• Please review our updated Terms and Rules here

Trouble low level formatting MFM drive...

MikeS said:
Good luck with that, especially if one of the platters is used for sync...
Click to expand...
This isn't something I know about, if it was used for sync and it was stored magnetically, surely the drive would die out anyway once it needed an LLF?
Or does an LLF rewrite the sync block?

I have a Maxtor 380MB (XT-4380E) brick that's supposed to be a spare for my IBM RT, it spins up and down, figured the motor/bearings were shot but if it has a magic sync platter I can't rewrite I need to avoid them.
 
MikeS said:
I've also wondered what actually goes wrong with some of these drives; I've got some Seagate drives that are nice and quiet, no stiction issues, etc., but they just go through the initialization and give up. They restore to track 0, sweep across the platters a couple of times, twitch rapidly back and forth around the middle of the disk and park the heads, not ready.

These drives (ST251-1) have 3 platters, all six surfaces apparently available for data unlike some other drives (e.g. Maxtor) which use one surface for sync; any idea how they locate track 0? I don't see any sensors, so do they perhaps use an out-of-normal-range track? Or what else the problem might be since it doesn't seem to be bad heads or surfaces and at least two worked fine the last time they were used?
Click to expand...

Well in the case of this TM-252 drive, there is an optical sensor to locate track 0 just like on a floppy disk drive.
 
geoffm3 said:
I wonder if some of these issues could be cleared up if you could degauss the media. Of course you'd have to open up the drive to have a chance to do that...which brings me to another point... if you could provide a sufficiently clean environment could you DIY a head clean/replace and/or platter swap. Seems like it couldn't hurt if you had several broken drives of the same type to mix and match parts as it's most likely a money-losing proposition to have them repaired professionally (if that's even possible). Seems like you could build a cleanbox to open the drive up in and work on it. It's a shame that so many of these old stepper motor type MFM/RLL and SCSI hard disks are going tits up now, as I consider them characteristic of many of these early machines. I know there are emulator options and replacements, but sometimes the original is more gratifying.
Click to expand...

My personal opinion (without any hard data to back it up) is that it's the medium itself that's slowly degrading. Early drives used an oxide coating similar to that of floppy disks or mag tape. Given that we're seeing binder failures on those media, it wouldn't be surprising to also see that occurring on hard disks. Plated media would be a different case--and I think it reasonable that drives using that might be a little more robust.

Then there's loss of lubricant. For a disk spinning at 3600 RPM, the lubricant has to be pretty light. Unfortunately, light lubricants evaporate with time. More friction means that the relatively weak spindle motor can't get the platters up to speed, so you get the spin-up, spin-down cycle. I've got a Maxtor XT-1140 like that--I actually witnessed the lubrication failure. The drive spun up and started to work just fine, then audibly began dropping speed and finally stopping. When power-cycled, the disks would attempt to spin, then give up. It should also be noted that heads dragging on the disk surface will have the same effect.
 
I don't have many (one or two) with the spin-up, spin-down problem. The vast majority of them spin fine immediately and stay spinning without any interruption whatsoever. They just cannot be LLFed.
 
Chuck(G) said:
My personal opinion (without any hard data to back it up) is that it's the medium itself that's slowly degrading. Early drives used an oxide coating similar to that of floppy disks or mag tape. Given that we're seeing binder failures on those media, it wouldn't be surprising to also see that occurring on hard disks. Plated media would be a different case--and I think it reasonable that drives using that might be a little more robust.
Click to expand...

That would be my theory too. Back in the 80's I did a lot of work with CDC 14" CMD's - had a fixed platter (1,3 or 5 data surfaces with a servo), and a removable cartridge (single platter with a data surface and a servo).

An uncommon, but not unheard of problem would be a perfectly functional drive that would suddenly start throwing read/write errors. A format would mark some bad sectors that weren't there before, and the process would start again - fine for a while, then random errors with increasing frequency. The cure was new media. Easy if it was the removable pack that was a problem - a little more involved if the fixed media was involved. In either case, the media appeared perfect and the heads were undamaged, ruling out a head crash. We had 3 affected media tested, revealing low magnetic coerciveity (manufacturer techno-crap meaning some of the oxide layer had dropped off). This was on media that was - at most - 5 years old, so I guess that finding this kind of issue on drives that are 25+ years old shouldn't be that surprising!

