Conference hydra::amiga_v1

    I have RD equivalent drives up.  What system (a1000/2000/500) are
    you using, what controller (Supra,C=,C-Ltd,etc), What version/type
    of drivers/formatters?
    
    You could also read note titled HARDDRIVE MADDNESS for some additional
    info...
    
    -john-
    

rep .-1
	I have an A2000 using an A2090 controller. The driver/formatter
etc are what comes with the A2090. I am having problems with the PREP
program, used to initialize the first two cylinders of the drive. I
choose option 0, which is user defined and feed it what I believe are
the correct values for an RD52. PREP diddles with the disk for about
10 seconds, then either hangs the system (nice), or returns the ever
so helpful message, "Device not PREPed". The great documentation
supplied with the A2090 assumes everything will work just fine, since
it tells you nothing about errors. Off the top of my head, I can't
remember the version number of PREP.

Frank

    re .-1
    
    An additional question I have is: What are the parameters for an
    ST-225 (RD31??). I just borrowed a drive without any manuals or
    specs and I would like to try it. So if any body has this data
    PLEASEEEEE drop a line
    
    thanx
    
    ... george...

RD31:

	0-614 cylinders
	0-3 heads
	precomp at cyl 256
	NO reduced write current
	park heads at 670
	drive will auto-recal at 671 (used by DEC for auto-size)

	cyl 614 has whats know as an FCT (Factory Control Table)
	so you might want to use this info - ie don't format over it.
	This is a known-defect map and corresponds to whats on
	the label on the drive. If you want the format of this table,
 	let me know - I'll xerox the relevent pages from the RD31
	spec.


John (who writes the software which evaluates the RD drives for DEC)

    re: .0
    
    Sorry, I can't help on the 2090 controller.  I have Phoenix and
    C-Ltd controllers on an A500.  How about a call right to CBM???
    
    -john-
    

    RE: RD53
    
    Heads:   8
    Cylinders: 1024
    Sectors: 16,17  (18 sectors required reduction of num of accessable
                                                                   cyl)
    
    RE: RD31 (ST225)
    
    Formatting for 
    4 Hds
    17 sectors
    0-611 Cyl   will give you approx 20 meg formatted and keep the
                factory BAT block intact.
    
        
    -john-

.4

   What is precomp?
   Also is it a good idea to have the heads automatically parked?

Frank

>   What is precomp?

Precomp (or pre-compensation) is a technique which is becomming  unnecesary
with newer drives. The layout of a typical hard disk (and floppies) is by
concentric rings, called cylinders. Each cylinder has subdivisions called
sectors in it.  Typically, for simplicity of design and mapping, all cylinders
have the same number of sectors (for DEC RD drives this is 16 or 17 depending
which controller you use - PRO/Rainbow/ RQDX1 and RQDX2 have 16, RQDX3 has 17
"normal" sectors and an 18th one is used for primary bad block replacement
(BBR)) This pattern is repeated for each head on the drive (neglecting a
dedicated servo surface for the sake of simplicity).

Now picture a pie (just like mom used to make!). Its round like disk platters.
When you slice it up, you create pieces of pie (sectors). A strange phenomena
of pieces of pie is that thier width (if you will) gets smaller and smaller as
you move toward the center of the pie. Well, on disk's this same thing happens
- as you move toward the center, the sectors (a slice of pie) get closer and
closer together. Note that this analogy of sectors to slices of pie si not
totally correct. A piece of pie is a single entity (until the fork hits,
anyway). On a disk, there is a sector in each cylinder so now the piece of
pie is sliced into cylinders.

So one can now see that individual bits on inner cylinders are packed closer
together than are the bits on outer cylinders. Each bit on a disk can be
thought of as a tiny bar magnet. We've all played with magnets and found that
if you put two N poles (or S poles) together, they repel and an N and S
attract. So at inner cylinders where bits are close together, they tend to push
eachother around. What precomp does is pack them *even closer* BEFORE it 
writes them on the disk so that when they finally settle thier terratorial
disputes, they end up in the right place.

This is a pretty simplistic view, but is generally what happens. Again not
all drives require precomp (most new ones don't). Putting precomp on a
drive that does NOT require it does not hurt (doesn't help either). NOT
putting precomp on a drive which DOES require it has the *potential* to cause
problems (won't always).


