• Please review our updated Terms and Rules here

Seagate ST2383 5.25 form factor ide drive

k2x4b524[

Veteran Member
Joined
Jun 13, 2009
Messages
1,521
Location
Tacoma, Washington, USA. Zip code 98404
Another one, this one is seagate 338mb 5.25 ide hard drive, the seagate st2383. Formats, reads, writes, but no matter what bios translation i use for this sucker, no matter what i do, scandisk reports it can't check the drive due to a data error in cluster 2, here the fat is. Spinrite reports no errors, yet chkdsk and scandisk wont touch it, and the seatools 1.1 doesn't even see it. This drive is going into a 5x86-133-p75 system as a secondary, for the win 95 swap and such. Is there a way to fix this drive, or just the fat area, or is this thing boned?
 
yes, every time i changed the geometry of the drive, i would wipe the partition and start over, but there are 4 different geometries this thing uses. the bios translation reports 837 cylinders, seagate tells you 737 cylinders, the chs is around 1700 cylinders, and the lba is about 383 cylinders. It had bad sectors, which spinrite fixed up, but the first few clusters ALWAYS check out clean. Also seatools still will not see the drive at all, i may have to fall back to the old ontrack disk manager for that one. But i would like to get this cluster 2 issue thing fixed so i can use this drive.
 
yes, every time i changed the geometry of the drive, i would wipe the partition and start over, but there are 4 different geometries this thing uses. the bios translation reports 837 cylinders, seagate tells you 737 cylinders, the chs is around 1700 cylinders, and the lba is about 383 cylinders.

:huh: ? 1700 cylinders? What?

If the area in question is near the start of the disk, create a primary partition that include the flawed area, change its type (use a partition editor or even a disk editor) and then create a new active primary partition.
 
yep, the chs reports 1700 standard cylinders. so what i should do is give it a partition like i've been doing, then change it's drive type, which disk editor would be good for this, and what bits would i look for to do this? and feed it a new active primary partition, that might work, once i figure out what i need to edit. I also did some research on the drive itself, according to seagate, the drive has a jumper that tells it to use sector sparing or not, i'm going to see if it's turned on or not, would that make a difference?
 
An IDE drive? Try the following utility and see what it says. It reports the results of the IDENTIFY command:
 

Attachments

  • IDESDI.ZIP
    7.5 KB · Views: 1
A potential last-resort utility that I have not tried yet. Has anyone tried this?

This is an interesting utility that tries to repair bad sectors by re-writing the magnetic field around bad sectors. http://piotrkn22.republika.pl/drev/index.html
He has a freeware version on the page that is DOS only for older drives http://piotrkn22.republika.pl/drev/free/drev.zip

"DRevitalize is a program that repairs bad sectors (physical defects)
on popular magnetic media (hard drives and also floppy drives to some
extent) by generating a special sequence of high and low signals around
the physically damaged area."
 
I call BS. The mechanics of controlling the magnetization patterns are buried so far in the drive electronics that you can't really do anything.

Of course, if you believe that putting a magnet on your fuel line will double your mileage, then maybe it's for you. ;)
 
tried that program, it gives me what the seagate site tells me, 737 by 16 by 56, so thats what i'm using, it's formatting now, see what happens with that one, i'll also give that program, drev, a shot, might prove useful.
 
yep, just tried it with the translated amount, but it returned 9mb bad and scandisk still returns

"Scandisk encountered a data error while reading the FAT entry for cluster 2"

This error prevents Scandisk from fixing this drive

My Scott Mueller book of hard drives shows 873 by 14 by 54, This drive is proving most interesting and perplexing.
Anyone encountered anything like this before?
 
I did some digging. That drive may have a Seagate label, but it's really a CDC Wren VI under the hood.

Physical configuration 1747/16/54. Translated: 952/11/63 or 757/16/56, depending on firmware.

You may have a way out of this--take a look at the jumper block on the drive: There should be 7 sets of them (look at the TH99 entry here. Set the C and D jumpers for sector sparing. That reduces capacity some, but gives you a spare sector on each track to use. Repeat your diagnostics.
 
you think i might be getting all these errors because the seagate parameters may be off by a little bit? The sector sparing is what was giving me the 9mb bad at the end of the drive, the 952/11/63 didn't work and neither did the 757/16/56. Right now the drive is sitting my my p-166s as secondary master to get it ready for it's swap job. No translation appears to be working, why would it keep reporting cluster 2 having a data error? MSD is telling me it's 368/32/56 but the cmos is reporting 378/32/56.......


Update, i even tried at it's native parameters, before translation, it took the format but there is still an error in the fat entry for cluster 2, so the next question is, how do i fix that little bugger, or is the drive a 338mb brick? Will it still work for what i need it to?
 
Last edited:
IIRC, when FDISK asks you if you want to use entire drive for the partition and you say No, it asks you to input the starting and ending CHS. For the ending, use the same one from the standard translation ending cylinder, but, for the beginning one, use cylinder 1 if it starts at zero and cylinder 2 if it starts at 1.

This will reduce the capacity slightly, but move your MBR and FATs to the second cylinder on the drive.
 
I've never seen FDISK do that, Dru. You get to say what percent or how many megabytes, but that's it. Is there some alternative version of FDISK for your scenario? I suppose you could temporarily hook this drive to something that's a 486 or above and then use something like GPARTED to put the partitions anywhere you'd like.
 
i've found a couple of dos based partition managers but they don't let you select a starting cylinder, i've tried cute partition manager, but all that did was botch up it's mbr, which scandisk found and fixed, but it still chokes on that cluster 2 error. I'm going to try partedmagic and see if i can make a bootable disk, like you can in partition magic, and try it that way, and if you got fdisk to allow you to pick a starting cylinder, what trick did you use?
 
all i'm finding is dr-dos 7.02 and 7.03, but it appears to be the complete os, just found the images though. I'll see what i can do with that, the free dos one is just a mirror of dos. This is most frustrating *sigh*
 
Last edited:
Back
Top