floptical problems
27 messages in this thread
Well, after months of using it on the IBM I finally tried to use the
21MB floptical drive on my Amiga. I've got a 2000 with a 68020
accelerator and a 2091 interface card. There's a Seagate 177N as the
primary drive on the card. Oh, I'm running 2.04 at the moment.
Anyhow, HDToolbox recognizes the Floptical as Device 4 on Unit 0 (the
Seagate is 0 on 0). It can read its properties OK. But when I tell
it to do a low level format the lights for both the drives (Seagate
and Floptical) come on and nothing happens. At least it doesn't wipe
the Seagate. Oh, once I've done that I have to remove the power from
the Floptical to get the system to boot.
"Well, after months of using it on the IBM I finally tried to use the
21MB floptical drive on my Amiga. I've got a 2000 with a 68020
accelerator and a 2091 interface card. There's a Seagate 177N as the
primary drive on the card. Oh, I'm running 2.04 at the moment.
Anyhow, HDToolbox recognizes the Floptical as Device 4 on Unit 0 (the
Seagate is 0 on 0). It can read its properties OK. But when I tell
it to do a low level format the lights for both the drives (Seagate
and Floptical) come on and nothing happens. At least it doesn't wipe
the Seagate. Oh, once I've done that I have to remove the power from
the Floptical to get the system to boot."
Well lets go through some of the basics first. Is your Floptical on
the internal SCSI bus or external? If it's external, and the last
drive on the bus, is it terminated properly? If it's internal and
shares a SCSI cable with the Seagate, did you remember to remove the
termination from the Seagate and then terminate the Floptical?
Ummm, boy. Will there ever be plug and play?
The Seagate is on the internal and the Floptical on the external.
The Floptical works fine on an IBM external (which does not have a
drive on the internal) so I presume it is terminated OK. Appreciate
the input.
Did you click on the "save changes to drive" button after entering
the drive configuration into toolbox before trying the low level
formt? I made this mistake myself a few weeks ago with mounting a
Sequest on a 4000 with a 2091 and had a similar situation.
I had a floptical drive mounted on my 3000 without any problems, but
it died and I haven't tried it on the 4000..
John mentioned something that is very important… termination.
Each end of the physical scsi bus should be terminated and no other
termination should exist.
Have you "changed drive type" to get the drive characteristics set in
the RDB? Have you partitioned the drive?
I would not necessarily worry too much about the lowlevel format.
You actually don't lowlevel format most scsi drives anyway. Have you
tried to DOS format the drive (from Workbench or CLI)?
As I replied to John, the Floptical works externally on an IBM so I
presume termination is right. Change drive type seems to want to
write to the drive. You can't do that 'til it is formatted. Format
di0: doesn't find anything. It seems like a chicken and egg
problem. Maybe if I IBM formatted it first. I'll give that a try.
The Amiga scsi.device is not going to be able to understand an MSDOG
hard disk (there may be a way to get it done with CROSS-DOS).
HDToolBox DOES want to write to the drive. It wants to write a thing
called the "Rigid Disk Block" that defines the drive. This is the
same thing that scsi.device reads when the computer is reset or
powered up.
The RDB contains such information as the size of the drive,
synchronous ability, filesystem type, max transfer value, type of
scsi device, removable media or not, size of data blocks, bootable
device or not, boot priority, automount or not and even the device
name (not volume name).
The Ami can not format the drive until after the RDB and partition
information have been written to the disk. Watch out on the
partitioning part as HDToolBox usually want to create more than one
partition (an will do so unless you tell it otherwise).
As to termination… probably most machines out there are NOT properly
terminated. On the Amiga side the problem is that (as far as I know) the
A3000T is the only machine that was ever set up properly. The A3000T has
a terminator installed on the external scsi connector and on the internal
drive which is at the end of the internal cable. There is no terminator
set on the motherboard.
Most A3000s (and A2091s) have terminators on the motherboard (card). This
is ok if you only have either internal or external drives. Though it
usually works, it is incorrect when you have both internal and external
drives.