Does this help?


>   Also is it a good idea to have the heads automatically parked?

YES. "Parking" the heads assures that the heads will touch down in a dedicated
"landing zone" and not on your data. Not all drives require the user to park
the heads manually. Most will use the power generated by the spindle motor,
(it's really a generator at this point) as it slows to a stop, to drive the
heads to the landing zone.



BTW - RD53 has just reached over 100,000 hours MTBF !!! (MTBF is continually
being proven by all products measured in this manner)


Following is an exerp from a program we (RD buyouts) send to the drive vendors
to verify the FCT (defect map). It's the drive parameter table.

(I've left out RD33 since its not a formal product yet (80mb 1/2 height))


DRVTBL:
	; RD31		type 1	(SEAGATE ST225) 
	.WORD	615.
	.WORD	4.
	.WORD	256./4
	.WORD	671.
	.WORD	670.

	; RD52		type 2	(QUANTUM)
	.WORD	512.
	.WORD	8.
	.WORD	256./4
	.WORD	512.
	.WORD	0

	; RD53		type 3	(MICROPOLIS 1325)
	.WORD	1024.
	.WORD	8.
	.WORD	1020./4
	.WORD	1024.
	.WORD	0

	; RD54		type 4	(MAXTOR 2190)
	.WORD	1225.
	.WORD	15.
	.WORD	1020./4
	.WORD	1225.
	.WORD	0

	; RD32		type 5	(SEAGATE ST251)
	.WORD	820.
	.WORD	6.
	.WORD	1020./4
	.WORD	911.
	.WORD	0

Sorry, I forgot to include the format of this table... Here it is:




        ;;; DRVTBL parameter block format :
        ;;; max cyl (1-nnnn)
        ;;; max head (1-nn)
        ;;; precomp/4 (1-1020 /4)
        ;;; recal cylinder (0-nnnn)
        ;;; parking cylinder ( 0-nnnn, 0 = no park cylinder)


DRVTBL:
	; RD31		type 1	(SEAGATE ST225) 
	.WORD	615.
	.WORD	4.
	.WORD	256./4
	.WORD	671.
	.WORD	670.

	; RD52		type 2	(QUANTUM)
	.WORD	512.
	.WORD	8.
	.WORD	256./4
	.WORD	512.
	.WORD	0

	; RD53		type 3	(MICROPOLIS 1325)
	.WORD	1024.
	.WORD	8.
	.WORD	1020./4
	.WORD	1024.
	.WORD	0

	; RD54		type 4	(MAXTOR 2190)
	.WORD	1225.
	.WORD	15.
	.WORD	1020./4
	.WORD	1225.
	.WORD	0

	; RD32		type 5	(SEAGATE ST251)
	.WORD	820.
	.WORD	6.
	.WORD	1020./4
	.WORD	911.
	.WORD	0

John, thank you very much for the info, especially the last table. Its
exacally the data I have been looking for. I Have tried both an RD52 and
an RD53 on the Amiga, and have not (so far) been able to get either to
work correctly. I can PREP and FORMAT them (A2090 controller) as long
as I tell PREP that they are ST-225's! Doing this only gets me 20MB.
I will try again tonight using the info you supplied and let all know
what happens.

By the way, either the A2090 or the hard disk format is painfully slow!
Its better than a floppy, but not a whole lot. I took about 1.5 mins
to copy a 2700 block file from one hard disk to another. To copy an
entire hard disk to another (about 27000 blocks), took about 20 mins!
IBM PC's are like lightning compared to this. The disk seems to thrash
itself to death simply copying a file.

Frank

Well, I tried an RD53 with the data supplied in the previous note, and
could only get it to work if I told PREP it only had 4 heads. As expected,
I ended up with only half the size I should have. I tried the same with
an RD52 and that didn't work at all. I think that either the A2090 controller
or the PREP software has some problems.

Frank

I know some controllers have physical geometry limitations, but any I've seen
are limited to 8 heads, 1024 cylinders.

The following may offend some people so if you are sensitive, please skip 
this -

At the software store last night someone asked -

	Does a 3 1/2 inch floppy disk turn into a 5 1/4 inch hard disk when 
it gets excited?

