Hard drive lubrication?

[beachycove]

Like many of you, I have a number of apparently failed, or irritatingly noisy small 50-pin SCSI drives (most 250mb or less). I have been going through them. Several of the ones I thought were dead have turned out just to need reformatting (Apple's utilities here were, however, useless; I used Silverlining v5.8.3a with much more success).

The more interesting thing to report is an experiment with a Quantum ProDrive LPS, 230MB. These were never particularly quiet drives, even back when new, but I had one of these that was whining badly, and so I regarded it as expendable. These drives shipped in 1992-93 machines like the later Quadras. Whining of the sort in question is caused by bearings, and as it's squeaky wheels that get the grease, I got out the oilcan....

A definitely dead drive was disassembled first to take a peek. That is something best not done to a drive you'd like to keep running. I discovered from the takeapart that it is possible to get oil down into the main bearing. Whether THAT is a good idea or not is, of course, another matter. Still, let's press on.

There are three stickers on these drives, each of which says, "Warranty Void If Seal Broken." Well, warranties are no longer an issue. The sticker to remove is the centre one. It has a small screw underneath, which, when removed, offers access down into the driveshaft. There is a lip and a seam (platters rotate around the shaft) separating this from the platters, so it looked possible to keep the oil where it belongs by using just a little and applying it carefully.

I removed the screw on the working, but whining drive and dipped a needle into some high quality machine oil a few times, inserting it down the hole into the driveshaft area. I then left the oil to soak in for a few minutes, and finally popped the drive in a Quadra 605.

I didn't expect this to work, somehow. However, the Quantum drive is no longer noisy. I am not sure how long it will continue to work, but I have been running it in a Silverlining testing loop for an hour, and it seems just fine.

I then tried the same thing with another of these drives that died a couple of years back. It was making a grinding noise. It started spinning, quietened down in about 90 seconds, and was actually recognized on the scsi chain. However, the heads were parked, and I could not get any further. I am not sure if other utilities like Lido would get me more joy.

Anyway, given the rate of drive failure these days, I thought I'd report this. It was only an experiment, but I'll try the drive again in a few days and see what happens.

 

[ClassicHasClass]

The real question is what lubricant you used.

 

[markyb86]

Did you forget the at the end of that question?

 

[ClassicHasClass]

No, I'm seriously interested in what he put in there. If it was a petroleum distillate, I'd be worried either about evapouration or gumming up, depending on how volatile it is. I would have used a molybdenum-type lubricant or something similar, which is hard to get, but lasts a while (it's what I used to use to lubricate floppy drive mechs, like the old Commodore 1541 Newtronics mechanisms that always seemed to stick).

 

[markyb86]

I'm just ribbing hehe. Not sure if its beneficial to the hardware or not, but I know my dad used WD40 at some point on our commodore floppy.

 

[beachycove]

Sewing machine oil.

 

[Trash80toHP_Mini]

On a peripherally related note:

ISTR that HDDs are assembled in a Nitrogen filled, clean room environment, sealed against Oxygen and obviously dust. I never questioned this, figuring Oxygen is deadly to anything subject to corrosion.

1) is this true?

2) if a drive is opened is it toasting, just waiting for the timer to pop?

3) can it be re-assembled in a home-mad clean room box, similar to a sandblasting box?

4) is Nitrogen available in quantities less, and more convenient, than a welding tank?

I've always wanted to make a clear plastic cover for one of the (Quantum?) drives with the stamped Aluminum pan cover. It's part of the long term goal of creating a ClearProtoHoaxMac of one of the Frog designs or a modified Mac design.

Lately I've been leaning toward a taller IIsi/BeigeG3 Mobo w/low profile PCI cards kinda deal. A clear lidded HDD would be a killer feature.

Were Teflon lubricants available at time of manufacture, what was the original lubricant spec?

 

[ClassicHasClass]

Sewing machine oil is a synthetic white oil, so it would be better than something volatile like WD-40, but I would be concerned it would be hard to control where it went. If it accumulated somewhere, it could gum up. At least a dry lubricant, assuming it doesn't fragment, would not do that.

 

[Gorgonops]

ISTR that HDDs are assembled in a Nitrogen filled, clean room environment, sealed against Oxygen and obviously dust. I never questioned this, figuring Oxygen is deadly to anything subject to corrosion.
1) is this true?

No.

Most hard drives, if you look hard enough, will have a tiny hole in them somewhere, and behind that is a very fine filter. That hole is there so the drive's internal pressure can equalize with the ambient atmospheric conditions. If it wasn't hard drives would have a nasty tendancy to bulge when shipped via air or taken up mountains.

