Static electricity can be shockingly funny, but the joke's over when a rack goes dark Returning to work on Monday often imparts a rude shock, which is why The Register opens the week with a new installment of Who, Me? It's the reader-contributed column in which you admit to your worst moments at work and explain how you survived them. This week, meet a reader we'll Regomize as "Connor," who told us about a job …
I remember working at an Italian observatory in Switzerland on my BSc thesis project, and the dry air at 3200m altitude caused a lot of static electricity. I was given strict orders to earth myself by grabbing a metal handle on the desk, each time I wanted to type something. Now I had to issue commands to the scope control software every few minutes or so, due to the tracking system being a bit wonky on the 1.5 m aperture infrared scope. I had to keep a guide star centred in a little box I would have to draw on the CRT screen with a whiteboard marker pen before each observation, by issuing short commands to steer the scope up, down, left or right. Even sitting still for a minute or two, you would build up enough static electricity to get a nasty shock. I never blew up a keyboard (which previous astronomers had done) or worse, but it was far from pleasant.
I may have told this one before, but here goes anyway:
Many many moons ago, our physics class was given a somewhat free access to the various items of demonstration hardware and told to pick one, figure out the physics and demonstrate to class.
One classmate picked the static charge generator.
When time came for his demonstration he was in front of the class and proudly turning the handle with one hand, holding on to the metal sphere with the other, and his hair was standing out nicely.
Then suddenly an "eep" was emitted, his eyes rolled up, and he collapsed behind the desk.
When he came to after a while he explained as follows.
1: The desk on which the generator had been operated was a conductive steel desk.
2: He had been wearing boxers that day, and his "fuse" had somehow found it's way out beyond the protective cotton.
3: He was a young man comitted to style, so instead of using a zipper, his jeans were buttoned with a row of nice and conductive brass buttons, agains which his "fuse" had been resting.
4: In order to achieve maximum hair-effect, he had been turning the handle of the apparatus with vigor, resulting in a full body swinging motion that ended with one of the aforementioned brass buttons touching the steel desk, completing the circuit.
Apparently the fuse did not blow completely, but it was a physics demonstration that no one in class was eager to duplicate.
Well, that is basically what he did, using his "third leg".
Joke aside: The goal of the exersize was to electrostatically charge him, and use his hair as a visible charge indicator. For that purpose, his rubber sneakers served as insulators (as would rubber in other areas also have, but there is such a thing as being too prepared).
Adding a grounded strap would have negated the whole experiment.
I believe that that one was originally invented in 1837 by a Welch farmer, who was using the lower intestine of a sheep.
In 1842 the invention was then improved upon by an Englishman (Mr. James Tuddly Condom), who figured out that it was possible to dispose of the rest of the sheep.
Our UPS's have been pretty good (even when the mains voltage got up to 255V one day), it's automatic transfer switches that have caused problems for me. One in particular (in a comms rack, just to cause extra problems), that seems to freak out at minor supply changes and rather than switch to a feed from a UPS, just shits the bed.
Had a remote site call that the power had gone out but no one could access anything even though the power was back on. Caller finally mentioned that the black box on the floor wasn't beeping anymore. Took me a while to realize he was talking about the UPS for the server that had been beeping for months. Yup. Dead batteries.
I've had "we've got a power cut and there was a box beeping so we turned it off. Now nothing works."
Thankfully that UPS supported cold start and I could talk them through turning it back on.
I did explain that the beeping was signifying loss of power, and that the correct resolution would be to restore power to it.
I generally defend users. I've never been a proper tech pro and all my work supporting users has been firmly from their perspective, But I do draw the line at something is obviously not working as it should but we'll ignore/silence it
FFS I don't care if it's a bleeping box or a leaking roof you damn well tell someone (usually me) and get it looked at, you don't just ignore/silence it.
We have offices in America, where the UPS alarms go off every time it rains or it is very windy...
The alarms on the European servers have only gone off in one location, once, in the last 6 years, apart from the times we've replaced the batteries.
We had an AS/400 at one place I worked. The APC UPS said the batteries were 100% on the weekly self-test. Then we had a power failure and the batteries held for a whopping 2 seconds, before the power dropped completely. All the batteries were dead, but the UPS said they were all healthy and 100% charged! Needless to say, after that the tech in charge of the UPS didn't rely on the self-test and did regular load tests, after putting in a redundant UPS. When the AS/400 came back up, it had lost one of its DASD's, the drive had been running for so long the bearings had dried up and as soon as they cooled down, they seized solid. That was an expensive lesson.
Our little IBM PS/2 Novell Netware server in the corner carried on merrily during this, its UPS didn't have any problems.
We had what was reported as a failed interface cable on an AS/400. I left my evening shift operator to deal with the engineer, it turned out it wasn't the cable, it was the DASD itself, and neither of them had thought it would make sense to take that drive out of the array (it was RAID0) before starting.
Had I been there I've got a few tricks that have worked in the past, including taking the top off the drive and spinning it by hand to get it moving - it worked long enough to recover the data. As it was we spent a few days with restores and rebuilds.
