilarious computer related information I can think of
is found watching a host sell a PC on the Home Shopping Club (or equivilent!
Tonight, this lady tried to sell a Motorola 286 PC by saying things like

	This system has a 20MB hard disk (she stressed this well) unlike
	most other systems which only have 256K (oh, she clarified this too:
	"That's 256 hundred thousand")..

Sorry. it's late.

John

    Maybe the following can help you. Below you will find the example
    Mountlists as mentioned in the German article.
    

    ----------------------------------------------------------------------------
    DH0: Device = harddisk.device
         Unit = 1         (should be always 1)
         Flags = 0
         Surface = 4
         BlocksPerTrack = 26 (17 for MFM and 26 for RLL controlers)
         Reserved = 2
         Interleave = 0
         LowCyl = 0
         HighCyl = 400
         Buffers = 10
         BufMemType = 5
         
    #
    
    DH1: Device = harddisk.device
         FileSystem = L:FastFileSystem
         Unit = 1
         Flags = 0
         Surfaces = 4
         BlocksPerTrack = 26 (17 for MFM and 26 for RLL Controlers)
         MaxTransfer = 256
         Reserved = 2
         Interleave = 0
         LowCyl = 401
         HighCyl = 612
         Buffers = 10
         GlobVec = -1
         BufMemType = 5
    
    #
    ----------------------------------------------------------------------------
Three remarks are made in the article that may be important to mention.
    
    - BufMemType on 3 (For CHIP_MEM) or 5 (For FAST_MEM)
    
    - When the drive is formatted with wformat.bas, the article gives
      the following information.
    
      The BASIC-program asks at the beginning for the Unit that you
      want to format. Here you should answer always "0", since the 
      harddisk.driver supports uptill now only one unit 0. After that
      the program asks for the number of Cylinders,heads,sectors/track,
      the interleave-factor and the steprate. The steprate should be
      between zero and seven. The steprate can be taken from the following
      table:             
      nr        Steprate
    ---------------------
       0        3 msec
       1        10 usec, buffered step
       2        25 usec, buffered step
       3        50 usec, buffered step
       4        200 usec, buffered step
       5        70 usec, buffered step
       6        3 msec
       7        3 msec
    
    When the drive has error sectors, they must be entered. In this
    case the wformat.bas program asks in which order cylinders, head
    and byte of the error-spots should be entered and whether this will
    be done manually. Is the drive error-free, return should be pressed,
    untill the question "Bereit zum formatteren (J/N)?" appears.
    
    When after the reply "J" you hear a soft "schnurren" and a regular
    "klacken" (I don't know the good translation, but the words indicate
    the sound well I think) the interface works fine.
    
    Very important is the last message of the program. It contains the
    number of the last cylinder that can be used. All cylinder with
    a higher number are for sparetracks. When later in the mountlist
    a higher number is used then the by the program mentioned, the DOS
    command 'format' stops at the track with the next higher number,
    without any command.
    
    Before the DOS command 'format' can be used, the drive has to be
    mounted first.

    ----------------------------------------------------------------------------
    The above is a translation of a part of the German article, I hope
    this will help you out. Keep us informed.

>      the interleave-factor and the steprate. The steprate should be
>      between zero and seven. The steprate can be taken from the following
>      table:             
>      nr        Steprate
>    ---------------------
>       0        3 msec
>       1        10 usec, buffered step
>       2        25 usec, buffered step
>       3        50 usec, buffered step
>       4        200 usec, buffered step
>       5        70 usec, buffered step
>       6        3 msec
>       7        3 msec
>    


FYI - Most modern ST506 drives will require a buffered seek to meet their
specified access time. If you don;t have a drive spec, try either
 2 (25usec) or 1 (10usec).

Otherwise keep increasing the step rate till it works!

RD53,RD54,RD32 all get a 27usec step from the RQDX3.

John

    
    As far as Reserved vs./with LowCyl is concerned, hear the words
    of the wise nerd from the east... ;-)
    
    Reserved indicates the number of BLOCKS that will be resevered at
    the start of a partition to hold any kind of bootstrap code (or
    viruses :-( ). This is currently 2, from which the standard CBM
    bootblock uses about 20-40 bytes.
    The LowCyl does just it seams to indicate, define the lowest physical
    cylinder of a partition.
    