Taking a drive apart and replacing the media is definitely doable BUT if you have a drive with a logic issue and one with a media issue I'd swap the circuit boards and bin the other drive - it's much easier than putting another media in the drive with the good electronics!
 
Chuck(G) said:
My personal opinion (without any hard data to back it up) is that it's the medium itself that's slowly degrading. Early drives used an oxide coating similar to that of floppy disks or mag tape. Given that we're seeing binder failures on those media, it wouldn't be surprising to also see that occurring on hard disks. Plated media would be a different case--and I think it reasonable that drives using that might be a little more robust.

Then there's loss of lubricant. For a disk spinning at 3600 RPM, the lubricant has to be pretty light. Unfortunately, light lubricants evaporate with time. More friction means that the relatively weak spindle motor can't get the platters up to speed, so you get the spin-up, spin-down cycle. I've got a Maxtor XT-1140 like that--I actually witnessed the lubrication failure. The drive spun up and started to work just fine, then audibly began dropping speed and finally stopping. When power-cycled, the disks would attempt to spin, then give up. It should also be noted that heads dragging on the disk surface will have the same effect.
Click to expand...

Yeah, I've thought about that too... you would think though since the flying heads aren't touching the surface of the disk that any loss from lubricants would be an issue only at start up or shut down since that's the only time the heads should be touching the platter. Rather than when it's running steady state. The TM-262 will seemingly run for some indeterminate amount of time (and not always the same time) and then stop.

I suppose you are right though about the media degrading over time. I suppose that's perfectly plausible, but there are a couple of obvious distinctions between a hard disk and floppy... namely the floating heads and rigid disks. So you would think binder failure wouldn't be as much of an issue there, unless the material sloughs off on its own from the centripetal force of the spinning disc.

I wonder if you could 'anneal' the media by cooking it similar to how you do with floppy cookies.
 
Phil Saunders said:
That would be my theory too. Back in the 80's I did a lot of work with CDC 14" CMD's - had a fixed platter (1,3 or 5 data surfaces with a servo), and a removable cartridge (single platter with a data surface and a servo).

An uncommon, but not unheard of problem would be a perfectly functional drive that would suddenly start throwing read/write errors. A format would mark some bad sectors that weren't there before, and the process would start again - fine for a while, then random errors with increasing frequency. The cure was new media. Easy if it was the removable pack that was a problem - a little more involved if the fixed media was involved. In either case, the media appeared perfect and the heads were undamaged, ruling out a head crash. We had 3 affected media tested, revealing low magnetic coerciveity (manufacturer techno-crap meaning some of the oxide layer had dropped off). This was on media that was - at most - 5 years old, so I guess that finding this kind of issue on drives that are 25+ years old shouldn't be that surprising!

Taking a drive apart and replacing the media is definitely doable BUT if you have a drive with a logic issue and one with a media issue I'd swap the circuit boards and bin the other drive - it's much easier than putting another media in the drive with the good electronics!
Click to expand...

...but if you have a couple two platter drives with one bad platter each, you could make one good one with a media swap! Who knows how long that would last though.
 
Phil Saunders said:
That would be my theory too. Back in the 80's I did a lot of work with CDC 14" CMD's - had a fixed platter (1,3 or 5 data surfaces with a servo), and a removable cartridge (single platter with a data surface and a servo).
Click to expand...

I think I remember those. I think we had one on a VAX 11/750.

I wonder if any of the older drives, say an 808 or 6603 (Bryant) are still operable with their media intact...
 
Stone, are you trying to kick up a fight?

"Centrifugal" force (the imginary force that seemingly causes a particle to "flee from the center" in a rotating frame of reference) doesn't exist as a real force. What you have instead is a real force due to inertia; things tend to travel in a straight line. Call it centripetal, if you will.
 
Chuck(G) said:
I think I remember those. I think we had one on a VAX 11/750.

