30 Oct
2005
30 Oct
'05
12:16 a.m.
ard at p850ug1.demon.co.uk (Tony Duell) wrote:...
Hmmm. I thought it was where the taps stuch to itself,
or to the drive
belt, when the catridge was stored. In which case your idea might not
help much.
Aha! There goes my original hypothesis. Well, that is relatively easy to
check. I took apart two old tapes and found the following:
Assuming that the tapes were rewound before storage, there are two
stretches of weakened magnetic coating on both tapes I looked at; the first
was short while the second longer. These correspond to the length of tape
in contact with the drive belt
In
http://www.series80.org/Articles/tape-cartridge-lr.gif
you can see that on the take up reel (left) there is less tape in contact
with the drive tensioning belt, while on the right hand reel (which has most of
the tape and hence a longer radius) more tape comes in contact with the
belt.
Looking closely at the belt, I can see remnants of tape oxide stuck on it
(in fact its stuck so well, I could not remove it).
So how come we get an EOT error when using these tapes?
The tape has tiny holes near the beginning and end signifying BOT and EOT.
(BOT) (EOT)
+-------------------------- .... --------------------------------------+
| |
| o o o o o |
| |
+-------------------------- .... --------------------------------------+
I assume that the first hole signifies the end of the data region and then
if the drive sees one hole its EOT, while if it sees two holes (prob. more
than one) its BOT.
So when it sees the blank region, it interprets it as one hole and
reports end of tape.
Using a marker to paint (on the back side of the tape) over the region
does not work, since the markers I tried allow some light to go through.
I stuck a piece of splicing tape and that took care of the hole, but I
got a DATA error. I guess the 85 needs some of the info that is missing
from the tape.
I have also noticed that the tape does indeed stick onto itself on the
right hand reel, but if you unwind it slowly you can get it unstuck without
damaging it. I assume that conditioning the tape by warming it up before use
may reduce this sticking problem, but I still would not want to fast wind
a tape in that condition.
Given the above, I am even more confident about data recovery, as we may
be able to use the existing tape mechanism of the 85 to read the data
by skipping the regions that appear blank until the tape un-spools and
the drive stalls.
This may end up being a software project after all :-)
So the new plan is to see how I can get the tape moving at 10 ips
(I'll look at the 9815 diagrams) and see how I can tap the output of
the read/write IC (1820-2418) to get the bits from the tape.
If this does not work, I'll have to get somebody to help me built the
preamp for the heads.
I wish I could understand this infernal assembly language of the 85
so that I could look at the internal routines that access the tape
for ideas.
------------------
ard at p850ug1.demon.co.uk (Tony Duell) wrote:
Does anybody know if those drives are tape and track
compatible
with the 85 drive? In other words, would the 2 track head of a 9815
correctly read the tracks on an 85 tape (yes, I know the controllers
are very different).
There was a program on the Series-80 User's Library
(8291029) that
allowed retrieving 9845A SAVEd programs. The program was for the
HP-85 and did not specify any ROMs or peripherals. You may infer
from that, that the Series 80 could read 9845 tapes.
On the other
hand, the 9845 used dual directories (the second one was a backup)
per tape while the HP85 only uses one, so this may be an indication
that they are NOT compatible.
--------------------
"Joe R." <rigdonj at cfl.rr.com> wrote:
Both ideas sound do-able to me. It might also be a
good way to find out
how HP "protect"s the program tapes and possibly duplicate them.
I would
guess some bit in the directory entry (like LIF volumes), but
if my plan works, we'd be able to copy tapes at the bit level so
HP's protection would be irrelevant.
Where are you located? I have a couple of spare scopes
and I might be
presauded to loan you one. But I'll have to do some checking to be sure
that that these specific models can survive shipping.
Thanks for the offer, but I work at the College of Engineering here at
Drexel, so I can find a scope (as well as someone who can tell me how
it works :-)
--------------------
So, back to the drawing board...
**vp
30 Oct
30 Oct
2:24 p.m.
ard at p850ug1.demon.co.uk (Tony Duell) wrote:...
This would seemm to imply my guess was correct _in this case_. Of course
other tapes might fail in other ways.
Hmmm. I thought it was where the taps stuch to
itself, or to the drive
belt, when the catridge was stored. In which case your idea might not
help much.
Aha! There goes my original hypothesis. Well, that is relatively easy to
check. I took apart two old tapes and found the following:
Assuming that the tapes were rewound before storage, there are two
stretches of weakened magnetic coating on both tapes I looked at; the first
was short while the second longer. These correspond to the length of tape
in contact with the drive belt Using a marker to paint (on the back side of the tape)
over the region
does not work, since the markers I tried allow some light to go through.
I stuck a piece of splicing tape and that took care of the hole, but I
got a DATA error. I guess the 85 needs some of the info that is missing
from the tape.
Of course if you did the trick you were suggesting and read the tape at
the flux-transition level, you could at least try to recover the data
that was on the undamaged tape. It might be possible to piece together
some of the files.
I assume the tape format is somewhat LIF-like. That is, there's a single
directory at the start of the tape, and files are contiguous.
So the new plan is to see how I can get the tape
moving at 10 ips
(I'll look at the 9815 diagrams) and see how I can tap the output of
I think the 9825 controller is easier to understand (and easier to use
outside the machine if you mamange to find one).
the read/write IC (1820-2418) to get the bits from the
tape.
I would not try to use that IC. AFAIK it's a single-channel device, you
can't read both heads simultaneously. Look at the 9815/9825 schematics to
get an idea how to build a read amplifier using standard parts. You'd
need to build 2 of them of course, one for each track.
ard at p850ug1.demon.co.uk (Tony Duell) wrote:
I know for a fact that the 9845 tape drive is the same as that in the
9815 and 9825.
I think the data format is very different, but the drive is compatible in
the snese that the track positions are the same, etc (if not, the 85
wouldn't have a hope of reading the 9845 tape). Which means you might be
able to do the reverse.
I would not try to use a 9845 tape controller. There's a big, heatsinked,
HP custom chip in the middle, linked to a DAC for the motor speed
control. The servo loop, read/write circuit, etc, is all partly inside
that ASIC. And it's not even software compatible with the 9825 (it uses
different lines of the I/O bus, IIRC the 9825 uses DMA, the 9845 doesn't,
etc).
-tony
Does anybody know if those drives are tape and
track compatible
with the 85 drive? In other words, would the 2 track head of a 9815
correctly read the tracks on an 85 tape (yes, I know the controllers
are very different).
There was a program on the Series-80 User's Library
(8291029) that
allowed retrieving 9845A SAVEd programs. The program was for the
HP-85 and did not specify any ROMs or peripherals. You may infer
from that, that the Series 80 could read 9845 tapes. On the other
hand, the 9845 used dual directories (the second one was a backup)
per tape while the HP85 only uses one, so this may be an indication
that they are NOT compatible.
6954
days inactive
6954
days old
1 comments
2 participants
participants (2)
-
ard＠p850ug1.demon.co.uk -
vp＠cs.drexel.edu