    As far as the real problem here is concerned, I'd also say that
    the software is to blame.
    But the vague memory of the instruction hints that maybe the LowCyl
    shouldn't be 0.
    
    Regards,
    
    <CB>

Sorry about that. But I still don't completely trust AW :-)


John

    Wow!!!....what can I say except that the response has been
    overwhelming. Thanks for all of the help and advice. However, I'm still
    stuck.
    
    For the benefit of new readers, I had no trouble with the "WFORMAT.BAS"
    part. The thing just clicked away, counting up cyls  and heads until it
    got to the end, telling me that the last usable area was cyl 510, head
    7. Interestingly, the program never got into the subject of steprates,
    so I can only assume that it takes care of that when one selects from
    the menu of drives. This should mean that the interface and drive
    works..doesn't it???
    
    
    as an update, I tried the process under 1.2......same result. More than
    ever, I am now nearly convinced that the Driver(?) - "harddisk.device" is a
    heap of s@#$!!. Methinks that when we "mount" a device the S/W only
    goes looking for the presence of a driver, NOT checking if it works.
    
    One could logically expect that I would get all sorts of different
    errors if my mountlist were in error, but I always come back to the
    "cannot find handler" message under "SYS:SYSTEM/FORMAT"
    
     I have written to INGO via All in one to ask if he can check with the
    people who have got this working, with a view to making certain that
    all of th S/W was uploaded AND that they used the same versions.
    
    As a last resort I have thought about dis-assembling the "driver" and
    seeing if it really does contain code, but then I wouldn't know what to
    look for.
    
    I'll certainly double check all of the things that you folks have
    advised on and keep you posted.
    
    Cheers/ Glenn

    I have an A1000 and a Microbotics Stardrive SCSI interface, and
    it cannot auto-configure.  I have to put the device driver in the
    Expansion drawer and then add appropriate entries to Mountlist,
    then in startup just Binddrivers and mount.  By the way, the first
    time I got "Not a DOS disk", but all I had to do was format.  I
    dont have any device driver in the DEVS drawer, only Expansion.
     You might want to give it a try.  Who knows????
    

    
    Ok let's try this:
    
    A fellow Amigoid who started to work recently at this location has
    been using the C't interface for quite a while now. I'll grab all
    the related software from him and upload it ASAP.
    
    That should do the trick...
    
    Regards,
    
    <CB>

    
    I just uploaded a working "harddisk.device", which is about 15 KB
    in size (so you were right about yours...) to:
    
    FRAMBO::BZOEVFL:[BALZER.WORLD]harddisk.device
    
    I'll zoo the whole disk over the weekend and make it available on
    Monday. This should get ya going.
    
    Regards,
    
    <CB>

    Thanks <CB>, but still no joy. Even though the Harddisk.device was 15K
    long, approx 13.5 Kb was Zeros. Still got the "format failed...cannot
    find handler.."
    
    As an experiment, I borrowed another guy's boot disk with a commercial
    driver on board. Of course it did not work...BUT the formatter did
    "run", ie, ...."formatting cylinder O.... verifying Cylinder zero.."etc
    The downside was that between each of these messages was a two minute
    delay. It would have taken fifteeen hours to run.
    
    I guess my salvation lies with the complete upload, or, if you could
    ask your friend to take a look at the uploaded driver.
    
    Thanks/ Glenn.

    Re: .26
    
    Well, just in this very moment the upload of CT-HD.zoo is underway
    to "FRAMBO::BZOEVFL:[BALZER.WORLD]".
    
    All I can say it that this stuff works for Oliver, the fellow I
    got it from. I'll ask him to participate in this conference...
    
    Please note that the CT-HD.zoo file does NOT replace Ingo's file,
    it merely is an addition. I would ask Glenn to generate a "final"
    version of this, when/if he get's the sucker going.
    
    (Still) Hope that helps, ;-)
    
    <CB>
    

    Looking into the source code for the harddisk40.i it is clear you
    have to change the diskparameters according your own harddisk. 
    However this source is an older version.
    Some history, the German magazine published the first article in
    april '88 and the second article in feb. '89 which made some changes
    both in the hardware and software. The source is the older version
    and the hardware taht Ingo descripted is the latest.
    
