• Please review our updated Terms and Rules here

MiniScribe harddrive crash?

D

Denniske1976

Experienced Member
Joined
Jan 17, 2011
Messages
480
Location
The Netherlands
Hello everyone,

I have a little bit of a situation here: Last night I was playing Commander Keen 4 on my HeadStart (I have the CGA version hehe) and we had a power outage :-( So, the computer of course unexpectatly shut down... after switching it on again, my hard drive makes a really scary noise (twice, like a rattle) and after that I get a "please reboot system" message. So no more DOS starting etc.

Took the drive out, and switched the computer on with the drive outside (but connected) and I see a little rod (the sticker says "interruptor") go round until it can't go any further, and trying it again. That's what's causing that rattling noise at boot. After that the drive appears to make that typical "MiniScribe-sound" and trying to read off the disk.

So uhm, is there anything I can do about that or is the drive shot? Was it a head-crash? And of course: Does anyone happen to have a MiniScribe 8450XT drive around that they would like to sell? (it's the 40MB XT-IDE version I think, same drive as the Explorer has?). Really too bad, took me ages to get my CGA monitor replaced with another original one and after solving the problems with the disk drives (that happened to be a bad CMOS Battery) now my hard drive is dead :-( :-(

Thanx for the help and have a great weekend!
 
The drive may still be somewhat salvageable, provided its not in a boot sector or damaged head. In that scenario, you could partition around the bad sector, or use software to map a single partition around the defect. Personally, I've used drives with a few bad sectors for years without problems, when I partitioned around them.

In any case, if the drive is toast, and the heads did crash, I would replace it with a ide to cf adapter and a 32-256mb cf card. Just my 2 cents though...
 
Sounds like it's trying to read the sector that the head crashed on. Hopefully you didn't have anything important on it, so try a low level format.
 
I presume we are talking about a Miniscribe 8450XT hard drive. It's 40MB, 3.5" form factor with an IDE interface.

I don't think that losing power really counts as a head crash. By the time that drive was made heads were automatically unloading on power loss, or at least moving to a safe track. To have a head crash you need to have a violent thump while it is running - just a spindown due to power loss should not cause damage.

The rod/interrupter mechanism sounds like something that I've seen on an old Teac drive. A solenoid locks it into place to prevent the heads from moving around while the power is off. Having the rod go to the stops just indicates that the drive is trying to reset or seek to a home location.

At this point I would reinstall it and make sure that the fan that was cooling it is still doing it's job. If you can't boot from the drive or read it at all then it's time for a low level format. If a low level format does not restore the drive to usability it's probably done. Remember, these things were designed to last 5 years. You are at least 4x outside of that.


Mike
 
Chuck(G) said:
I've heard that, in some cases, a brick works just as well as a Miniscribe drive. :wink:
Click to expand...
I have several still working miniscribe drives, they seem to be fairly reliable based on the track record I have had with them, that said sometimes the "interrupter" will stick and keep drive from seeking, and while I know it says never to move by hand, a little turn one way or the other (only a few steps either way when powered off) will un-stick it and will usually start seeking again next power-up.
 
If you don't care about what's on the disk then by all means do a low-level format, fdisk and high-level system format.

But if there is important data on it then it might be worth while to boot from a floppy and run some diagnostics before you permanently erase everything by formatting.
 
RWallmow said:
I have several still working miniscribe drives, they seem to be fairly reliable based on the track record I have had with them, that said sometimes the "interrupter" will stick and keep drive from seeking, and while I know it says never to move by hand, a little turn one way or the other (only a few steps either way when powered off) will un-stick it and will usually start seeking again next power-up.
Click to expand...

No slam on Miniscribe per se--I've got a couple of big (FH 5.25" ESDI) Miniscribes--one even has a Miniscribe labeled PCB, but carries the "Maxtor" label. It still works to this day.

No, I was referring to a particularly ugly episode in Miniscribe's history:

The suit was brought by Kempner Capital Management and by U.S. National Bank of Galveston, which bought Miniscribe debentures and then sold them at a loss after accusations were made that Miniscribe had falsified its financial statements, by, among other things, shipping bricks as disk drives and counting them as sales.

The episode at Miniscribe has become one of the most scandalous in computer history and has already tarnished the image of both Mr. Wiles, the turnaround expert, and of Hambrecht & Quist, a San Francisco company.

Miniscribe filed for bankruptcy protection in 1990 and its remains were purchased by the Maxtor Corporation. Several shareholder lawsuits are still pending in Denver against Miniscribe as well as its outside directors, its auditors, and Hambrecht & Quist, as well as a suit by the Securities and Exchange Commission against some officers.
Click to expand...

