• Please review our updated Terms and Rules here

Need an opinion on a bad drive...

R

rvdbijl

Experienced Member
Joined
May 8, 2017
Messages
112
Location
NH, USA
I have had an old ST251-01 sitting on a shelf for a while, but didn't have a controller to drive it. Today I picked up an old IMB 5150, which had a WD1002S-WX2 controller in it, but no HDD.
So I put 1+1 together, so to speak, and tried to low level format the drive. The first try succeeded according to the LL format tool in the ROM of the card, but it only formatted 10MB. I ran FDISK, and got some odd noise from the drive while generating the partition, but it did say it succeeded. Reboot, format c:, got to around 10% or so, and then says format failed after a few "attempting to recover ...". So I'm sure there are some bad sectors on this drive. What I don't understand is why Format gave up? I thought it would just continue and mark those sectors bad.. Or is there some threshold? I tried with both DOS 6.22 and DOS 3.30. Both behaved identically with the high level format. After about 10%, they just gave up.
I figured out how to set the jumpers for 20MB, but the LL format was taking so much longer with many recalibrations (the stepper driving back to track 0) that I just gave up on that.

So .. I have a number of questions:
1. The drive does not have the termination resistor pack on it, could that be the cause of the noise during low level format & FDisk and the high level format failing?
2. The controller supports up to 20MB drives with slightly different geometry than the ST251. However, the geometry of the 10MB and 20MB selections on the controller fits in the geometry of the ST251. I assume that not using some cylinders and heads shouldn't cause issues with the drive, correct?
3. Based on all this, what are the chances this drive would work ok with a resistor pack and another controller?

Thanks!
 
I would start with new cables or at least cleaning up and checking the ones you have, then try a new controller (or different drive on that controller).
Personally a drive that works fine for 10% then suddenly fails isn't one I have high hopes for, but you might get lucky with a connection issue.

Terminator is great, but I've never fitted one and had it fix my problems [still worth getting one if you can as it's not expensive].
Also with the drive settings, Cyl and Heads will be fine, but things like Precomp etc may not be - however I would not expect that to be an issue until at least 50% through your 10MB format so I dont think it's relevant here.

Just my two cents
 
It sounds to me like the drive is just full of bad sectors. I would try it with the resistor pack just to be sure, but I would not expect that to make much difference.

What kind of odd noise did it generate? A high pitch, but not necessarily loud or continuous, screeching noise is bad.

If the geometry matches, I would expect that drive to more or less work with this controller.

Chances are the controller is returning a seek error, or something other than just a read error. The DOS formatter won't know how to handle that. (Also, the DOS formatter will bomb out AFTER getting to 100% if there are bad sectors in the sytstem area).

I'd suggest trying SpeedStor and letting it to do a read test. It will give you more details about what is wrong. It may also let you set up partitions and format with less hassle dealing with the bad areas.
 
DEFINITELY put a terminator in there! I've seen some really odd behavior from un-terminated mfm drives.
 
ibmapc said:
DEFINITELY put a terminator in there! I've seen some really odd behavior from un-terminated mfm drives.
Click to expand...

I would like to... But not sure where to find one, besides on a dead drive off eBay or something like that. :) Is it a standard resistor bridge I could get at Digikey or Newark? And if so, anyone know the part #?
 
ST506-type terminators are the usual 220/330 ohm network variety. From the schematic on bitsavers, the one you need appears to be a Dale CSC10A05-221/331G or equivalent. There are plenty for sale out there. Datasheet
 
ibmapc said:
DEFINITELY put a terminator in there! I've seen some really odd behavior from un-terminated mfm drives.
Click to expand...
Absolutely. Unless one has the power to rewrite the laws of physics not having a terminator where one is specified is inviting trouble.
 
I tried with a termination resistor pack, but unfortunately it still won’t complete the high level format. Looks like a section of the drive from cylinder 38-45 is bad on head 0. Low level and fdisk seem to handle this ok, but DOS format still fails after you can hear (and see) the stepper go back to track 0 to realign itself a few times....
I guess this disk is unfortunately no longer viable... Although I remember back in the 80’s keying in a bad track table as part of the low level format. But it doesn’t seem like there was an option to do that in the controller I have...
 
Does DOS format eventually complete? It may flaw out a sizeable part of the drive, but it shouldn't give up unless the boot tracks are affected.
 
Try running SpeedStor http://minuszerodegrees.net/software/speedstor.htm and do a read verify test. Depending on the controller, it may add that to a bad track list and should give you more meaningful error messages. That by itself might not make a difference, but if you then use SpeedStor's partition creation tool it should stop you if a bad sector would wind up in a partitions FAT or root directory. If it does that, try creating a smaller primary partition.
 
I suspect that there's a strong probability that the media flaws with expand over time, so this may be a fool's errand. (I've been caught in those)
 
Chuck(G) said:
Does DOS format eventually complete? It may flaw out a sizeable part of the drive, but it shouldn't give up unless the boot tracks are affected.
Click to expand...

It does not complete. It gets to about 10% (or around cylinder 40-something) and then just drops back to the prompt with "Format Failed". I may try to hook it back up and try Speedstor, just to see if it makes a difference. But I suspect this thing has been sitting on a shelf for too long probably. :)
I was planning to put in an XT-IDE anyway, but just wanted to see if I could resurrect this old drive first...
 
SomeGuy said:
Try running SpeedStor http://minuszerodegrees.net/software/speedstor.htm and do a read verify test. Depending on the controller, it may add that to a bad track list and should give you more meaningful error messages. That by itself might not make a difference, but if you then use SpeedStor's partition creation tool it should stop you if a bad sector would wind up in a partitions FAT or root directory. If it does that, try creating a smaller primary partition.
Click to expand...

Hmmm.. I think that may do the trick, even without Speedstor. If I tell fdisk to create a primary partition that is < 38 cylinders in size (should be large enough for a DOS boot partition), then created the extended partition in the rest of the disk. Logical drive D, small again (cyl. 38-40-something), and never format it, logical drive E starts after cylinder 40-something.. Maybe that will work... I guess I'll have to try that this week sometime!
 
It kinda sounds like Track 0 might be hosed and if that's the case there is little to no chance of getting any partition table to work.
 
Track 0 (Cylinder 0, head 0) can't be hosed; otherwise you couldn't write a partition table and DOS wouldn't start a FORMAT operation.

That being the case, you can start the boot partition pretty much anywhere.

Personally, I wouldn't bother, however. Chances are that the situation will only get worse.
 
You must log in or register to reply here.
Back
Top