Ask a builder to fix a server and out come the vastly inappropriate power tools With Friday upon us, the green-thumbed among you may have some gardening planned for the weekend. Before you dig into that chore, The Register presents another instalment of On Call, our weekly reader contributed tale of being asked to get into the weeds of malfunctioning tech. This week, meet a reader who asked to be …
They started putting protection on the inputs of the 4000 CMOS chips at some point; someone told me it was similar to Zener diodes. They were a lot less likely to pop by the time I got to play with them. They always were moisture sensitive; presumably from the ultra high input impedance.
"Also static damage isn't, to coin a word, binary ... it can degrade the transistors on the chips to the point they're more likely to fail earlier"
Thanks for this, I have seen this a lot. We had quite a few big data centres each with a dedicated engineer at each site, many of the systems upgrades involved handling the electronics. One of the sites had many random system failures ongoing for a couple of years, all other sites were reliable, this was a huge problem. One day a more-experienced engineer was on-site and saw the local dedicated engineer failing to follow the anti-static precautions. So they swapped out all parts that the previous dedicated engineer had touched (10s thousands $) and there were no more problems.
Static damage frequently causes degradation, perhaps components shift out of their specified tolerances, not always obviously destroying a part, but when it is under a specific load the shift away from the specified tolerance eventually conflicts and causes an overall failure. One slacker who thinks they know better break things for others.
Indeed dress sense matters. At one data centre in the very early 80's we had machine room operators, who changed tapes, watched the pretty lights, and occasionally types a command into one of the many consoles. We had a lot of problems with our Cray-1, which was running some very long jobs, often for days at a time. It would just stop. Needed a reboot and was fine, but we lost the whole of that job.
After some investigation we found the a number of the lady operators were walking past this iconic and very tactile machine, and just ran their fingers over the outer panels. It seemed that the use of a nylon overall, and nylon tights created quite a static charge.
The unions were furious when all female operators were banned from going close to the cray, but when the data centre manager explained that the alternative would be a check that they were wearing silk underwear it was decided to go with the first option.
Yup. Back in the days of 4000-series chips a few electronics mags ran stories about ESD which included photos of actual chip-level damage (usually a chunk blasted out of the side of a track, leaving the remainder to act like a fuse and pop weeks/months later)
It wasn't just CMOS chips either. A lot of RF frontend FETs will pop if you look at them funny and it wasn't unusual to have them supplied with a pull wire shorting all the legs together. This was meant to be removed AFTER soldering into position and I know a production line that had to redo at least a weeks worth of radios due to one employee insisting he knew better (these things weren't cheap either, so it was a financially painful exercise)
Technically, "early days of computing" is all electric "valves" instead of transistors, or perhaps the extraordinary Engines of Charles Babbage in the 19th century; hand operated, I think.
Valves had a fairly terrible rate of failure when you were using a large number of them at once, heat was a factor too, but I think you were in more danger from your valves than your valves were in danger from you.
It's often repeated that in winter during the Second World War, women's underwear was to be found beside or on the secret British electric code-breaking computer, as it was in the only warm room on the premises. It couldn't help it. I think I saw a report that in hot weather, women's underwear and not anything else was found on the women working in there, as it was still the warmest room. I expect none of this is relevant, but I find it interesting. ;-)
When I worked with testing and fixing electronic equipment as day job then static used to an issue. We'd work on rubber mats for protection and movement on those could generate quite a bit of static (sliding back the chair to change something in rack would generate a painful charge).
When have I done a bit of electronics at home static has never been an issue
One of the things about damaged caused by static is that it doesn't always cause an instant failure.
The over-voltage caused by a static discharge can punch a tiny hole in the insulation between the layers of a chip that allows metal and other ions to migrate over time, and this eventually leads to the device failing as some unpredictable point in the future.
I have recently retired from a career in IT hardware servicing.
There was always a high focus on following ESD procedures correctly (wristband and anti-static mat etc).
I would sometimes be challenged for not following the guidelines.
I would reply that I knew that I was safe because of the reason I lost my previous job at the bus company.
They said I was a very poor conductor.
I did once corrupt the firmware in a print server circuit board with static. Walked across an office carpet and took the board from a colleague and saw the spark jump as I reached out.
Reapplied firmware and all seemed well again. Guess that's why they call it flashing firmware...