    The differences:
    
     The adress for the OMTI controler has been to from 810640h to EF0640h
    to enable working with a 68020 processor. So you have to change
    this adress in the source code accordingly.
    Also a major bug has been solved. The article mentions a bug in
    the sourcecode of HD01.i, in the TRACK buffer routine the line
    move.l #MEMF_PUBLIC,d1|MEMF_CLEAR has to be changed to
    move.l #MEMF_PUBLIC|MEMF_CLEAR,d1
    
    This bug was causing that the softwaredriver was not correct mounted
    after a reset.
    
    I hope this additional information will help.

Nearly there!.....

I did the assembly last night and the assembler found some errors on pass 2
which I made some assumptions about.

1.
"File exec/io.i Line 22 ERR 44: Not enough Macro parameters supplied"

The contents of the line were "LIBINIT"

I changed this to "LIBINIT 0"         (assumption!)

2.
"File xtra.i/harddisk40.i Line 12 ERR 44: Not enough Macro parameters supplied"

The contents of the line were "DEVINIT"

I changed this to "DEVINIT 0"         (assumption!)

3
"File xtra.i/harddisk40.i Line 124 ERR 1: Error in expression value"

the contents of this line were  "STEPRATE   EQU %011 ; 10uSec"
                                
I changed  this to "STEPRATE   EQU 031 ; 30uSec"   (assumption!)

I assumed that if 011 (10+1) = 10 uSec, then 30uSec = 031 (30+1)
This could be totally crazy but I was told that the steprate on the
Q540/RD52 was about 30Usec. However, I might have set the Steprate to Five
years!!!  Would any learned readers care to comment?

Also, 
1.	Jos mentioned a patch to HD01.i. Do you mean HD01.ASM ?

2.	Change the OMTI addr to $EF0640. In the Basic formatter (which works
OK), the addr is $EF0641. Which is correct? or are both?


Note that after incorporating my assumptions, the assmbler went through without
errors. I then "Blinked" it and installed it in the DEVS directory. Voila, the
formatter program started, saying "formatting cyl 0, 510 to go". It never got
any further.

It looks like some progress has been made.

Please be assured that, when I get this thing going, I'll put everything in a
note for the benefit of those that follow.

Glenn

                                

    I forgot.
    
    I accidentally creamed my copy of Cheaphd.zoo.
    
    I get a protection violation when I try to copy from
    tape::user2:[amiga.harware]cheaphd.zoo
    
    Help
    
    Glenn

    I serched this conference I figure this is an appropriate place for my
    question.  I too have built a hard disk interface for my A100; well,
    sort of.  What I built is a 5 slot expansion backplane for the A1000
    accordding to Commodore's specification.  I think it came out of the
    A1000 technical reference, I'll check if anyone cares.  Anyway, I
    plugged my newly acquired used 2090 controller into it and it doesn't
    work.  The interesting thing is that I'm not yet sure whats not
    working: the backplane or the controller (or both).  
    
    The symptoms are kind of wierd.  With either a SCSI or ST-506 disk,
    when I run prep, he program just hangs after the answer to the last
    (according to the docs) question - "..this could kill you, do you want
    to continue?" - which is supposed to return in a few seconds "or
    minutes" with "now reboot and format."  Stranger still is that the
    prep won't run past the banner after a reboot from this hung condition
    unless I do something to change the disk's identity, such as changing
    the unit number jumper or powering the drive off untill after the
    workbench is running.  On the ST-506 drive its worse: it won't boot
    past the bindrivers command, never mind execute prep.  
    
    What am I doing?  Well, I have the correct data on the ST-506
    interfaced ST-225 drive (I think): 4 heads, 17 sectors/cylinder, 615
    cylinders.  For the SCSI drive I don't have this information.  I
    assumed that the real number of surfaces, cylinders, etc., wouldn't
    matter as long as the number of blocks multiplied out correctly, or
    even if the numbers supplied resulted in a smaller number of block
    represented than actual.  Does anyone know if this is true?  Anyway,
    I'm using the original distribution disk that came with the 2090 which
    has Prep V 33.19, with Kickstart and Workbench V 1.3, 512K A1000.
    
    Any ideas?  I've spent many hours on this one, I'd like to get it
    running. 
    
    Thanks much,
    
    John
    

    With the CLI command INFO, you should get the "not a dos disk" message
    for DH0: ( the ST-506 drive).  At that point you can now "prep" DH0,
    but you should have a RES0: entry in your mountlist.  PREP will ask you
    the size of the first partition on your hard drive, make it about 30
    cylinders this will make it about 1 meg.  Then you can format DH0, it
    will be OFS (Old File System).  If this works at this point, I can
    prompt you for the rest.
    