    Now that my WEDGE and miniscribe 40M drive on line...any suggestions
    as how to partition up the 809 cylinders???  right now, Im still
    learning how to set this baby up, so if I have to blow it away and
    re-format no big deal...but in a few weeks...now that a horse of
    a different color....
    
    so.....
    
    What kids of trad-offs are there to many devices (like ram buffers,
    directory access, etc) between one (or two large partitions vs.
    more (4,5,??) smaller ones???
    
    this sounds like a good "religious" question, so express your selves!
    
    thanks...Scott
    	     keeper of mvcad3::user0:[amiga...]
    

    I set up my 40MB as a single partition.  I'd rather not hassle with
    fixed-size partitions--with my luck one of them will be too small,
    and I'll have to reformat and start over.  Instead of partitions
    I use directories to keep things organized.
        John Sauter

    
    
    I run my 40MB disk as one large partition. I works fine and you
    don't have to worry about runnig out of space on one side.
    
    If I could do it over I would split it into 2 parts ( 20MB/20MB).
    
    1) When I got more than 20MB on the disk I could tell the preformance
       difference. New files and directories take a lot longer now.
    
    2) With 20MB some of my windows are getting very full of Icons.
       I use Workbench as much as CLI and a lot of Icons are big.
       Even my top level is full of just drawers.
    
    
    		Steve Peters
    

    I have a couple of disks and have been fooling about for some time
    now with this very question.  I set up a disk with 2 large, equal
    partitions, and 1 or 2 small (1 Meg) partitions to hold about a
    floppys worth.
    
    The two large partitions give me better performance than 1 large
    one, plus, if you lose a partition, you can reformat it without
    blowing the rest of the disk.  In my case, I back up one partition
    to the other.  This has gotten me out of a jam at least once while
    I was testing drives.
    
    It is a personal choice in how you configure your environment for
    yourself.  It might even be worth having a partition set up for
    languages,applications, whatever,  a partition for development,
    one for documentation, etc.  It helps keep the windows/drawers from
    getting out of hand.  Like it was said, it's religious!
    
    john
    

    
    
    	I just remembered one more very important thing to keep in mind.
    Many programs that use requesters for file names have limits to
    the number of chacters in a file name. This means that you can't
    make the directory structure too deep or drawer names to long. So
    directory trees need to be wide not deep. It's the width that kills
    performance. So, split it up.
    
    
    		Steve
    

    I split my Supradrive into a dh0: of 10 megs and a dh1: of 10 megs.
    I use dh0: for the everyday stuff, and dh1: for archives etc.
    
    One nice thing about equal sized partitions is the ability to perform
    an AmigaDOS 'COPY dh0: to dh1: ALL' command to repack the files.  This
    dramatically speeds up performance if the partition becomes fragmented.
    
    Ed.

Does anybody have the following configuration:

2000 + 2090 + Rd51 + Rd53?

Either one of the drives works perfectly well as unit 1 (dh0)
but the other doesn't show up as dh1 (yes, I changed the drive select jumper
and removed the terminator pack and made approp. entries into mountlist etc..)
Cables are O.K.(btw: made from old unibus extension cable) 
- also the jumpers on the Rd53 - i tested it 
with a second Rd53 as unit 2 (total ~140 Mb - geeeee, too bad a have to give
the stuff back...)

Does anybody know about the jumpers on the RD51?
I'd like to test the Rd53 as #1 and Rd51 as #2, but I don't know
what's the drive select jumper on the '51??
(I had a look into the microfiche tech tips - no avail)

BTW: I had also very little luck with 2 different Rd52's (also on loan) neither 
as #1 nor as #2 nor as single drive - when prepping prep hangs or the
famous 'device not prepped' message appears... 
(I couldn't step out of a hanging prep with ctrl-c - as mentioned in
an earlier note!!)
When I decreased the number of heads down to 1 (ONE!) and the 
last track used by 1st partition to 4 or 5 the drive started 
rattling and rattling and got finally prepped.
DH0 was the only partition I could format, it gave me about 300 kb
out of a 40 mb drive.....
It seems to me that this type of disk has a sophisticated bad block
logic and tries to fill up all his replacement area (in this case
almost the whole disk).