I once TRIED to fry a NEC printer with static arcs - we were shooting 1/2-1" arcs from a small tesla coil into the I960 chip's legs for 20 minutes and it was still printing fine! End result was, the printer lived. The SDRAM chips were similarly unaffected, at least in the short term.
I too have never, to the best of my knowledge, killed something from static electricity and I go back to Atari 8 bit computers.
That doesn't mean however that I haven't damaged something with static electricity, and it later failed and I didn't connect the two. I don't go all out with wrist straps and such, but I always try to touch a metal surface before touching sensitive areas like a motherboard or DIMM, just in case.
Funny thing though - it seemed like static electricity was WAY more of a thing when I was a kid. I remember getting some BIG shocks sometimes growing up, but I honestly can't remember the last time I got even the tiniest shock. Last I can remember for sure was that I used to get one when I touched the outer door handle of a previous car sometimes, but that was 5+ years ago, and I think that was more likely an electrical fault in the door to be honest.
Are children more conductive? Did my parents house contain a hidden Van De Graaff generator? One house had shag carpets, but even when our family moved to a new house that had regular carpets and no shag I remember getting some pretty good sized shocks. House I live in now has plenty of carpet, but I could walk around in socks dragging my feet and not elicit even the tiniest of a shock touching a grounded surface like the screw on an outlet cover or my cast iron stove. Even when I used to regularly visit my then girlfriend in Vegas, one of the driest places around that you would think conducive for static electricity, I don't recall ever getting a shock in her fully carpeted condo.
Did the CFC ban to save the ozone layer have a side effect of removing static electricity from the air or something? :)
but I always try to touch a metal surface before touching sensitive areas like a motherboard or DIMM, just in case.
Yes, if it's not something you are doing every day, all day, AND you understand what you are doing, you can be perfectly safe without using the proper and approved precautions :-)
"Funny thing though - it seemed like static electricity was WAY more of a thing when I was a kid. I remember getting some BIG shocks sometimes growing up, but I honestly can't remember the last time"
Nylon was still the new "wonder" material in the time period you are speaking of, I also lived through it. It wasn't unusual for underwear, trousers and shirts to have a nylon content if not actually 100% nylon. It wasn't just women's clothes. And plastic "patent leather" shoes.
Likewise, nylon carpets in the home and workplace.
Some of that still hasn't left us, even if it's not always nylon. Other man made static generating materials are available :-)
Not quite related but once, long ago, when I was eating crackers and peanut butter while I was on my computer, I accidentally bumped the butter knife as it sat on the table and it spun and went right into one of the USB ports on the front of the laptop.
There was a loud snap and the laptop shut off. I pulled the knife out and powered it back up and all was good.
It was such a fluke that it went right in the USB from about a foot away, I still laugh about it.
I have never seen a bit of equipment killed by static electricity.
ESD protection is a feature in chips these days. Not that equipment is invulnerable now, just much more robust than it used to be. Go back to about the 486 or Pentium 1 era, and motherboards would just suddenly die if you touched them before putting on a wrist strap... micro static you don't even feel, and it is just gone, black screen and no activity on power-on. Making it worse, components were quite a lot more expensive back then, so it was a significant loss whenever it happened.
Don't blame you as I thought the same although I believe IBM had training vids about it. Then in about 2012, under someone's desk just shifting their PC because of graphics issues on the graphics card. I caught the back of the case, due to being in carpet. Was a tiny pop and her PC turned off. Ah shit. Killed the graphics card.
Talking of which, a colleague was sent on a project to a flour mill. The IT manager of the customer set up a trestle table wih power strips in the production hall for the consultants to use.
"It was dreadful, smell of mould and flour and the air was thick with flour dust!" He recalled.
Cue H&S appearing on the scene & evacuating the whole site, before shutting down the power to the section of the building where the table was! The IT Manager had just ignored the overhead pipes transferring flour from one side of the building to the other, which weren't 100% air-tight and were constantly letting flour fall down onto the table...
"Sounds amusing unless you happened to be in it."
Indeed it does! From the linked article:
"In 1981, there was a dust explosion at the Bird’s Custard Factory in Banbury. A hopper – a container used to hold particles that have been collected from expelled air – became overfilled, creating a dust cloud of corn flour that ignited due to nearby electrical equipment. The explosion blew the roof off the building and 9 workers were injured. As custard is made when heat and water are added to custard powder, the water from the fire engines that came to put the fire out created gallons of custard inside the building, which then came pouring out."