    Jean
    

    Oh wait a minute Jean, I think you've got something there... Although
    the mount command seems to work, I never see the drive with the Info
    command (I tried).  I think I may have backplane hardware problems, and
    I'm tracking down some leads now.  Should I really expect to see a "not
    a dos disk" with "Inof" immediately after mounting it?
    
    Thanks,
    
    Joh

    In fact, you should have a blank icon on your WB screen for Dh0: with a
    NDos name attached to it.  This is because the A2090 will automount the
    first partition of each drive attached to it.  As long as your hardware
    is accessing the controller correctly, and the drive is correctly
    cabled what I described should happen.  
    
    Did you check the controller and drive on a 2000 to make sure those
    parts work correctly?  doing that can avoid a lot of problems.
    
    Jean
    
    

    O.K., I need a volunteer with a 2000 to check out my controller/drive
    combo.  Could someone nearby please help?  I work in the mill in
    Maynard, live in Wakefield, Ma.  
    
    Also, I know this sounds stupid at this poin, but what SCSI i.d. should
    the first SCSI drive have?  I can't find this in the 2090
    documentation.  What SCSI i.d. does the controller assume?  
    
    Thanks,
    John

>    Also, I know this sounds stupid at this poin, but what SCSI i.d. should
>    the first SCSI drive have?  I can't find this in the 2090
>    documentation.  What SCSI i.d. does the controller assume?  
 

For the A2090a :

Device		1st 2 cyls	Unit*	cyl 2-? (1st partition)
-------		----------	------	------------------------
1st ST-506	RES0:		 1	 DH0:
2nd ST-506	RES1:		 2	 DH1:
1st SCSI 	RES2:		 3	 DH2:
2nd SCSI 	RES3:		 4	 DH3:
3rd SCSI 	RES4:		 5	 DH4:
4th SCSI 	RES5:		 6	 DH5:
5th SCSI 	RES6:		 7	 DH6:
6th SCSI 	RES7:		 8	 DH7:
7th SCSI 	RES8:		 9	 DH8:


* A unit number of 0 is not allowed.

John

    thanks John, but I don't think that's it.  I saw that table in the 2090
    manual, but it doesn't indicate the "Amiga Unit Number" to SCSI bus
    unit number; e.g., SCSI unit number range from 0 to 7, not 3 to 9.  The
    problem becomes guessing perhaps) if the 2090's SCSI port is SCSI unit
    0 or 7, making the first SCSI drive SCSI unit number 0 or 1, or
    something like that.
    
    Thanks,
    
    Joh

Huh?

For the first SCSI drive, set it's PHYSICAL ID to 0 (all jumpers out).

The software then calls it RES2: for prep, and unit=3 for mountlist.
The first partition is then automounted as DH2:


This is what that table says, right? Or am I missing the point?


John

Oops, just re-read your message.

Looks like the table does not specify scsi ID's

The manual I have does state that for SCSI devices 1-7 (first through seventh)
set the SCSI ID to 0-6 respectively.

John

I know this is standard stuff, but just LOOK at the crazyness of this...


For the first SCSI drive, set it's PHYSICAL ID to 0 (all jumpers out).
        ^^^^^                                     ^
The software then calls it RES2: for prep, and unit=3 for mountlist.
                              ^                     ^
The first partition is then automounted as DH2:
                                             ^

This is a classic! Just imagine if installing ignition wires on
an engine involved attaching one end of the first wire, wire 0, to sparkplug
1, which fires on cylinder 2, and attaching the other end to hole 3 on the
distributor...

Oh well... %-)

re .-1

I agree!

And I thought they had a pretty reasonable solution to the problem (how to
get 2 ST-506 and 6 SCSI driver under one roof).

There are better ways, but this one works... Such is the philosophy


John

    
    
    
>The manual I have does state that for SCSI devices 1-7 (first through seventh)
>set the SCSI ID to 0-6 respectively.
    
    Got it, thanks!  Yet another variable filled in on the debug trail.
    
    John

    Update:  tried SCSI drive as ID 0, did get it to step with Prep, but
    Prep hangs machine eventually; Workbench disappears, etc.  Probably
    have hardware problem with backplane.
    
    John