2nd BTW: it's true - you can't have the whole Rd53 as a single 
partition....I tried it out of sheer curiosity, the formatter
doesn't complain (yes, needs some time but it's not slower then on a 
DEC-Rainbow), INFO shows a 69MB dh0, you can see the trashcan on it
(I wonder how the format wrote that onto it) but when you try to copy to it
'key error xxxx' (or something like that) appears.

3rd BTW: my prep version is 33.19 -
now eagerly waiting for WB1.3 and the famous FFS!!!

    If I am not mistaken the 2090 can only recognize ST506-type drives
    as DH0: and DH1:. I believe that is what the 2090 manual says.
    
    BTW, RD31's (half-height 20 Mbyte) are a dime a dozen these days,
    and RD32's (40 Mbyte half-height) are getting to be.
    
    Paul

	Re .21

	If I'm not mistaken all DEC Rdxx's are ST506-type drives...
	- and all works pretty well for 2 RD53's!!

	Your BTW: If you happen to get other RD's even cheaper then
	RD31 or RD32 ??? (compared $'s/Mb)

	Horst

    Well, I am in the market for a hard disk controller. I will be using
    two ST506 type drives, and would like to know:
                                         
    	1.  Has anyone used the Great Valley Products Impact disk
    	    	controller with the Adaptek A4000 "scuzzy" to st506
    	    	adapter?                 
                                         
    	2.  Has anyone had any problems with the Adaptek A4000?
                                         
    	3.  How many ST506 drives can one Adaptek A4000 handle?
    
    	4.  Does anyone know any reason that this config would not
    		work with KickStart V1.3  and/or with the future
    		Fast File System (FFS)?
    
    	5.  Will the FFS do a better job of handling bad blocks?
    		Bad Block replacement - or simply a Bad Block File?
    
    Thanks for your help !!
    =jbf=
    

	re .-1 and the GVP SCSI controller:

	I just bought the A2090 HD controller, after spending an
	unsucessful two weeks trying to get some information from
	Great Valley Products.  There's apparently only one person
	on their sales staff (guy named Greg Garnet), and every time
	I or my local hardware dealer tried to get him, he was "out
	of the office right now"...

	I finally decided that, even though it sounds like a great 
	piece of merchandise, if they're this difficult to contact,
	what will it be like if I need technical assistance or
	repairs ?  Take it for what it's worth.

	-Art

    The 2090 handles 2 ST506 devices and 7 SCSI devices. I have one
    connected to my 2000 and NEC 40 Mbyte drive. Works just fine.

    
    Re: .24
    
    I would be working through my local dealer. You might remember the
    one -- the one that trades cables for free, does local hardware
    support, cares about the customer!!  So I would use them to avoid
    contact and frustration with GVP.
    
    However your point is well taken. I will consider this before making
    a purchase of their controller. Thanks for the warning.
    
    
    Re: .25
    
    Some questions:
    
    	1. Does the A2090 do DMA transfers?
    
    	2. How is the speed ?
    
    	3. Will the A2090 be upgraded to allow AUTOBOOT ?
    
    	4. Can the ST506 devices be partitioned ?
    
    
    Sorry for all the questions, but I started to total up the cost
    of the components & decided that I had better be certain about the
    cost before I plunk down the money.
    
    Thanks to again to one and all.
    

    
>    Some questions:
>    
>    	1. Does the A2090 do DMA transfers?

	Yes, it does; on top of that, it's got its own Z80 on board to
	handle most of the donkey work.  There's also a FIFO buffer on
	the controller to reduce some of the contention for memory.

>    	2. How is the speed ?

	I don't have the stats with me, but I remember its maximum transfer
	rate as being pretty good; at this point the real limiting factor is
	the slow speed of the AmigaDOG file system; V1.3's FFS should clear
	some of that up.

>    	3. Will the A2090 be upgraded to allow AUTOBOOT ?

	Probably not, since the autoboot feature will require some ROM's
	and there's no space on the board to plug them. I thought about
	that before I bought the 2090, and it seemed like a minor problem;
	once the startup file is running, most of the real work can be done
	from the hard disk anyway.

>    	4. Can the ST506 devices be partitioned ?

