MFM/RLL data recovery
nemesis-lists at icequake.net
Tue Apr 5 12:55:31 CDT 2005
Replying to a bundle of posts again:
> If you have an MFM drive used on an AT or better you have a good chance of
> reading it on any AT or better system with most any controller.
> RLL or 8 bit based systems have no such "standard".
I think this is the best summary of the situation I've seen yet.
> A terminator is also required at the last drive in the chain. Assuming
> you are using just one drive, not having the terminator installed would
> probably cause the drive not to work properly.
Yeah, both drives have terminators. One of the Miniscribe drives
doesn't have a terminator but I remembered to move it when I tried that
> If you are using a HD that originally had a twist, the jumper select
> will most likely be set to drive 1. Using a straight cable will cause
> the drive to appear as drive D instead of C ... I don't remember what
> problems that will cause. So if using a straight cable, make sure the
> drive select is at 0 (of 0-3) or 1 (of 1-4).
That's where it appears to be on both Seagate drives.
> If you are using an 8-bit card with the bios enabled, make sure in the
> 386 setup that there is NO HD installed (the bios will take care of it.)
Yup, did that.
> If you are trying to save files on the HD AND it was installed with an
> 8-bit controller, you don't have any choices but to use the 8-bit card
> bios to access the data AND an identical controller to the one that was
> used to low level format the drive. Fred can probably comment on whether
> it has to be the same type of controller, or the same controller for
> this to work. If you have any idea of what the original machine it came
> out of was, that would help :).
Here is the machine that the MFM drive came out of:
Intelligent Data Systems PC-88
> Finally, the clicking you hear on the drive is not a good sign. I would
> power up the drive with no cables connected (except of course the power
> cable) and if the clicking still continues, the drive is probably bad.
Clicking only happens when I've connected it to one of the 16-bit
controllers and the BIOS is attempting to find the disk. Otherwise the
drives sound fine.
> If it is important enough, you could also send the drive to a data
> recovery service and leave it to them to deal with it.
Not really important, but the drives used to be in BBS systems in the
early 90's and we are quite interested in the files and messages that
are on there for archival purposes.
> A few times I've seen different PC controllers from the same manufacturer
> (i.e. WD or whatever) work from one model to another but not be able to boot
> without a low-level formatting (booting from a floppy allows access).
Now *this* is interesting. I was assuming that if the drive was going
to work at all, it would be bootable too. Why is is that you wouldn't
be able to boot from the drive, yet you'd be able to read it if you
booted from some other media?
> As was pointed out if you know the drives were formatted on a 8 bit
> controller just set the BIOS to no hard drive and don't even try the 16 bit
> controllers, they would just be a waste of time. Your best bet would be to
> try different 8 bit cards and boot to a floppy then do a "dir c:", repeat
> until you've tried all cards.
Okay, will do.
> Does the controller require a real 8 bit machine, yes not all 8 bit
> controllers even worked in 16 bit computers.
Any examples of problem controllers in this area?
> Is the drive bad.
> Has the drive formatting been messed up.
Hoping not :(
> Is the drive terminated properly.
> Is the drive jumpered properly.
As far as I can tell, yes.
> On IDE all of the truly low level stuff is 100% hidden so the only
> compatibility problems with IDE tended to be drive geometry (early problems
> when the same drive could be addressed by different geometries) .
Yeah, I had a Fujitsu IDE drive die on me recently. I was able to get
the data by swapping the drive PCB with that of another drive which was
the same model, but not identical (6 months newer, different stuff on
the label). I was surprised that it actually worked.
Ryan Underwood, <nemesis at icequake.net>
More information about the cctech