I agree that your problem is probably not termination but with scsi one
does have to be careful. Also, the PC boxes usually run the scsi bus at a
MUCH lower speed than the Ami does which makes issues such as termination
not quite so serious.
-bill
Until I've disassembled every device on the SCSI bus, I never believe that a
device is terminated or not. If the floptical is in an external case, and even
if it has two bare SCSI D-shape connectors, it might still have the resistor
packs in place inside the external case, on the drive. If that's true, it
might still work on a PC. It might work even if you doubly-terminate it with
an external terminator. You never know until you double-check.
Another possibility is that your 2091 SCSI controller card has old ROMs. I had
to upgrade to version 6.6 ROMs before lockups between my hard drive and the
floptical would go away.
As for formatting a floptical disk, I usually insert a previously formatted one
while starting HDToolBox. Then, switch to an unformatted disk and low level
format it. That way HDToolBox doesn't try to read or write the unformatted
disk.
– Alex
You don't HAVE to low level format the floptical disk. The disk
already are by the manufacturers. Just go into HDToolBox or whatever
and give it a name and/or partition it. I've got a similar setup to
yours with a Seagate HD and a Quantum on an A2091 and a DMI (Insite)
floptical drive as a third unit. No trouble. The fact that two drive
lights come on when you ask the softare to format the floptical
suggests to me that your terminator resistors aren't set properly on
one or more SCSI devices. — Ron Legro from Milwaukee via AUTOPILOT
I haven't managed to get inside to check the termination yet but I
expect that is most of the problem. I don't think you are correct on
the low level. It is definitely necessary for the IBM and hard to
imagine that they would have them Ami formatted.
Paul: I was expressly advised not to worry about doing a low-level
format. And, indeed, some of the diskettes I've purchased have a
stamp on them indicating they already are low-level formatted. The
CompUSA dealer here advises that ALL floptical diskettes are indeed
low level formatted at the factory. This low level formatting is NOT
computer specific. I'm sure it couldn't hurt to go through that
routine, but I've never bothered. — Ron Legro from Milwaukee via
Well, I finally opened up the Amiga. I've pulled out the 2091 and am
stumped. Where are the terminator resistors? There is nothing
socketed that looks like a resistor to me. Do some of the jumpers
control this? I bought the board used and got no docs with it.
Paul;
The terminators on the A2091 are labeled "RN1", "RN2" and "RN3" and are
the resistor packs right next to the 50 berg SCSI connector (on the side
toward the back of the machine). They may NOT be socketed.
Why do you want to remove them?
-bill
Well, I've got a drive mounted on the board and am trying to get an external
drive (a floptical) to work also and it has been indicated in previous
discussions that the on-board termination is fine for one or the other but not
both.
Since you seem to be in the know: What do the option 1,2,4 jumpers do? Does
the memory that will go on the board simply act as amiga memory or is it a disk
cache? What kind of chips does it take?
I appreciate the info.
——Paul—–
Paul,
J5 Option 1 jumper: LUN enabled if ON. LUNs (Logical Unit Numbers)
allow more than one device at each SCSI address. With LUNs enabled, in theory
it is possible to put 8 LUNs at each SCSI address (ID)…giving you the ability
to connect up to 56 ( 7 x 8) SCSI devices to one card.
J5 Option 2 jumper: Time-Out/Parity usage, Jumper OFF, Time out
disabled, short wait period for drive to become ready, parity checking enabled.
ON…Time out enabled, Long wait period for drive to become ready, partiy
checking disabled on the message-in phase.
J5 Option 4 jumper: Reserved for future expansion.
Don
Paul;
The A2091 (and probably the A version too) is not correctly designed. The
terminators on the A2091 are not removable. Thus you SHOULD use only the
internal or the external bus (though you can use a drive on the card
itself (without terminators) and still use the external bus (the cable to
the internal drive will be less than the stub limit for a drive mounted on
the card).