"Cue H&S appearing on the scene & evacuating the whole site, before shutting down the power to the section of the building where the table was! "
Not surprised - dust explosions are a major hazard in flour mills!
https://www.ukflourmillers.org/dustexplosionsfire
--> For the weekend!
Once had to do a site refresh at a remote office that was getting new servers, switches, UPS, the works.
But we also had to replace the little 12U rack with a full height one, so it was appropriately ordered and delivered, and waiting for us when we got there.
Problem was the network cable people had routed all the cables through a little hole in the top of the rack, and wired them into the back of the patch panel, meaning we simply couldn't get the patch panel out of the rack without ripping all the cables out and having to get it all re-terminated.
One trip the site workshop saw the application of a power saw to simply cut the top/rear of the rack out so the cable loom could be removed with the panel intact.
Luckily the new rack had a full size slot on top we could easily pass the patch panel through.
"Problem was the network cable people had routed all the cables through a little hole in the top of the rack, and wired them into the back of the patch panel, meaning we simply couldn't get the patch panel out of the rack without ripping all the cables out and having to get it all re-terminated."
I'm guessing that simply walking off and demanding that the project manager "fetch the buggers back" to do the cabling job properly wasn't an option.
Never used keystone, so can't speak on that. I use Panduit Minicom, in wall outlets you get the issues with plastic (Although not as often as with keystone it seems like), but on patch panels they are retained by a piece of metal with a notch in the plastic. Very solid and if it doesn't hold well anymore you can bend the metal a bit to make it tighter. Solid to the point where you can barely remove them without the tool (a small flathead will do in a pinch).
Also, they can be a pain to service. You'll usually end up taking out the whole thing if you just want to work on one cable, and they are prone to be abused by idiots who strip back too far or terminate them with different wire lengths.
And don't even get me started on random cable breaks...
I get your point on keystone and clips breaking off sucks for sure.
At a previous job, we had a lot of problems with brittle keystone jacks. The previous Network admin had cabled the building about five years before I started working there. There were a bunch of issues with how they ran the network cables (bundles of cables laying directly on top of ROWS of florescent light cans, etc.), but the biggest problem was the keystone jacks. The same jacks were in the wall plates and in the patch panels. These were good, name brand jacks, but there was something "wrong" with how the plastic was made, and they became very brittle. At one point, you could actually crumble them into tiny pieces in your bare hands.
I had been pushing to have the building re-cabled for a while, but because of financial issues, it was continually postponed. When things finally got to the point where it was impacting company operations to a significant degree, I decided it was time to make a point. I kept a few if the really brittle jacks that I had cut off. I went to see our CEO, and in front of him, I crumbled a keystone jack into tiny pieces in my bare hand. I handed one to him, and said "here, try it yourself". His reaction was simply "wow!". I told him, these jacks are what our entire network is relying on. His response was: "Okay, let's get some quotes on re-cabling..." About a month later, we had all new cabling.
There was a programme on TV a couple of nights ago about an f-ing builder for the super rich.
He had specified programmable electrics, so any switch could be programmed to control any function, so that meant miles of wires... unfortunately the guys had got a little too creative when they got to a partition in one of the rooms.
Instead of stopping and asking if someone could punch a hole or what they should do, they simply looped the dozens of cables round the end of the wall and left them hanging in full view
Ah yes, I've worked on such building control systems aka 'dumb home'... The switches are godawful and plasticky for the money, and yes, there are MILES of cable literally, every single light or bank of light has to be wired back to the distribution board. We had around 1200metres of mains wiring (1.5mm thank gawd, larger would have been unmanageable) and nearly double that of cat[x]. Switches only used RS422 or 485 (can't remember which) so only needed cat5, but everything else got cat6A. The building inspector was actually worried about the structural integrity of the joists due to the number of holes, so we had the builder double them up. Fun job but absolutely ludicrous price.
