back to article Abracadabra! Tales of unexpected sysadmagic and dabbling in dark arts

Congratulations, Reg readers – you've made it through the week, and those of you in England and Wales are now headed into a three-day weekend. To celebrate, we've got a bumper edition of our weekly On-Call column of tech support crises. This time, it's a selection of tales from sysadmins who looked to the "dark arts" to fix …

Case sensor

Oh Timmy,

probably a case sensor that needed a bit of persuasion that the case was closed before powering up.

My other halfs iPhone screen & touch sensor has been misbehaving. i gave it a few taps in the right place and it worked flawlessly for a couple of days. I'm just hoping its misbehaving when she takes it into the apple store and they don't fob her off with my fix!!!

41
0
Ogi

Re: Case sensor

Once I did the same thing with a tower box, but it involved a fist bang on the top rather than a mallet.

The PSU fan was on its way out, and would get stuck when the machine was turned off for an extended period of time. So when someone turns off the PC and leaves it off overnight, in the morning the PSU fan won't spin up unless you give it some taps to get it started ( I guess the PSU would detect the fan not working, and cut out to prevent overheating/fire hazard).

The PSU was mounted at the top of the tower, so the lazy "fix" was to power on the box and thwack the top of the case and get the fan spinning, after which it would work until left off for a few hours.

Fun fact was you had to get the timing right as well. If you wait too long before the power on and thwack, the PSU would cut out before the fan started spinning. Likewise you power on and thwack too early, the fan hasn't been powered up and won't start spinning, and then it would cut out. So for those not in the know of what the problem was, it seemed like a magic touch. Hearing other people in the office thwacking the box before they would come and ask me to do it was quite funny.

It was a "low priority" fix, primarily because the desktop was scheduled to be replaced in the next hardware refresh, and the PSU replacement was a fiddly job, involving removal of the motherboard to get the PSU out.

This kind of fixing is quite common, so much so it even has its own term "Percussive maintenance", which I first heard on a (now old) comic:

http://ars.userfriendly.org/cartoons/?id=20010326

38
0

Re: Case sensor

I had an Apricot 486SX PC where the HD had probably got a coil gone in the motor so once in a while it wouldn't spin up. Easy enough, turn it on and just give the box a bit of a rotational tweak. Again, had to be timed right between power coming on and POST reporting no boot disk, but that wasn't too much of an issue.

Way back we'd had an HD that was really sticky. Can you run a disk without the top of the case? We could then but only long enough to retrieve the data having got it moving "by hand".

25
0
Ogi
Pint

Re: Case sensor

> Can you run a disk without the top of the case?

Heh, I used to to do that with old disks that were destined for scrapping (they were fine, but company policy on data destruction meant we had to wipe then destroy the drives). See how long they would last with the top off. In some cases pretty darn long actually (a couple of weeks of badblocks testing), assuming the area they were operated in wasn't too dusty.

Once I had the bright idea to fit a plexiglass top to a disk and use it in my casemod (which was a completely clear plexiglass mini tower case, complete with blue LEDs, which had just come out on the mainstream market, so had to use some).

It worked well for a year or so. but my limited fabrication skills (a Dremel and my hands) meant I didn't get the fit perfect, so dust and dirt would get in between the gaps, and the drive eventually failed, to be replaced with a standard top one. Still, it looked amazing watching the speed the arm would move around on the spinning platters, especially if it would start swapping. I found it mesmerising.

So, in short, yes, they can work, but only for a short time, so not very useful for actual data storage/access.

Also, modern drives with many platters stacked close together, or the ones filled with helium or whatnot, probably can't be run with the case off at all. The drives I talk about above are 40-80gb, so relatively more hard wearing than modern multi TB drives you can get.

Icon, cos its Friday!

42
0
DJV
Silver badge

Re: Case sensor

My first hard drive for an Amiga 2000 (a Quantum 80MB, if memory serves) started failing to turn on after about 2+ years. The solution was to open up the case pull the hard disk and attached card from the slot, give it a good shake and shove it back in. That usually fixed it. It was a regular occurrence for another couple of years when I could afford to replace it with a HUGE(!) 540MB one.

25
0
Happy

Re: Case sensor

The drives I talk about above are 40-80gb

What are these gb of which you speak?

I'm going back to the megabyte days!

First HD I had in my own PC was 20MB, stretchable to 32MB with an RLL Controller*.

*They packed more data on the outer edges of the disk where it spins faster, ACT had done something with floppy drives that appeared to play music as they actually varied the spin speed.

22
0

Re: Case sensor

"Percussive maintenance" was also known as "Gentle Persuasion" round here. Any large heavy object will do.