(Without significant design changes drives would leak around the motor spindles and the points where wires enter and exit the drive case anyway, so I imagine the reason for the equalization filter is to make the air flow "predictable" and avoid air forcing its way through the bearings and drying them out/possibly loosening debris from them or the other possible egress/ingress points.)

Not to say you don't need a clean room, or close to it, if you want to put a clear plastic cover on a hard drive and expect it to function for long afterwards. People *have* done it for show by fiddling with the parts through a plastic bag or other jury-rigged expedient but I certainly wouldn't trust a drive so modified with *anything*. I'm sure you've seen the illustrations showing how large various sorts of debris are compared to the flying height of a Winchester drive head already, so there's no need for me to Google them up. Short answer is that even a drop of oil from a fingerprint, a below-average-size pollen grain, or a fleck of cigarette ash is more than enough to cause a crash.

 

[James1095]

In my experience, you can usually get away with opening a hard drive for short periods in a reasonably non-dusty location, a clean room is not really necessary. Do not get fingerprints on the platters and avoid touching the delicate head assembly. If any dust does get on the platters, a gentle dusting with a can of compressed air should be fine. Obviously I don't recommend opening up a good working drive, there is always some risk, but I have opened them on several occasions for various reasons and had them work for years afterward. Every drive I have taken apart has a small air filter inside it that will clean up dust that does manage to get in, hopefully before the heads come out of the parked position.

Do not operate the drive with the cover off though, at least not if you intend to keep it working for long. Running the open drive is where you get into trouble. Air currents and debris can interfere with the tiny air cushion that allows the heads to float over the surface of the platter and cause head crashes. You may be able to make a transparent cover or install a window into the lid, but anything with a different shape may alter the airflow inside the drive and lead to head crashes.

 

[onlyonemac]

Interesting thought on the clear lid! I've also been thinking... }

More related: I was wondering what people think about lubrication of the head assembly. I was going to try doing so when my drive wouldn't unpark, but that's was only after I'd wrecked the top platter with my finger when trying the temporary "manual unpark" fix :-/ . But anyway, what do you think?

Do not operate the drive with the cover off though, at least not if you intend to keep it working for long. Running the open drive is where you get into trouble.

I did that with my drive when I was trying to fix it, and eventually it made a funny smell, and then the heads stopped moving properly. Now, every time I switch it on, it goes "click chick, click chick, click chick" about five times, and then stops. Why??? That's not air currents, is it?

 

[beachycove]

Well, as promised, I fired up the drive again. It ran an Anubis Utilities test, top to toe, for several hours straight and did not report errors. I then powered down and the drive began to screech as it spun down. Now it has the Click of Death.

Most likely the oil is to blame, and that is what I expected at the start, but I suppose it is possible that the utility, which ran it real hard, did the old gal in.

Still ran pretty quiet, though, right until the end!

 

[James1095]

I did that with my drive when I was trying to fix it, and eventually it made a funny smell, and then the heads stopped moving properly. Now, every time I switch it on, it goes "click chick, click chick, click chick" about five times, and then stops. Why??? That's not air currents, is it?

I wouldn't think so, maybe it's a coincidence? Could also be static damage I suppose, if you managed to fry the head amplifier chip it would be unable to locate the servo track and just click.

 

[CelGen]

I have added windows to hard drives before. The key to keeping the drive useable is a ziploc bag, lots of canned air, and a triple format that writes across the entire surface of the drive. If there's any substantial contamination you'll pick it up on that last step because will get REALLY cranky.

 

[onlyonemac]

It's not static damage, because the heads stopped while it was running open; I didn't touch it when it broke.

 

[James1095]

Well you may have had a head crash on the head that reads the servo track. The click of death occurs when the head seeks across the platter but crashes into the stop without locating the track. I killed a couple of drives back in the day before I knew just how bad it was to run them with the lid off. It doesn't take much of an air current or dust particle to interfere with the tiny cushion of air the heads fly on.

 

[onlyonemac]

It's not a click of death. It's unparking, moving the head to the middle of the platter (not running into the stop) and then parking it again, five times over. BTW, it always used to move its head to the middle as part of the driver loading process before it broke, so it's only doing what it used to.

 

[James1095]

Well maybe a head crash corrupted the driver or some other data the drive needs to initialize? No way of knowing for sure.

 

[techknight]

its crashed. the repeated seek and fail are bad heads for sure.

The heads cannot touch, even together while off the platter assembly. I dont know why, but they just cant. it kills them. HDD science is still somewhat a black art. Scott moulton seems to know alot about them and made several defcon conventions and they are on youtube.