Very hair-raising experience
I once had someone want to weld components to a printed circuit board of a stage light controller I had built and maintained for a student drama society. I suggested soldering was the better option. As she was doing some practice soldering on a little spare PCB, she kept confusing the terms, so when she next called the soldering iron a welding tool I couldn't resist going "KZZZEERT!" when she next put the soldering tip on the PCB. Apart from confusing the terms, she was very capable, and did an excellent job. I kept up the occasional "KZZZEERT!" until she got the terminology right.
> weld components to a printed circuit board
It is just melted metal. The distinction between lead-base and iron-base, and filler vs base metal, is not observed in all languages (even all trades in one language).
If she brings a LINCOLN stick arc welder to the PCB table, stop her.
"I kept up the occasional "KZZZEERT!" until she got the terminology right."
The terminology is very important. Lot's of people use "loctite" when they mean "threadlocker". I made a bunch of money when some production staff went to the cabinet to get some "Loctite" and grabbed cyanoacrylate (super glue) instead of the threadlocker they were supposed to use. The company brought me a pallet of messed up bespoke product to see how much of it I could recover as the parts were very expensive. Loctite is a brand and they make loads of different adhesives and stuff. I also used a different brand of threadlocker in my own products as Loctite was too expensive if I couldn't find a bottle on eBay cheap.
"until she got the terminology right."
I remember back in the day when Americans always seemed to refer to a soldering[*] "gun" when doing PCB work and the visuals in my mind were horrifying!
* Also usually pronounced with a homeopathic L sound in the middle, something you can't quite tell is actually there when spoken aloud :-)
"I remember back in the day when Americans always seemed to refer to a soldering[*] "gun" when doing PCB work and the visuals in my mind were horrifying!"
Soldering guns are great. I find at least one at every estate sale, buy them for $2-$3 and flog them off on eBay for $20. I was tempted to keep one of them that was shaped like a cowboy revolver but decided the one I had is more than I need. The electronics bench has a Weller 50w station as standard fare. I also have an 80w station and a box full of classic 24w stations along with loads of tips and replacement parts. I used to find deals on eBay when companies would put all of their soldering stuff on one auction and I could get a great deal on what I wanted out of the listing and either sell or give away the rest. I don't see those sorts of deals online that often anymore so I keep a sharp eye out on what's being offered at local estate sales. One of the last ones had a massive Lionel train set up that I traded taking apart and hauling away for the track, a bunch of accessories and a load of good lumber (it took up the whole room and was 3 tiers). The garage was also loaded with tools so I scored some good finds there.
Yes, somebody that attacks a PCB with a soldering "gun" is a menace.
We had a guy that solved a desktop stability problem with nail trimmers.
See, the users in that department were slightly technical, and had figured out they could overclock the Pentium 100's they'd been issued by changing a jumper. Sure, their build times dropped, but the machines became a little flaky. You'd think that folks making unauthorized modifications to company hardware might keep it quiet, but not these folks. They openly admitted to overclocking and *whined* about their applications crashing.
On about the third or fourth trip over there where he found a machine behaving badly because they'd changed the clock speed, he noticed a pair of nail clippers on a desk across the way and was inspired.
Snip! Off came one of the pins in the jumper block, and with it, the ability for the motherboard to be clocked to 120mhz.
"Besides, what's the point of screwing with someone if they don't know they're being screwed with? :)"
You could have removed the metal part of the jumper, leaving just the plastic bit. And then had a quiet giggle when they moved the jumper and claimed it was "so much faster" :-)
Sometimes it's even more fun to watch the placebo effect on "semi technical" people. Maybe tell them later, if you really must see the look on their faces :-)
> Even better would be to snip just one pin, and put the jumper back in place so it looked like it was still overclocked.
Nah, you have to remember that in this era that clock speed was displayed (along with other config stuff) for several seconds on bootup, so they woud have known it was no longer overcooked.
Funnily enough, that relates to a situation my wife and I had with the first PC we bought after we got married and couldn't just use my dad's any more. Got the new machine, set it up, turned it on... wait, that's not the clock speed we paid for! (IIRC it was 60MHz instead of 90MHz, or 90MHz instead of 150MHz. A fairly substantial amount slower than it should have been, in any case.)
A little reading of the motherboard manual later to check on which jumpers were involved and how they should be set, and breaking the "do not open" sticker to look inside, revealed the problem: there were two pairs of jumpers for the CPU speed and for this configuration both pairs should have been set. And indeed they were, but the wrong way around - like = instead of ||. Took them off, put them on the right way around, problem solved.
