#Bridge HDisk Problems
4 messages in this thread
I am having problems with frequent soft errors on the Amiga partition of
my hard drive. It is a FUJI FK309-39R RLL drive,and the controller used is a
Western Digital WD 1002A 27X. The only other card on the IBM side is an AST
6Pak+. I also have replaced the 68000 with a 68010.
Using debug to do a dynamic low level format I used the following:
Interleave=5; # of cylinders=615; # of heads=4; Starting reduced write cyl.
=616 (Drive manual states that onboard Write precomp cyl. =616
uProcessor handles these two ) Max correc. error burst length=11 bits CCB
option byte,Step rate select = 3 (Taken from drive manual)
The low level format proceeds fine,and detects no bad areas. Moving to
FDISK,shows 940 cyls,of which 470 cysl. are used by the MSDOS side. ADISK
shows 470 cyls. and these are allocated to ADOS. High level format occurs
without incident on both parts.
Writing to the AMIGA side produces no errors,but reading from it does,and
they are frequent and inconsistant (meaning the same file does not always
give an error). Also,clicking RETRY allows the read to continue,with no
apparent ill results..text and data files are complete,and programs run
without a hitch. After several days of this,I went through the formatting
again but before doing the ADISK step, I ran a battery of tests on the
drive,and nothing turned up wrong. I then completed ADISK,and ran a drive
diagnostic called AUTOTEST,which appears to to be based on the example in the
Libs & Dev manual,and errors turned up, again with the inconsistency
regarding location. The errors returned were 16 & 17 (often) and also 2 (just
once). None of these are listed in the RKM,so I looked in the ENHANCER manual
to see if they were defined there. That's where I read that parameters in
trackdisk.device are extremely hardware dependent. Could this be the cause of
the problem? Or has jdisk.device been optimized for the 68000? Or does the
problem lie with values chosen in the low level format?
Mark J. Morales 72307,2503
Mark,
Do you have any JLinked devices to the PC side? If so, unlink them
before using the PC side HD for the Amiga side.
As a matter of curiosity…try a lower interleave value…that will
slow down the data transfer from the drive and that may solve the problem.
What I'm thinking is that the drive is pumping out data faster than the
interface can handle it…so it's getting 'lost'. I realize that changing
the interleave to a lower number will slow down the drive for both the PC
side and the Amiga side…but it might work and therefore may be worth it to
have a somewhat slower drive that works rather than a fast drive that doesn't
work.
I might also be way off base, Andy may have other thoughts on the
subject.
Don
Don,
I don't normally have any JLINKed drives, but I do have Device=JDisk.SYSin the config.sys file. I've also fiddled with the
interleave factor, but knocking it up or down by only 1 really slows down the IBM side's access, and since I use both sides
equally, I woul prefer not having to 'baby' one or the other. Besides, going thewhole 9 yards in the setup is a real pain, so my
experimenting with the settings is over until I get some new leads. I will let you know how things finally turn out, and for
anyone else reading this who has the bridge card w/ a hard drive on the IBM side, let me know how things are going for you.
Mark
p.s. Liked your column in the Amiga Transactor!
I'm not sure what the problem is; I've never tried an RLL disk on the PC side
and attempted to have an Amiga partition on it!
We're going to get one in and see what happens. It might be connected to
that in some way. Or it might be something else. I don't know at this point.
andy