15
0
Silver badge

Re: Case sensor

I've often found with faults like that you dont even need the violence, sometimes the machine just fancies a walk around the building with you and when you get it to your workbench to start invasive surgery the machine will get all squeamish , start working perfectly and demand to be returned to its user.

If you're in a hurry a good slap around the casing will often persuade the lazy **** to get back to work without the walk. (the pc, not the user)

27
1

Re: Case sensor

Not Computer related but one of the engineering companies that I worked for had a Refrigeration department which did installs and maintenance of commercial equipment and the occasional domestic job when work was slow.

Some of the fridges used a coolant that would on occasion separate out and lose effectiveness. This was fixed by loading the fridge into the back of the works van and going for a drive around The vibration / acceleration / deceleration would churn everything up and hey presto one working fridge.

27
0

Re: Case sensor

Many, many years ago I was the service manager at a local computer store.

One day we had this very nice older gentleman come in with a really old PC. He was an accountant, and (you can all see this coming) had no backups. He was about in tears. There was something like 15 years of his clients records on the hard drive that just failed.

We mentioned the option of sending in the drive for data recovery. In those days there were only two reputable companies in the US that did data recovery, and the bill usually ran between $5,000 and $15,000. He told us that he couldn't possibly pay that kind of money (this was in the early 90's).

I told him that I would take a look at the drive, but that we didn't have a clean room, and we might make the drive unrecoverable by opening it. He told me to go ahead and try.

It seems that somehow the drive managed to seek too far towards the outer edge of the platters, and the heads went over the edge. Since they are kind-of spring loaded, they were stuck. So, I very gently spread the heads apart and slipped them back onto the platters. What do you know, the drive worked again. The fact that I opened it in a dusty back room meant that the drive would not run very long, but it ran long enough for me to get the data off of it.

We sold him a new drive, copied his data to the new drive, and he was off and running. The joy on that man's face made up for a lot of the BS from cranky customers we otherwise had to put up with! I kept a copy of his important data in our safe on a zip disk for quite a while, as I figured that even after this event, he would never back up his data.

42
0
Silver badge

Re: Case sensor

"Percussive maintenance" was also known as "Gentle Persuasion" round here. Any large heavy object will do.

Such as the boss's skull?

11
0

Re: Case sensor

In the WAAYYY back in the day, you did the same with the Sparc IPXs and IPCs. If for some reason the system was shutdown over a weekend, you did not hit it with a mallet, you dropped it on the desk.

It seems the lubricant used on the drive spindle would congeal when it cooled off, specially if the system was a couple years old, so a good slam on the desk top while powered on would crack it free and it would boot up nicely.

11
0
Silver badge

Re: Case sensor

I still remember opening the cases of computers when "sticktion" was a problem, reversing a screwdriver and gently tapping the outer edge of the hard drive. That was on IBM compatibles. I would have been seriously upset, given how much I'd paid for them, if the drives for my Amiga 1000 (40 MB, wheee!) or A2000++(80MB) had that occur.

8
0
Anonymous Coward

Re: Case sensor

Percussive Maintenance:

Frequently have customers who want us to make sure that we totally erase their old hard drives, especially if they are failing. First thing we do is ask if they want them back or not. If they don't, then lots are drawn, and the winner grabs a big hammer and takes the drive out back to smash it to pieces. Once the platters have broken into tiny fragments, the risk of anyone bothering to try and recover data from them approaches zero point zed zed zed zed zed. And then we charge them for "proper disposal".

6
0

Re: Case sensor

"even after this event, he would never back up his data."

Pretty safe bet. I am rather impressed you were able to re-load the heads without damaging them, or them already being damaged by seeking off the edge of the platter. I am also curious as to what kind of error or event would cause a seek off the edge.

4
0

Re: Case sensor

>First HD I had in my own PC was 20MB

God yes, same here. The version of DOS I had couldn't even conceive such a large volume could exist, so I had to split the disk in a 16 MB and a 4 MB partition.

And I managed to play "The Secret of Monkey Island" all the way through, even though my 8086 PC only had a CGA graphics card *and* a green on black monochrome monitor. At some point in the game, the player gets a list of ingredients he needs to collect, written with multicolored characters. On the 4 color CGA display, different colors were merged, so the writing wasn't recognizable - only a few pixels of each letter could be seen. I didn't even realize it was just an issue with my bottom of the barrel display. I thought it was another puzzle, and the list was intentionally written in some secret alphabet - and cheerfully spent some quality time decoding it.

Fun times!

11
0
Silver badge

At Alan W. Rateliff, II, re: head crash.

On a clear disk you can seek forever. =-)p

14
0
Silver badge

At Ogi, re: percussive maintenence.