	Oh, yes, quite easily.  In fact, its probably a good idea to do so,
	since under 1.2 larger partitions take longer to access.  Again,
	under 1.3's FFS, the speed-thing  shouldn't be such a big problem.

	One other note about the A2090.  The HARDISK.DEVICE driver that
	CBM currently supplies will ALWAYS mount the first partition on a
	hard drive using the slow file system; all later partitions can
	be mounted with the new FFS.   This may change somewhere down the
	line, but it means that for now (i.e. v1.3), it's a good idea to
	make the first partition something small, so you won't take a
	performance hit for the whole disk.

	BTW, if you're looking for cheap drives, there's a place in
	AZ called HARD DRIVES INTERNATIONAL (number supplied on 
	request) who have excellent prices on Seagate drives, and 
	their service is excellent; I found them in Computer Shopper.
	They guarantee alomst everything for a year after purchase.
	<End of unsolicited endorsement>

    Thanks for the help.  I have decided to give the GVP IMPACT controller
    a shot.  I was swayed by several factors.  
    
    1.  First of all I need some more memory on the system  --  and to
    think that I once thought that a PDP/11-45 with 128 KWords of memory
    was extensive !!  But I don't think that I will need more than 1MB
    of memory for a while.
    
    2.	Secondly, I like the idea of being able to use the AUTOBOOT
    feature of V1.3 of AmigaDOS. The GVP board has the two EPROM sockets
    and is ready to rock and roll.
    
    3.	I was able to obtain the memory a 'pre-inflated' prices. I probably
    paid more than I might have from a mail order place, but with the
    cost of memory chips -- who knows ?? !!
    
    4.	One of the employees has already purchased the board and was
    very pleased with their service and the product itself.  A satisfied
    customer is nice to see and hear.
    
    
    Though I have the board, I do not have any disk installed on it.
    However, the memory is available until I can get the rest of the
    pieces up and running.  I will be using a Adaptec A4000 controller
    for two ST506 disk drives.  So when the money is available, I will
    be purchasing all of the missing pieces.  I will keep everyone
    posted on the GVP board.
    
    Thanks for your help and suggestions.  I know I don't follow
    suggestions all that well, but the information and advice is
    appreciated.  At least now I know that I am probably going to make
    things confusing...
    
    =jbf=

    FYI:
      
    1. The IMPACT controller software currently does not support the
    Adaptec SCSI to ST-506 adapter.
      
    2. Late this month or early next month GVP will be releasing the
    next version of the controller software that will:
      
      	a. Support the Adaptec controller
      	b. Support more than one logical unit (disk) per SCSI id # -
      		the Adaptec adapter supports two ST-506 drives (I think
      		I saw an advertisement for a version that supports upto
      		4 ST-506 drives).
      	c. Provide Graphic support to configure a disk.
      		Yup, that's right !!  You will be able to have the
      		MOUNTLIST automatically generated...  Let's hope they
      		don't introduce to many limitations.
      
    3. I have been very pleased with the service and answers that they
    have provided. I received a possible work around from them, but
    I have not had the opportunity (read: TIME) to try it.  Here is
    where I could use some help.  It is my understanding that C. LTD.
    has a program that can be used to prepare the Adaptec controller.
    I would be interested in finding out how that program works. Depending
    on when I receive the updated software, I might try the work around.
    
    
    =jbf=

	Can anyone help? As a Enginneer this is embarassing, but
I don't have any documentation.

	I'm trying to install 2 RD53's on an A2000
with a 2090 card.

	I bought a 'dual drive' cable set  from a computer store:
	2 - 20 pin cables, 
		and
	a 3-connector 34 pin cable that has me a bit puzzled:

	from left to right the connectors are
	a 34 pin 'berg'
	a 34pin keyed edge card.
	a second 34 pin keyed edge card.

	Now for the 'gotcha':	 the cable
has been split and a 5 conductor section TWISTED before insertion in the third
connector. The split and twisted section apears to be conductors #25 - #29
which I suspect are the drive select lines.( I could be off by 1, did a
visual count of the cable ridges.)

	Question #0: Is this o.k. ? ( I want to make sure I've got a 
	{				st506 cable set.
	Question #1: 870.15 tells me how to jumper the first drive, but how 
			should the RD53 drive jumpers be configured for the
			second drive (the one with the third connector 
			plugged in)?