I fixed a 3.5 inch floppy drive with a similar application of sharp blades.
Brand new PC (tower case) wouldn't write or read fro floppy. Year about 2001. My uncle gave me a ring about it two days after he bought the thing.
Much messing about looking for loose cables, at some point we had the thing on with the case open and its front off. To our surprise, floppy worked fine. Front back on, floppy failed again.
The front had a plastic facade in front of the floppy and a thingamajig to allow the operator to push the eject button on the actual drive. The thingamajig was about two millimetres too long and would keep the button half pressed when the front was screwed on. Snip snip and all was well.
At one large company I teched at, I was warned by my supervisor about a particular department-project rep who ate erring techies for lunch. I did a few projects for her, she was reasonable, and we got along fine.
On one major office remodel/offices move/many-PCs replacement project she supervised, I was working on a PC which had a 5.25" floppy diskette drive and a 5.25" streaming backup tape drive. These were mounted on drive rails specific to the old computer, and which were not compatible with the new computer. Attempting to unscrew the screws holding the rails in place, I found I could not.
She saw me walk out of an office, look around, walk over, pick up a very large hammer, walk back into the office, heard WHANG-WHANG! ... WHANG-WHANG!, saw me walk out of the office, put the hammer back, and return to the office.
She didn't say a word. I guess she'd learnt she could trust me -- or was too flabbergasted to think of what to say.
(Icon for potential tool-misuse hazard.)
Dating myself badly here.
I once saw an IBM CE (Customer Engineer) fix a misbehaving IBM 3330-2 disk drive with the magic hammer.
It was spinning up to its working speed, but not becoming 'ready' after I had changed the disk pack (All of 200MB in capacity and weighted in at 10Kg), so he pulled the whole assembly out at the back of the unit on its rails and gave a solenoid a good wack with the hammer, at which point it became ready.
Used to work with signalling unit that used Reed relays. The reeds had gold plated tips they had a nasty habit of 'welding' together if the relay was energised for more that a few months.
The cure... put the card out of the rack to the full extent of the runners and slam it back in to place... contact broken... clear code: reseated card
Most people don't understand cold welding or the physics behind it (particularly in crimps) and decide to "do a little extra, just to make sure" - the result is almost always the crimp going unreliable due to the compressive force having been decreased by 90-95% over doing it properly and ion transfer (cold welding) failing to occur
A properly done crimp will easily outlast soldered parts
I knew a retired master machinist whose motto was, "Never use force. Just get a bigger hammer." On one job he worked on, they were putting together a press-fit pipe...6 feet in diameter. The "bigger hammers" were two guys with 100 lb. sledges driving the fitting home.
Yup.
I have a 48 ounce (1.3 Kg) dead blow hammer that's also soft faced. (It's fire engine red, can't miss it) I have the words "drive adjuster" scrawled on the handle in marker. It's been used to seat reluctant rack rails on their drive shelves during installation.
We also have in the team's shared toolkit a Stanley FUBAR Forcible Entry Tool that the boss bought because 'we needed a crowbar to unpack some of the gear that gets shipped to us'. I can't remember if I used the label maker to put "Maxim 37" (There is no overkill; there is only 'open fire' and 'reload') on it or not.
"We also have in the team's shared toolkit a Stanley FUBAR Forcible Entry Tool"
Wow. I wondered what that linked to, and it linked to exactly what you described. Now, AIUI, FUBAR= fucked up beyond all recognition, so top marks to Stanley for their product naming (a bit like the now sadly missed Screwfix "Sticks like Sh*t" brand). Not sure how Stanley got FUBAR as a registered trademark when it's been part of the language for seventy odd years, but I suppose we shouldn't complain.
A die grinder and laser tachometer, for hard-drive data recovery. Not, mind you, to cut open the drive or whatever, but for the actual, in-the-moment data capture. Oh, plus the oven in their kitchen.
I've told the story before.
The main piece of the puzzle not mentioned there is this (highly appropriate) MFM hard drive reader/emulator, about which I can't say enough good things. It does a very niche job -- but for that job, it's brilliant.
Ex apple genius here....
The first aluminium iMacs, circa 2010, I think, had 3 very large capacity fans inside (power, Logic board, HDD), and if you're going to blow air out, you gotta suck it in.... So, have seen my share of dusty (and other flotsam and jetsom) Macs.