You got an upvote for the UserFriendly pointer. Cheers & join me at the tavern so I can buy you a drink! =-D

4
0
Silver badge

At Roger Greenwood, re: heavy objects...

Like my ex boss' fat arse? =-)p

2
0
Silver badge

At Prst. V.Jeltz, re: slapping.

Sorry for the downvote but I want to slap the user instead. Let me slap the user & I'll upvote you in glee. =-)p

3
0
Silver badge

Re: Case sensor

"This kind of fixing is quite common, so much so it even has its own term "Percussive maintenance", which I first heard on a (now old) comic:"

It goes back even further than that. There's a cartoon referencing it in one of my archive magazines of "Radio Television and Hobbies"(August 1964)

Extra credits to anyone who can remember what that magazine changed its name to.

5
0
Silver badge

Re: Case sensor

"the HD had probably got a coil gone in the motor so once in a while it wouldn't spin up"

HDD motors have very low torque. The cause was a little wierder than that - called "sticktion" by many.

Grease from the motor bearing would get past the end seal, vaporise and deposit on the HDD platters. It wasn't much, but it was _just_ enough (when cold) to make the parked drive head (they used to park on the drive platter, not off the edge) stick to the platter(*). Of course this only happened when the head settled onto the deposited grease vapour when warm.

Giving the drive a sharp twist on the platter axis was usually enough to jar the head loose and allow the drive to spin up, but people would do all kinds of violence to get them spinning(**)

Startup friction is always a little higher than moving (going from stationary to moving friction before the heads get airborne) and makers used to pulse the drive motors quite hard to ensure they'd start but as drives got smaller this got less practical to do. Quantum's early 1990s 40MB scsis (used in Macs) were pretty notorious for this - probably pushing awareness into the mainstream.

(*) Sometimes they'd really stick hard to the platter, and then the spinup after the "twist" would result in the unmistakable sound of heads tinkling free on their wires.

(**) Hitting them with some kind of hammer or dropping the drives (or entire machine) was a favourite - and generally shagged bearings when overdone. Bear in mind that HDD platter assemblies have a motor at one end and are usually unsupported at the other. That's a lot of shock twisting torque on a heavy cantiliever when the drives are mounted vertically (and when horizontal sometimes the drive motors would simply detatch from the HDD body)

7
0
Silver badge

Re: Case sensor

"If they are failing. First thing we do is ask if they want them back or not. If they don't, then lots are drawn, and the winner grabs a big hammer and takes the drive out back to smash it to pieces."

Alternatively there's this: https://youtu.be/wb3Xa1h_RqM?t=24s

Satisfying to watch

3
0

Re: Case sensor

At home it is known as BFI, Brute Force and Ignorance.

1
0
Silver badge

Re: Case sensor

May have been stiction. One batch of 1GB IBM Spitfire disks, had the wrong lubricant in the drive bearings, which would vaporize and condense on the platter. When the drive was stopped, and the head parked, it sometimes stuck to the condensed lubricant, and would prevent the disk spinning up.

A quick shock would free the head and allow the disk to spin up.

I believe some other drive manufacturers also had this problem as well.

1
0

Re: Case sensor

Actually, I always knew it as BFFI.

0
0
Boffin

Re: Case sensor

Similar thing happened to me when I was running the Technical Publications Department of a large electrical manufacturing company in the Midlands. We had several Xerox desktop publishing workstations networked on Thick Ethernet, and one of them developed a sticky hard drive, which wouldn't spin up when switched on. I swapped it for the HDD out of my workstation, so I had the task of giving it a slight rotational wiggle every morning, and the other user didn't have the problem any more. I also wrote a small Terminate and Stay Resident (TSR) program to install on all of the drives to automatically park the heads after a few seconds of non use, as the HDDs were not self parking in those days. As the whole department was made redundant in 1992, I have no idea what happened to all that equipment.

1
0
Facepalm

Re: Case sensor

PSU fan won't spin up unless you give it some taps to get it started

Had a similar issue except it was an old saab car. Wouldn't start, roadside assist turned up lent under the car and gave the fuel tank a hard wack with a rubber mallet and it started.

It turns out the problem was the fuel pump in the fuel tank was on its way out and one of the classic problems is it wouldn't start. Bugger.

3
0

Kudos to the guy who came up with assumed name "Timmy" for a mallet story.

78
0
Pint

Gal, not guy.

I'm guessing Ms Hill gets the credit there.

Kudos to you for spotting it though, that one totally passed me by.

=Bank Holiday Pint

23
0
Silver badge

Pataphysicality

The mallet-wielder's name was probably Maxwell.