I wonder if any of the older drives, say an 808 or 6603 (Bryant) are still operable with their media intact...
Click to expand...

The RAMAC on display at CHM is still readable. The people that did the restoration read all 25 platters to see what was there (just test data)

I read mid-70's Diablo 30 packs regularly.

Those sorts of drives have pretty low bit density.

I don't know of any Bryants that are operable.
 
Al Kossow said:
The RAMAC on display at CHM is still readable. The people that did the restoration read all 25 platters to see what was there (just test data)

I read mid-70's Diablo 30 packs regularly.

Those sorts of drives have pretty low bit density.

I don't know of any Bryants that are operable.
Click to expand...

Probably destroyed in a deluge of leaking oil... :)
 
Okay, couple of observations on the drive from last night.

First off, when the heads land while powering down the drive, a fairly obvious "grinding" sound can be heard.

I attempted a fresh low level format. That passed. Then FDISK then FORMAT and loaded some basics. That seemed fine. Loaded a couple more things, and transferred Norton Utilities onto the drive. Tried Nortons on it and it flagged and marked three clusters as bad. Loaded up some software and attempted to use some of it. Data errors. Ran Norton's again, and this time it flagged a bunch more clusters than the first time... about 20-25 or so.

Retried this whole thing a second time. Attempted another low level format, etc. That passed with no bad sectors marked. And eventually the drive behaved worse than the first time. NU flagged a bunch more bad clusters than the first two runs.

So, the drive will boot on its own now, but not much else is reliable.

The grinding sound while the drive stops is troubling. I am making sure to park the heads when I stop it.
 
IMHO, the drive is dying and will only get worse in time. I have a huge box full of similar doorstops so I've seen this syndrome many times. It seems to be the fate of too many old MFM drives.
 
That may not be the head you're hearing. A lot of these drives have a friction brake on the spindle (basically a solenoid and a pad) and they can get quite noisy.
 
There is at least one machine shop in California that specializes in rebuilding Ford flathead V8 engines that Ford stopped producing in the 1950's. There is another in North Carolina the specializes in the "early" Chrysler Hemi made from 1955 thru 1958. There are more steel bodied 1932 Ford coupe's on the road today than Henry built in 1932. In short these exist because there is a market for them. Were there a market for, say, rebuilt ST225 drives someone would figure out a way to make money rebuilding them. However, the tide seems to be running stronger in the direction of emulating them on newer hardware.
 
Chuck(G) said:
That may not be the head you're hearing. A lot of these drives have a friction brake on the spindle (basically a solenoid and a pad) and they can get quite noisy.
Click to expand...

That's sounds plausible. The one thing that makes me think head is that it makes the sound close to when it comes to a halt and it slowly spins down. It has a ringy quality to it. Almost as if you were running your finger around the rim of a wine glass.
 
Well, if it's a problem, you can sometimes see what's going on by looking at the bottom of the drive. Removing the brake, so long as you don't move the drive while it's spinning down shouldn't cause any harm. What happens is that the brake "shoe" was bonded to a friction material, such as cork or felt and the material has degraded and fallen off with age. What you may be hearing is just metal-against-metal.
 
I'm just posting this out of interest, I don't think it's the case with geoff's drive but it gives a look inside the TM-252:

https://www.youtube.com/watch?v=XfpA63Reaic

That one suffered a proper head crash. I noticed it uses a different colour platter to what I'm used to seeing as well.
I have a couple of drives that make the wine-glass-ring noise on final spindown - squeaky brake would make sense.
 
SpidersWeb said:
I'm just posting this out of interest, I don't think it's the case with geoff's drive but it gives a look inside the TM-252:

https://www.youtube.com/watch?v=XfpA63Reaic

That one suffered a proper head crash. I noticed it uses a different colour platter to what I'm used to seeing as well.
I have a couple of drives that make the wine-glass-ring noise on final spindown - squeaky brake would make sense.
Click to expand...

IS TRACK 0 BAD??? lol

Thanks for the link. Good to see the anatomy of the drive... course not much different than most this vintage, is it?
 
You must log in or register to reply here.
Back
Top