Now having said that, I personally have not seen an A2091 that will not
work with three sets of terminators present (one set on each of the ends
of the internal and external buses and the one set on the card).
I also know of people that have cut the leads on the internal terminators
and that too has worked.
Something else that works nicely is to install a Maxtor MXT-540 drive
anywhere on the bus with the terminators "on." The active SCSI-II
terminated MXT-540 settles an improperly terminated bus even when the
MXT-540 "is not supposed to be terminated" because of it's position on the
bus. This was first (as far as I know) done by a Maxtor employee that is
also an Amiga owner. A bus analyzer indicated that putting the MXT-540 at
ANY location on the bus improved the bus termination. Subsequently other
Amiga users trying the same "trick" obtained the same appearent results.
Hopefully, someone will find their copy of the A2091 manual (I have not
been able to locate mine) and tell you what memory chips to use.
-bill
Bill –
You commented that the 2091 SHOULD be used only with EITHER internal OR
external devices, but not both, since the BOARD's termintors are soldered
and not removable, but that — with the internal on a short stub —
externals might well work fine.
This is indeed the case with ours. On-board Conner (with terminators),
external floptical and/or tape drive.
But we just acquired (used) a PP&S Zeus 040/SCSI board, and now find we
cannot install the 2091 and have anything on it work. Our 2nd HD is now in
an external box connected to the Zeus. Which is fine, of course, but can
you think of any reason we shouldn't be able to get the 2091 to work as
well?
. . . . JimDoc, on AutoPilot
Jim;
Actually your on-board Conner should NOT have the terminators installed.
However, as I also said (I think I said it in that message), I have never
seen an A2091 or A2091A that would not work with an extra termination
load.
What is the name of the scsi device driver for the Zeus card?
Are both cards trying to Autoboot?
Have you tried switching the order in which the cards are installed in the
machine?
-bill
Bill –
> What is the name of the scsi device driver for the Zeus card?
ppiscsi.device
> Are both cards trying to Autoboot?
AutoMOUNT, yes. AutoBOOT, no.
> Have you tried switching the order in which the cards are installed in
> the machine?
Can't. Zeus is in the co-processor slot.
. . . . JimDoc, on AutoPilot
The terminator resistors aren't on the controller card itself, but on
the devices attached to it. In the case of a stock A2091, the first
device usually is the Quantum ProDrive 40S that comes with the card.
That's the big square thing. Look on the drive's bottom side and
you'll see three terminator packs right next to the 4-pin power
connector. Pull 'em if you've added another drive to the controller.
Keep them and mark them for position in case you need to reinstall.
Ron;
That is not correct (at least for the A2090). There are terminator
resistors on the card.
-bill
I was reading from the A2091's manual, which shows in a diagram the
resistors on the bottom of the on-board Quantum ProDrive. So, yes, the
terminator resistors are "on the card," but only via the drive. At least
in the case of the A2091.
>I was reading from the A2091's manual, which shows in a diagram the
>resistors on the bottom of the on-board Quantum ProDrive. So, yes,
>the terminator resistors are "on the card," but only via the drive.
>At least in the case of the A2091.
I took apart my Amiga today, and had a look at the 2091. There are
terminator resistor packs on the 2091 card too! Look under the short
50 pin SCSI cable, they are there right next to the 50 pin socket on
the board. It's those three long enamel coated packages with about a
dozen pins in a single row coming out of each of them. They are in
addition to the hard drive and floptical terminators, meaning that
you can probably do without the terminators on the internal hard
drive.
The general termination idea is to have one set of terminators near
each end of the whole SCSI bus to stop electrical signal reflections
off the ends of the cable. Of course, the practical approach is to
add and remove terminators until it works.
Ron;
I can't speak to the A2091A with any authority but I have an A2091 (no
suffix) right next to me on the desk right now (not in a machine). This
card definately has terminator resistor soldered into the board between
the 50 pin berg connector and the blank area where the "on board" drive
would mount.
-bill