• Please review our updated Terms and Rules here

a little challenge, i hope you can help with a Miniscribe 3085 (85MB) in a 5170 Type1

L

luvit

Experienced Member
Joined
Feb 16, 2014
Messages
488
Location
east of Newark, Oh i;m luvit
a little challenge, i hope you can help with a Miniscribe 3085 (3085003-229) 85MB in a 5170 Type1

so I have this 5170 loaded down with the previous owner's stuff since 1985.
i used gsetup to configure the HD as type 12.. I could read everything. Office30 booted, Lotus 1-2-3 booted.. no errors at all.
i figured i'd back it up for the heck of it, but wanted to review the tree, so I typed "tree > tree.txt"
DOS complained that it couldn't write to the disk so I stopped that right away.

Now after I tried to write to the disk, I have errors during DIR in the root directory.. and the Office30 doesn't boot error-free anymore.

Any directory I see in the root (before the and errors with the DIR command) is flawless data, and I saved it.. but I'm wondering if what steps I may take to see if I can overcome some of the corruption I caused when I attempted to write to the disk.

For example.. the BIOS accesses the disk as a 52MB drive, there is 40MB of data on the drive (per CHKDSK), but I can only access 20MB.

I've been sitting on this for weeks just seeing if anything comes to mind for me.. everything read 100% until I tried to write to the drive and any inspections I do to it must be unintrusive.
No big deal, I just wanted to see if there were any software on the computer which I've never used before.
Here's a pic of SpeedStore.
 

Attachments

  • 20141206_121007.jpg
    20141206_121007.jpg
    85.5 KB · Views: 1
You have a few challenges here.

First, the real geometry of the drive is 1170 cylinders, 7 heads and 17 sectors per track. That works out to 69615KB of storage, or just 68MB.

An AT with the original BIOS table thinks a type 12 drive has 855 cylinders, 7 heads, and 17 sectors per track. That works out to 49.68MB. So if you use type 12, then you are giving up 18MB of storage on the drive.

The standard AT BIOS can not use a drive with more than 1024 cylinders, 256 heads, and 63 sectors per track. You can get around the mismatch by using a better BIOS which knows how to shift the bits around to make your drive look more like something the BIOS can handle. Drive overlay software can do this too.

Now, onto your corruption.

The drive might be fine. The problem might be the mapping of the drive CHS (1170/7/17) to what the BIOS understands. If the drive is setup with one mapping and that gets changed, then data will be unreadable or jumbled. You might be able to boot and get a directory listing because the first few tracks of data are mapped correctly, but midway through the drive who knows what is going on. Are you sure that any drive overlay software you are using is still in place? Has this drive always been marked as type 12 in the BIOS?

Assuming that is all correct, the drive might need a low level format. As the mechanicals of the drive age things shift. What might have been a perfectly centered track when data was written 20 years ago now might be slightly off because the head positioning is not as accurate. A low level format will take care of this problem, but at the expense of losing the data.

The defect table is interesting. Errors seem to be fairly evenly distributed on all heads, but mostly in the lower numbered cylinders. A low level format might clean that up.
 
mbbrutman said:
You have a few challenges here. ...the real geometry of the drive is 1170 cylinders, 7 heads and 17 sectors per track. That works out to 69615KB of storage, or just 68MB.
An AT with the original BIOS table thinks a type 12 drive has 855 cylinders, 7 heads, and 17 sectors per track. That works out to 49.68MB. So if you use type 12, then you are giving up 18MB of storage on the drive. ...The standard AT BIOS can not use a drive with more than 1024 cylinders, 256 heads, and 63 sectors per track.

..You can get around the mismatch by using a better BIOS which knows how to shift the bits around to make your drive look more like something the BIOS can handle. Drive overlay software can do this too.
Click to expand...
yeah, actually i just picked-up the world's sweetest looking eprom burner because of this computer. ;) screw the overlays. ..lol
The drive might be fine. The problem might be the mapping of the drive CHS (1170/7/17) to what the BIOS understands. If the drive is setup with one mapping and that gets changed, then data will be unreadable or jumbled. You might be able to boot and get a directory listing because the first few tracks of data are mapped correctly, but midway through the drive who knows what is going on. Are you sure that any drive overlay software you are using is still in place? Has this drive always been marked as type 12 in the BIOS?"
Click to expand...
There is no sign of drive overlay software.. now keep in mind, I'm only familiar with drive overlay software, like OnTrack (1995?). I can boot with any homemade floppy boot disk before the hard drive and still read the hard drive's contents.. something not possible with overlay software, like OnTrack.
The previous owner (the only owner since 1985) is elderly & does not remember the history of this PC.. He couldn't recall the fact that this is not the original HDD (dated 1989). The last user-created file I found was created in 1993. -- he was a preacher and had >500 sermons created. All have been accessible.

The BIOS was not setup when I got it, there's no history, no sign of HDD overlays.
I chose type 12 because I felt it best-matched the HDD geometry.. Since I attempted to write to the HDD (and initiated corruption) I have had flawless access to over 20MB of the data.. I have a good feeling that Type 12 is how it was configured.. it would be interesting to see results if I never would have written to the disk :)
Assuming that is all correct, the drive might need a low level format. As the mechanicals of the drive age things shift. What might have been a perfectly centered track when data was written 20 years ago now might be slightly off because the head positioning is not as accurate. A low level format will take care of this problem, but at the expense of losing the data.

The defect table is interesting. Errors seem to be fairly evenly distributed on all heads, but mostly in the lower numbered cylinders. A low level format might clean that up.
Click to expand...
yeah, i think the drive is fine, also. I was just wanting to learn if there was some near-miracle recovery program to play with before I low level format.
but first, I have the world's most beautiful EPROM burner... now I need to learn how it operates and order chips.. it's not like an eprom burner I used in the early 90's.
 
You could take the drive and controller and put it in a machine that recognizes 1224 cylinders. Then see how it reads.
 
my 5162 is my youngest desktop.. so that won't do.
based on a lot of read success, before and after the bad write attempt, i still collected 20MB of data after causing that issue. I truly believe it was set to Type 12.
I think Mike's theory is good about the alignment. -- I was lucky to read any data at all as I've seen 30 year old hard drives not even be usable until they were Low Leveled.
 
Which EPROM burner did you get ? I have a Needhams EMP-20 and a MCT MEP-1.
The MCT burners were originally sold through JDR Microdevices. The MEP-1 has an
ISA interface card. The EMP-20 attaches to the parallel interface.
 
You must log in or register to reply here.
Back
Top