However, those Macs had a pice of polycarbonate infront on the screen, but with a gap that wasn't sealed.... this meant that air circulated here.... Match this with people that had ash trays infront of their macs, and you had customers coming in wondering why there was a yellow tinge to their display.
Cleaning it only made things worse: we'd have to charge them for a new polycarbonate
My ashtray sits to the left of my laptop on the table to my right. When the window on my right is open, fresh air comes into the room through the window and exits on my left through the screen door. The two windows behind me in my 'office' don't normally have any effect on the smoldering cigarettes.
This is only a problem in winter when the glass door and all of three of the windows are closed. Sometimes it gets hard to see the opposite wall...
Years ago, (mid 90's) I was the service manager for a computer store. We serviced both PCs and Macs.
One day we had a very good customer of ours bring in his Mac. He was having problems with the electronic eject on his 3.5" floppy drive. I was never a big fan of the whole electronic eject concept, not to mention the whole "drag the floppy to the trash can icon" to eject the disk being the stupidest UI decision I've ever seen.
So, I took his Mac into the back room. I popped a key-cap off of an old, dead keyboard, straightened out a piece of a heavy duty paperclip, mixed up some five-minute epoxy, and glued the paperclip wire into the key-cap. I then crossed out the letter on the key-cap with a pen, and wrote "Eject" below it. I stuck this whole thing into the tiny manual eject hole on the 3.5" drive, and took it back out to the customer.
This was a customer that I knew very well, and I knew he had a good sense of humor, otherwise I would never have pulled a stunt like this. This eject problem had been an intermittent issue for a while, and was annoying the heck out of him. When I handed him back his Mac, he looked at it, and started laughing. At this point I told him that I was just messing around, and that I would order him a new drive unit, and I would call him when it came in. He said: "No, this will work just fine. how much do I owe you for the repair" I told him "No charge."
He brought the Mac in almost a year later for another unrelated problem, and it still had my "Eject" button on it.
"I left a delighted station owner who was oblivious to how close he came to a big bill and delayed opening,"
Never leave ANYONE oblivious to how close they came to a big bill, or you'll be back again next week. Some people only learn when faced with a catastrophic walletectomy.
Of course, making sure they know can lead to the opposite of a delayed opening, so stand clear.
I logged a PC fault with a supplier, the name of whom may or may not rhyme with Hell, and they sent out a new motherboard and a tech to replace said motherboard. I knew most of their field techs for my area and had a good relationship with them to the extent of switching the kettle on as soon as reception called me to say they'd arrived, so I was slightly surprised when a new face arrived in my cupboard. He introduced himself and proudly said he was getting into IT after leaving the RAF (I think he'd been a ground controller) after 22 years and had had the full training for his new job the week before. He carefully spread out his anti-static mat and smoothed it down, whilst advising me what he was doing and that it was to protect the fragile components from "random electro thingies". Okaaay, fairly good so far thought I. He put on his wristband and made sure all was connected properly. Then he placed the PC squarely onto the mat. On its rubber feet. That are not conductive.
"He put on his wristband and made sure all was connected properly. Then he placed the PC squarely onto the mat. On its rubber feet. That are not conductive."
To be fair, not only was he doing as he was told/taught, but by picking up the metal case, he'd discharged it via himself and the wrist strap and the mat is there to make sure any parts removed from the PC or new parts from packaging are at the same potential :-)
But yeah, I do see the funny side. Clearly the training (and I have done some Dell Servicer (and other OEM) training myself) only explained WHAT to do, not WHY to do.
Once had to remove a server from its rack that had been installed at an angle - the right hand side was one hole lower than the left.
I don't know how they managed to get the bolts in, but I ruined my favourite ratchet screwdriver getting them out. If I'd had an angle grinder, I'd have used it on the previous engineer.
Same, but with a drive shelf. Thankfully, we caught it early because the shelf wouldn't full seat right and it because bloody obvious what was wrong. :)
Other good rules to note:
Never let the apprentice use a power drill (even with a clutch cranked down almost all the way) to help rack stuff, unless you want to strip the absolute hell out of the screw heads.
Same, but especially on two post racks where the holes are pre-threaded, but all you have are the 'self-tapping' screws.