To the best of my knowledge, no customer ever received a brick-in-a-box.
 
It sounds like you can see the head actuator on the side of the drive, and that it is rotating OK through to the end-stop (calibration point). When low-level formatting it is possible track-0 could be bad. Someone posted a while ago (with an ST-412 IIRC) some success in adding a thin shim to the end stop (like a washer) to effectively cause the drive to calibrate to track-0 where track 1/2/3 was before, and was then able to run the formats OK.
 
Hi again,

Well... been trying to get a low level format working almost all night. I have the original HeadStart LX-40 installation disks and there's a utility called LLFORMAT on them. Seemed like the low level format was going OK, I can hear/see the drive formatting (ticking noise, nothing weird or scary and drive light on/off). Unfortunately after that was finished, FDISK was giving me an "Error reading drive" message all the time. So eventually gave up (yeah, I'm a quitter I know hehe) and went to bed.

This morning I was thinking: "Let's have a look through my old IBM and other PC stuff", since I know I had some kind of HeadStart mainboard somewhere as a spare (although I think it's from a HeadStart II or something but it looked the same as my LX-40 board). Opened up a box that was in storage for quite some time and the first thing I saw on top were TWO (yes, two!) other MiniScribe 8450XT harddisks. To be honest I have absolutely NO IDEA where they come/came from... must have saved them from one or another (2 actually) headstart computers. One of the drives even has the "Philips" sticker on it like mine has.

Anyhoo, took the two drives upstairs and swapped one drive for the other. I did change the electronics PCB though... wanted to know if it's the media or the drive electronics that were bad. Well, the other drive is in now and I have low level formatted it and it's currently installing DOS 3.31 :) So I guess it really was some kind of head crash?? Because obviously the media is defunct on the old harddrive? It also kept making that scary noise when turing on power, even after I rotated that interrupter thingie around a bit.

So uhm, thanx for all the help (again!!). Gotta love this site :) Luckily I have two 8450XT drives around... the only one I could find was on eBay for about 200 bucks (go figure??).

Now on to get Commander Keen back on and back to playing ;-)
 
Funny thing is: I have a small UPS standing here and it's not connected ;-)

Anyhoo, the "new" hard drive has one bad cluster. Is there any way to map those out?
Or is it not a problem if DOS knows about it?
 
If you run FORMAT (high level), bad clusters will be picked out and marked as allocated. After that, you don't need to worry about them--DOS won't use them.
 
Well that's what I thought as well... But the DOS format didn't report any bad sectors. But I was experiencing errors while installing Commander Keen and some other apps. I'd be getting "sector not found" or "error reading sector" and eventually doing a surface scan with PC Tools 7.1 (diskfix.exe) brings up some 5 bad sectors.

Weird or what? Or is the disk maybe having media problems?
 
Some formatting utilities have an option to "mark sectors (or tracks) bad", which guarantees that they'll never be used by DOS or anything else (a bit is set in the sector address header).

Find a good disk exerciser/test program such as SpinRite and let it run on the drive for a day. That should clear up the question of what you're seeing.
 
I always found it annoying that DOS would retry when formatting, so I wrote a utility to run a surface scan and mark any 'marginal' sectors (which took more than some preset number of ms) to read. Can't find said uitlity now though.

But bad sectors were totally normal and expected with MFM drives. There may evem be a list of bad clusters on the disk label.
 
Denniske1976 said:
Well that's what I thought as well... But the DOS format didn't report any bad sectors. But I was experiencing errors while installing Commander Keen and some other apps. I'd be getting "sector not found" or "error reading sector" and eventually doing a surface scan with PC Tools 7.1 (diskfix.exe) brings up some 5 bad sectors.

Weird or what? Or is the disk maybe having media problems?
Click to expand...
It sounds like your drive is behaving like the ST-4026 in my early 5170:

I low-level format the drive, declaring the bad tracks listed on the drive's track defect list. I then FDISK/FORMAT the drive. As expected DOS marks a certain number of clusters as bad. I run Norton Disk Doctor (NDD) to scan the drive (entire drive) and it reports no new bad clusters.

The following day, after the machine has warmed up, a run of NDD detects new bad clusters, and NDD marks them as bad. Rerun NDD - no new bad clusters found.

Same thing the following day. It takes about a week to settle down (no more no new bad clusters being detected).

The whole thing is repeatable.

I've seen this behaviour on some other drives of mine as well. I didn't see this behaviour in the eighties and so it must be due to deteriation of the drives. It's as though certain areas of the platter surface can now only retain an adequate amount of magnetism for a period in the order of days.

DISCLAIMER: Just because I quote NDD doesn't mean that I think it's the best product for the job.
 
You must log in or register to reply here.
Back
Top