VOGONS


First post, by precaud

User metadata
Rank Member
Rank
Member

I recently found four 42MB MFM drives that I used briefly in the 90's, sitting idle since then. The factory defects list showed 4 or 5 bad sectors on each of them.

I decided to low-level format them and see how the surfaces have held up. I used SpeedStor to LLF, and then it scans the surface for bad sectors.

On two of them, it found over 20 additional bad sectors before it had even gotten halfway through the scan. I decided that was not a good sign and scrapped them.

The other two completed with 10-12 bad sectors for the entire disk. That seemed reasonable to me given its age and inactivity. But that is an assumption. I don't know if that is in fact reasonable.

So it raises the questions:

What is an acceptable increase in bad sectors, not from surface wear from use, but just from age?

What is the recommended procedure for putting a hard drive back into service after a long period in storage?

Reply 1 of 15, by derSammler

User metadata
Rank l33t
Rank
l33t

What is an acceptable increase in bad sectors, not from surface wear from use, but just from age?

Impossible to say, as normally it should not increase at all just by age. Additional bad sectors arise if dust entered the inside of the drive, the heads were not parked and landed on the data area instead of the parking area, and if there were manufacturing errors.

You can use the drive for some time, doing intensive copying, reading, and writing of data. After that, do a LLF again. If you get even more bad sectors then, the drive is junk.

Reply 2 of 15, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

I'm not an expert on these old stepper-motor drives, having entered the PC scene around the ~50-100MB IDE timeframe, but... AFAIK, they can be cranky beasts. It's not necessarily fatal to have a test run go lousy with bad sectors. If you leave it running for half an hour and do the surface scan again, it could end with totally different results (for better or worse). To my knowledge, that's just... eh.. the "charm" of MFM.

If you haven't re-oiled the motor bearings, that could also help.

Reply 3 of 15, by Unknown_K

User metadata
Rank Oldbie
Rank
Oldbie

Other then factory defects if there is a growing amount of new bad sectors I just don't use the drive.

Collector of old computers, hardware, and software

Reply 4 of 15, by precaud

User metadata
Rank Member
Rank
Member
SirNickity wrote:

If you haven't re-oiled the motor bearings, that could also help.

Really? You oil the bearings? Never heard of it being done on a HDD.

Reply 5 of 15, by Errius

User metadata
Rank l33t
Rank
l33t

The motor bearings not the platter bearings. If you put oil on the latter it will end up inside the chamber.

Is this too much voodoo?

Reply 6 of 15, by precaud

User metadata
Rank Member
Rank
Member

Well I have never encountered a recommendation to oil the spindle motor bearings of a HDD. But one of the four drives in this batch has a bad track 0 and so is expendable, so I decided to experiment, comparing the spin noise before and after putting a small drop of synthetic oil on the bottom of the bearing (see pic).

I have to say, the drive is significantly quieter after the oil.

Reply 7 of 15, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

I wasn't referring to the spindle motor, but the stepper motor. (Although... maybe that helps too?) The heads on these drives move like on a floppy drive, and suffer the same kinds of issues. Namely, if they can't move freely, they have trouble seeking accurately. I've seen some videos on YouTube where the heads have been seized from lack of use, and a little oil and some manual persuasion got them moving again.

Reply 8 of 15, by precaud

User metadata
Rank Member
Rank
Member
SirNickity wrote:

I wasn't referring to the spindle motor, but the stepper motor.

But that requires removing the cover and exposing the platters, doesn't it?

Reply 9 of 15, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

No. At least, not typically. All the ones I've seen have a visible bearing on the side that turns when seeking. Some even have a little "flag" on them, although that's probably an older design. Wish I could remember what videos I saw on YouTube where someone was maintaining their old drives, I would just link you to them instead of being a terrible proxy of barely relevant info. 😉

Reply 10 of 15, by precaud

User metadata
Rank Member
Rank
Member

I did a little test today to satisfy my curiosity. I wanted to see if the lubing and lowering of noise (hence friction) would have any impact on the number of bad sectors it found. I took the above drive that I had lightly oiled the spindle motor, and ran the LLF and post-scan with Speedstor.

And the answer is, nada. No difference.

Reply 11 of 15, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

What is the recommended procedure for putting a hard drive back into service after a long period in storage?

Low-level format.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 12 of 15, by chinny22

User metadata
Rank l33t++
Rank
l33t++

As above, Low level format is about the only thing you need to do, and you already did that.
You can get tools like spinrite that may repair bad sectors, but on a drive that old I doubt it'll make much difference.

If it was me, I'd start using it and keep an eye on it.
It's not like you'll have any important data on it that you'll loose if it does go. Likewise MFM drives aren't all that common

Reply 13 of 15, by precaud

User metadata
Rank Member
Rank
Member
chinny22 wrote:

If it was me, I'd start using it and keep an eye on it.
It's not like you'll have any important data on it that you'll loose if it does go.

Yeah, it's not clear yet what I'd be using it for. certainly nothing important.

I found this on Stason.org for the drive model in question.

 1. Number of defects
+-----------------+---------+---------+---------+
|Model | MK132FA | MK133FA | MK134FA |
+-----------------+---------+---------+---------+
|Max. per drive | 30 | 40 | 50 |
|Max. per surface | 15 | 15 | 15 |
+-----------------+---------+---------+---------+

2. Defect-free tracks
Head 0 and head 1 of cylinder 000 have no defect for all models at
the time of factory shipment.

So from the factory, they considered drives with up to 50 defects acceptable, and shipped them. Yikes. That seems excessive to me for a 44MB HDD.

The one that I tested after oiling had 45 defects total.

Reply 14 of 15, by derSammler

User metadata
Rank l33t
Rank
l33t
precaud wrote:

That seems excessive to me for a 44MB HDD.

Well, you need to keep in mind that there was no translation to hide bad sectors as with IDE later. IDE drives could have had hundreds of bad sectors when they left factory. But the end user would never see those, as they were mapped out.

Reply 15 of 15, by precaud

User metadata
Rank Member
Rank
Member
derSammler wrote:
precaud wrote:

That seems excessive to me for a 44MB HDD.

Well, you need to keep in mind that there was no translation to hide bad sectors as with IDE later. IDE drives could have had hundreds of bad sectors when they left factory. But the end user would never see those, as they were mapped out.

Interesting point... thanks.