For that matter, just don't use a power drill on those unless you know what you are doing and can start the screws right- Cross threading rack holes is a Cardinal Sin to the God of Electrons.
On the same vein: NEVER EVER EVER EVER!! use self-tapping screws on a 2 post telco rack unless they came with the rack. And maybe not even then. (And the installers who abscond with the screw packs that come with the racks deserve a special place in the Bad Place.)
Follow those rules, and you won't get called every phrase in the Book of the Profane by the poor schlub that has to work on the thing afterwards. Maybe.
Ta for the reminder to put the robertson driver back in my toolbag.... I'm a british sparky and I do it to STOP plasterers removing boxes!
And yes, I keep a cheap battery drill with a 3.5mm tap in it for threading the holes on electrical boxes, it never hurts even (or especially!) on brand new boxes
"Ta for the reminder to put the robertson driver back in my toolbag.... I'm a british sparky and I do it to STOP plasterers removing boxes!"
Plasterers (and builders) shouldn't even be allowed within one metre of a back box. I'd rather the sparks did the plastering than the horrible mess builders make of back boxes. Crappy, single gang 25mm back boxes, fastened out of true with an effing clout nail into a lightweight block, or 16mm lightswitch boxes that are set so deep you need extra long screws to hold the plate on, mangled threads, bent tags etc etc and STILL the b@stards don't do a good job of plastering up to the box edges.
Back many decades ago, when I was in radio broadcasting (okay, the 1970s), the chief engineer at one of my gigs used to clean equipment by yanking it out of the rack, taking it down to the local "two-bit squirt" carwash (I said it was a while ago), and using the self-serve power washer to blast crud out of the equipment, then leave it in the sun to dry out.
As far as I know, he never had a failure, so I guess it worked.
I did that with a bunch of servers that ended up evenly coated in inside and out with about 1/4" of concrete powder after a renovation.
The servers kept failing and having to be power cycled.
Pulled the CMOS, and ran them under a hose...then let them dry for a few days with a fan.
It was a good bridge until replacement equipment could be picked, ordered and installed and they still worked when taken out of service.
As someone who spent half his apprenticeship having to wash down contaminated PCBs (mostly tills from the hospitality trade - ie: bars), I can understand why he did it and it would be quite effective at not only cleaning the boards but finding any dry joints.... :)
Many moons ago I ended up at a client. We'd supplied them with a PC to download stuff from their CAM to their CNC machines - early 90s so RS232 and CNC controllers with sod-all RAM. PC had decided to stop working. On opening it up, nearby grinding machines had generated a 5mm+ thick layer of metal dust all over the motherboard. Impressed that it kept working that long
I once had an HP laptop with a dodgy power connector. I managed to procure a
replacement from Alibaba for about 0.001c, but the original proved rather tricky to remove. My poxy 25 watt soldering iron couldn't generate enough heat before the metal can surrounding the connector radiated it away.
Judicious application of a 10,000 RPM 7" diamond cutting blade made short work of the metal can and I was able to remove remnants by "metal fatigue". The remaining pins were no bother for the mighty 25W iron.
Youngest daughter was much impressed and Dad's hero status reinforced.
I used to use a leaf blower to clean out my PCs all the time! Once or twice a year, I took them outside and gave them a good flush. While doing this, I placed them in the stream of a carpet-dryer blower to create an artificial breeze to carry away the toxic cloud. If I wasn't careful, I spent a day or two suffering from pollen allergies owing to the resevoir of crud within the case.
Another thing I did that I thought clever was to open a window about 6 inches, and set up my 12-core Dell T7500 workstation so that it backed up through this hole horizontally (padded on the sides with foam rubber blocks) which blew all of the hot air straight outside in the summer. In the winter, I flipped it around and it sucked in cold air directly from outside. I was able to do some serious SETI@home without concern for over heating. The poor beast had powerful fans, but undersized heat sinks inside and would ordinarilly never be able to run high loads continuously. Also, Dell had crippled the fan controller so that they would never ramp up in speed to respond to heat spikes. I used SpeedFan to manually turn them up as high as they would go. I would still do something like this in the summer, but it turned out that in winter, there was condensation and this led to rust on the case.