	Question #3: Do these drives have terminators that need to be dealt
			with when daisy chaining drives? I see what i suspect
			is a SIP terminator pak just to the right of the
			rd53 configuration jumpers.

	Question #4: There are other jumpers on the drives besides the
	drive select, what are they and are they properly configured:

	with the pc card down and loooking at the rear of the drive:
	It looks like:


	    _		     _  :::::: ------
	:: | |  uuuuuuu uu | | uuuuuuuuuuuuuuu uu


		: - Jumper pin-pairs.
		u - edge fingers,
	 ------     a sip.

	The left-most jumper, and the right-most three jumpers are installed
	on both drives.

Any assistance would be appreciated.

Regards, Verell Boaen

    I'm wondering if I can get some advice from the hard disk gurus out
    there (I know you're out there, o.w. I wouldn't be running a hard drive
    now!)
    
    I went to de-arc an arc file which I recently downloaded onto one
    of my hard disks; at one point the drive light stayed on for an
    unusual period of time, then ARC output an error message and
    subsequently hung its CLI.
    
    Popped a new CLI, tried to copy the ARC file to RAM:.  Shortly,
    the drive light again stayed on as before, COPY output an error
    msg, then gave up (at least it returned control of the CLI).
    
    I typically combine a bunch of ARC files into one master ARC file;
    I had done this with this particular file, so I re-extracted it
    from the master ARC file, this time directly into RAM:.  Now I can
    de-arc it and copy it ok.
    
    So I'm beginning to think that the hard drive has some type (format,
    surface) of error.  I've done a complete backup of the hard disk
    and now I'm trying to figure out what to do next.  I can either
    try to reformat the disk using AmigaDOS "format", or I can use the
    low-level formatting program (this is a C-Ltd SCSI controller on
    an A-1000) to scan the disk for surface defects.
    
    The low-level formatter is supposed to build a defect list as it
    scans the disk and write that list both to a file and to the (first?)
    track on the disk, where the Adaptec controller is supposed to read
    it on power up.  What I found odd is that when I originally did
    this, no defects were found (other than the list I entered by hand).
    This seems odd as I thought that surface defects were fairly common.
    
    My questions are: should I go ahead and try another low-level format
    (takes a llooonnngg time) or should I just do AmigaDOS format? 
    If there is indeed a new surface defect, should it show up during
    the AmigaDOS format as an error?  In other words, does AmigaDOS
    format look at all areas of a disk, or just those it needs to alter
    to put its structure on the disk.
    
    As always, thanks very much for advice and comments.

    Re .31:

    If you were able to do a complete backup of your disk after this error
    occurred, doesn't that suggest that the disk is not the problem?

	Paul

Re: .31

>In other words, does AmigaDOS format look at all areas of a disk, or just
>those it needs to alter to put its structure on the disk.

AmigaDOS format will verify all the blocks of the disk not marked as
bad by the last low level format.  This means for a hard disk that if
it passes the AmigaDOS format that the disk is free from errors.  (This
doesn't work for floppies since AmigaDOS writes floppies without regard
for the index mark:  A track may show up as good under format but then
be shifted to a bad spot at the next write!)

If the AmigaDOS format turns up bad blocks, you will have to do a low
level format to inform the Adaptec controller of the bad blocks so it
can handle them.

Does your low level format really take longer than the AmigaDOS format?
On my disk, the low level format goes ten times faster than the AmigaDOS
format!  (I have a ST-157N SCSI disk with a Pacific Peripherals controller.)

What error did COPY return?

the INFO command has an "errors" column for the disks, you might check it
from another CLI if it happens again (at least you'ld know if the disk
was charged with an error).

Do you have a fan for your drives? If not, maybe you got the error, turned
things off for awhile, then came back and backed up a cool disk...



John

    Re .33:
        Sorry.  Randy.  Any comments on the 157N and the Overdrive?  I have
    both of them on order, and I hope to hear that you've had no problems
    with either part, and that you're thrilled with them.
    
    Back to formatting...
    