14
0
Anonymous Coward

Re: Pataphysicality

No. He would have had a hammer

(Probably a silver one)

27
0
Silver badge
Coat

>Kudos to the guy who came up with assumed name "Timmy" for a mallet story.

I'm just very pleased that there are no pictures of him in a yellow polka dot bikini...

What he probably needs to cover it all up right now -->

10
0

If I had s hammer

I'd hammer in the morning.

(But it still wouldn't annoy those builders with the pile-driver 3 streets away)

11
0
Bronze badge
Coat

@Oddbodd.. ref: Timmy.

I have a wood planer called Nigel if that qualifies. :) PP

3
0
Stop

For those in the UK...

Isn't a long weekend up here :|

28
0

Re: For those in the UK...

Scotland?

5
0
Silver badge

Re: For those in the UK...

I have the benefit of working in Scotland for an England based company, so I do get the Monday off. Most others don't though

</smug mode>

19
0
Silver badge

Re: For those in the UK...

Ah is this the trade-off bank holiday for the ever-sensible 2 day holiday at Hogmanay?

17
0
Headmaster

Re: For those in the UK...

Sadly, no. Scotland had their bank holiday the FIRST Monday in August, the rest of us have to wait until the LAST one.

10
0

Re: For those in the UK...

And going beyond that, parts of Scotland ignore bank holidays completely, meaning my work is open all year round except between 24th December and 2nd January.

So we're open Easter Monday, Good Friday, May bank holidays, etc.

11
0
Silver badge

Re: For those in the UK...

A former employer did the open bank holidays thing - absolutely brilliant

A) it’s really easy to get to work on the bank holiday as most folks are still at home; and

B) you get the extra days off when everyone else is at work. :)

24
0
Happy

Re: For those in the UK...

So do I but we observe local holidays.

3
0
Silver badge

Back in 1981 one of the regular problems we had, about 1 or 2 AM in the morning, was the top AN/WSC-3(V)1 satellite communications radio going TITSUP. I'd go up there, pop the case using a big-ass flat-tip screwdriver. Every time, it'd pass all 24 of the BITE (Built-In Test Equipment) tests with flying colors. I didn't have a clue what was happening, nor did anyone else. The third or fourth time a thought occurred to me. (That does happen from time to time.) Each time I popped the case, I set that big-ass screwdriver on top of the transmitter module. So, I dutifully waited for the next time this came up.

Popped the case, set the screwdriver on top of the transmitter module for a minute, picked up the screwdriver, shut the case and screwed it back down. I then went to the duty radioman how it was working and he gave me back "working great!" I dutifully let all the other duty technicians what I did and perhaps what it means (some coil or other magnetic component intermittent). It was added to our lore.

[We were a rather arrogant lot, for good reason actually. We used to go to other divisions, ships, even other ports/bases to fix the unfixable. Our motto: "If We Can't Fix It, It A'int Broke!" Right on the door to our shop.]

30
0

@Jack

Seen this quite a few times. Equipment that fails, open the case start fault finding and the whole thing works perfectly.

In all cases it was heat, either the object was located where it couldn't get enough air flow to cool it or one of the components was starting to be come a little heat sensitive. A can of freeze can help to sort that out.

1
0
Silver badge
Headmaster

Couldn't close filing cabinet.

"But the fix that brought Mike the most amusement was when a manager couldn't shut the top drawer of his filing cabinet. I noticed that the bottom drawer wasn't quite closed"

I don't think it works like that. He wouldn't be able to _open_ another drawer if any others are open. This is to stop the thing toppling over. Nothing stops you closing the doors, as our friend did with the bottom one.

3
27

Re: Couldn't close filing cabinet.

The one next to my desk does work exactly like that. Which is super annoying as, when left alone, the bottom drawer has a tendency to roll itself out just enough to activate the lock.

30
0
Bronze badge

Re: Couldn't close filing cabinet.

I've had it happen to me on an older cabinet - the drawer was only slightly forwards (due to Stuff in it, I think), but enough to push on the locking bar that stops it opening, and that in turn caused the top locking bar to block the top drawer from closing.

Might have been a flaw with the design, or just age, but it can happen like the story describes.

33
0
Silver badge

Re: Couldn't close filing cabinet.

If you can find a way to get two drawers open at once (and I suspect the office move mentioned in the story might have had something to do with that), the the locking bar that stops one from opening multiple doors does usually prevent them from being closed as well.

18
0
Anonymous Coward

Re: Couldn't close filing cabinet.

>if you can find a way to get two drawers open at once

Just tried that - simples. just pull them both open together. But then the locking mechanism didn't prevent either drawer being closed.

Clearly, it depends on the design of the cabinet.

9
0

Page:

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2018