I use my leaf blower to clean my home entertainment cabinet. It does a great job without having to move any of the kit or the cabinet. Those pesky spiders don't stand a chance! I've got an open plan living area with wooden floors and large sliding doors both sides. OK, I use it to clean the entire living area when no one else is at home. I also use it to clean out my coffee grinder, outside of course.
> was "amazed that all the components were still on the board" – such was the force of the leaf blower's output.
Leaves are large and light. (If they are wet like today, no blower will move them.) Blower blows about as fast as a car. I routinely used 100psi/7bar compressed air through a needle nozzle. Exit velocity much closer to a jet plane, the speed of sound. Maybe it would dislodge the sink on a low-price Dell but never any real harm.
Of course leaf-blower lofting deep deposit IN the computer room is the setup for the next failure....
> It's a miracle the server survived, dust + leafblower = static electricity!
I'm with Peter2- in most civilized climates, forced air is not a static danger. And yes I handled CMOS, on boards and bare naked. And had long-term care of the chips I blew. After that first generation they were severely protected against finger-shock.
> walking across a carpeted floor and just touching it I gave it a good zap.
In my youth (before CMOS) we thought that was funny. The first generation of CMOS were not prepared for Real World.
I ever saw used to repair a piece of equipment was... me. I was sent out to repair a SCADA master site system at my local POP site with zero training, zero support, and zero manuals. I guess I was lucky they told me it was a SCADA system as I was at least able to find the equipment. After looking it over, wiggling a few wires and tapping some cards or something, I shut the power off and turned it back on which seemed to fix it. I was only ever sent the one time, never was trained after, and after over 20 years I don't really even remember what it looked like. Worked through them remotely for years so I know exactly what they're for, I've even worked remotely with a few field techs who were sent out to repair the SCADA connection for their site, but that was the only time I ever confronted on face to face.
For those who don't know, a SCADA system is a remote telemetry polling system for alarm reporting. The master unit will poll each site, one after the other, and when polled the site would kick back useful information like open doors, HVAC faults and other housekeeping alarms for sites. Anything that can be read via a normally open or normally closed switch can be sent over a SCADA system, and it can read, I dunno, 30 remote sites or thereabouts and something like 12 alarms points per site.
" After looking it over, wiggling a few wires and tapping some cards or something, I shut the power off and turned it back on which seemed to fix it."
You have to be careful of that. When I worked on audio electronics I was the only one in the shop that every repaired a particular amplifier that Fender put in a powered mixer they sold. From then on I would have to have a look at every one that was brought in. Not that I nor anybody else was successful repairing one again and we were THE Fender repair center if you sent your item to Fender for repair. This means we also received all for factory support we could ask for to no avail. I got some atta-boys, but I also got lumbered with having to work on them to no end. I have to wonder if anybody else that did spot how to make a broken one go again didn't make the repair so somebody else would be assigned the next one.
If you work on repairs for very long, you accumulate a list of those cases where bang, tap and wiggle does the job and you never know what the real problem was in the first place. My advice is to push for the accolades of your repair-foo and never own up to the real story.
I used an axe a couple of hours ago to fix a data security risk on some no longer required hard drives. The 2.5 inch did the decent thing and made tinkly sounds, I wasn't so sure about the 3.5 inch, so in addition to some massive whacks to knock the spindle out of true, I ran a drill through it a couple of times. I can't see Kroll being interested, but if they are I don't fancy their chances.
"I used an axe a couple of hours ago to fix a data security risk on some no longer required hard drives. "
I wasn't the one wielding the axe but a stage lighting company I worked for disconnected power that way once. The facilities person that had the key to the electrical room buggered off for the night before the end of the show with our power feed cables connected inside and led out through a hole at the bottom of the door for the purpose. 4/0 wire is expensive stuff and we also had to leave behind our lugs, but there was no choice as we had to leave from that venue for a show down the road the next night. I supposed we could have attacked the door with the axe, but there may have been a damage clause signed somewhere that wouldn't apply to us chopping our own cables. We did apply a thick layer of gaffer's tape and put a sign on that the wires were live.
I was driving from a customer's site back to base to collect some parts and pulled into a lay-by to answer a call of nature. When I got back into the car it wouldn't start. An investigation showed the distributor contacts had closed up. Unfortunately I'd left my tools on site so I had a scout round the lay-by and found a half-brick and a length of steel rod whch I used as a hammer and drift to open the points.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