    Gary

    The COPY error was simply "Error while copying, destination file
    removed" or something quite similar.  After each attempt to access
    that particular file, the error count in the INFO output would go
    up.  So...
    
    I deleted the file, then did a complete backup of the harddisk,
    which completed without errors.  Then I attempted AmigaDOS format;
    it formatted and verifying track 1,2,3... then stopped.  No error
    message, it just stopped (I've heard this before, perhaps
    comp.sys.amiga?).  Tried it three more times --- same thing.
    
    So I fired up the low-level formatter.  I figure, it's *got* to
    find some bad sectors this time.  It does all its passes, says it's
    finished with no errors, yet there are *still* no entries in the
    bad sector list, at least not on screen.  This is C Ltd.'s SCSI
    format program specially written for the Adaptec controller; it
    has a small window which is supposed to display the size of the
    bad sector list for the current disk.  This remained at 0.
    
    Ran AmigaDOS format, this time it completed without stopping.  Restored
    all the files from the backup disks...now everything seems ok. 
    At this point I suspect C Ltd.'s formatter program may be lying
    about its bad sector list.
    
    RE: .33
    
    Yes, this formatter takes *much* longer than AmigaDOS format.
    It makes several passes across the disk, apparently writing and
    then reading back test patterns, and it is supposed to keep track
    of any bad sectors it discovers; then it is supposed to write this
    bad sector list onto track 0 (?) of the disk; when you power up
    the Adaptec controller reads this list into its own memory and remaps
    bad sectors into good ones on the fly.

Well up until now I've been pretty much ignoring all hard disk notes thinking
I'd never be able to get one.  But now I'm thinking of getting an A2090A.
Having quickly looked at all the notes with the hard drive keyword I'd like
to run what I think I've learned by the hard disk masters to see if I've got
it right...

1) An A2090A ($325) will control 2 ST-506 devices and up to 7 SCSI devices.
2) All DEC RDxx drives are ST-506 drives.  RD3x are half height, RD5x are full.
3) A half height drive will fit inside the A2000 (Where is the correct place
	to mount it?  Where the 5� drive goes?) and can run off of the A2000
	power supply.
4) A full height drive requires an enclosure, and a power supply (having a fan
	is recommended) which run about $100.00.
5) The "A" after A2090 means AutoConfig and AutoBooting (?).

After I have an A2090A (provided I can get the proper cables) I could easily
hook up an RD3x.  Then for a little more $$$ I can get an enclosure etc. and
hook up an RD5x.

..Dan

>3) A half height drive will fit inside the A2000 (Where is the correct place
>	to mount it?  Where the 5� drive goes?) and can run off of the A2000
>	power supply.
 
Yes on power and the 5 1/4" bay.

>5) The "A" after A2090 means AutoConfig and AutoBooting (?).

A is for AutoBooting, the A2090 will also autoconfigure.

You might want to look back in the high 1900's in this notes file for my note 
on adding an RD32 to my 2000. I had a lot of trouble because I had a dead 
drive but the cabling and mountlist info in the replys is correct and very 
valuable.

	Dick

Thanks for info and the pointer.  The Note # is 1981.  I can't wait to try this.
Perhaps the moderator could add the HARD_DISK keyword to that note.
Guess I should have done a "DIR/tit=RD32"

..Dan

    
    HEEEEEEELLLLLPPPPP!!!
    
    Last evening upon arriving home with a quantum pro-drive in hand
    I discovered that my A2090 Hard Disk Software Diskette was VERY
    VERY unhappy.  Disksalv, and Diskdoctor were unable to recover 
    much.  OH well.
    
    What I need, Is there anybody in the greater Marlboro Region (Read
    Worcester to Boston, Lowell to Milford) with this disk that will let
    me copy it???? 
    
    thanks.
    
    brad

      Well last night I tried to prep and format a RD53 drive as drive
    #2. (DH1:)
    It made a little noise and lock up my A2000.  Now my st-225 drive DH0:
    is not working at all.  It was fine just before I tried to prep the
    RD53.  Now all I get is the st-225 drive is not mounted. I tried to re-
    mount the drive, but no luck.  What caused this problem?   Also, is
    there a size limit per partition with amiga-dos? 
    Also, What is the right mount list numbers for the RD53?

    Thanks  Bob