I had a similar case: UPS says "15 Minutes". But we were lucky to do a real test with one of them (two in failover config): Started at 72 seconds, and lost 10 seconds runtime every two seconds, and then was off after a bit over 10 seconds. Investigation result was: The UPS and the batteries are fine, but for that high amount of power the batteries could not deliver. After extending the batteries with external packs the UPS showed 52 minutes, and the test confirmed it indeed works that way - enough if one of the two UPS failes and the other left would have to do it all, which resulted in 20 minutes during the next test.
You've done it right: testing for real. All the metering in the world is but an assumption until you calibrate it with real life testing.
Bonus test: restarts. If the situation got so bad that it demanded a full preventive shutdown, testing a restart is also interesting. Unless you have sequenced startup (manual or automatic), the very startup pulse may throw breakers or trip overload limiters, and that's just the electrics. Then you get to discover which services do not auto start etc etc. The first time it can offer hours of entertainment, which is why you want to have the first time when you planned for it..
Yeah wrong wording... handymen were at the door to discuss planning. More important than TheRegister :D.
I was planning to write "We don't implement automatic startup unless the customer explicitly asks to, only manually". And if automatic it includes a lot of procedures information for all possible cases we can think of, and warn that unexpected thing might happen as well.. (like an F16 smacking down the wall - should the servers automatically start in such a case? What about the water from the floor above seeping through?)
However the <> failed at that task and instead shut down ... which seems to apply in over half the cases where a <> is required to actually do its job.
Unfortunately one can pretty much substitute anything for <> at the moment.
This post has been deleted by its author
A long time ago, we used to have crazy office chairs. No matter what type of cloth you wore, any time you unseated yourself, you were loaded with a massive electrical charge.
I swear, it was so painful I got used to grab a key before quitting my chair and approach it to a grounded furniture in order to discharge.
I got some 2 mm flash every time !
My father worked for a company that did a lot of the overhead electrical pylons during the 50s and 60s in the UK. When he was an apprentice, one of his first "field" assignments was a downed power cable in a field after it had rained. All the others got out of the van and put on wellies, my father didn't have any and didn't know better... Until he took a step into the field and got blown off his feet! The power was still running through the downed cable!
if this is a true story, your father was lucky. If he was not made to wear wellies there was a failure by the others. There would have been a 'voltage gradient' in the ground, radiating out from the point where the cable touches the ground surface. This can be hundreds, or thousands, of volts per metre, and since the length of a pace can be a metre or a large fraction of one, that's a voltage difference berween one foor and the other, enough to kill humans or animals such as cattle. Power workers in America and maybe elsewhere are taught to either stand with their feet together or 'bunny hop' away from downed lines.
Surprised as I know for example when they paint the pylons here in the uk, they phone the control centre and let them know where they are working and they deenergize that segment, though apparently there is still a noticeable tingle when you have 2 250000 volt segments running parallel due to inductive currents
As a junior dev, when business attire was mandatory I used those more comfortable rubber sole dress shoes. It charged me up quite a bit, and I would frequently distribute my energy around the office. A particular favourite was my IT director, which I liked very much to shake hands with. Sometimes even my colleagues saw the sparks!
"I swear, it was so painful I got used to grab a key before quitting my chair and approach it to a grounded furniture in order to discharge."
I had a car like that. I'd get a shock not from the car door handle, but from the house door handle as I discharged into it. So like you, I got into the habit of keeping my keys in my hand and using them to discharge into the house door handle before opening it.
> A friend had a keychain with some capacitor or whatever it was in it...... a one-time thing, we never saw them sold again.
Resistor. 1Meg is good. Any wattage. Wants to be "Thru hole", not SMD. Friend in the corporate archives (excessively dry for storage of historic papers) needed one.
"we never saw them sold again."
Check your favorite online chinesium tat emporium for "pocket static discharge". Under half a buck each.
The ones that light up can make for fun party favo(u)rs for a kid's birthday. Can be used as a teaching moment. I got a couple as freebees from either Granger or McMaster Carr at a trade-show about forty years ago (the keychains with the logo are long-gone).
Caveat emptor.
Aeons ago I worked for an outfit that made, amongst other things, high speed drum printers - 600 pages per minute - onto fanfold paper.
We'd do a test run then lift the stack of printout off the top of the machine, flick the laser line of perforations to start a tear, rip off the stack and take it to Reg, the QA guy, to OK or not.
There was a strip of grounded tinsel at the paper delivery slot, which was supposed to strip any static charge off the paper.
Only on this machine it hadn't been connected.
Mugging here lifted the paper stack, went to give the perforations a flick and received one GODALMIGHTY belt, a bit like being hit in the crook of my arm with a baseball bat.
It's moments like that which confirm the colour of adrenaline is definitely brown.
We had a line of VAX 11/780s in the mid 80s. A DEC engineer came out to do a memory upgrade on one of them. All the jobs were shifted from the affected machine and it was shutdown (console said it was shutdown and the power could be turned off). The DEC engineer disappeared behind the boxes and threw the power breaker on the wall... And nothing happened, he reappeared and the ops sitting at the console looked at him questioningly. Then the screams started, from the next VAX in the line, the one where all the users and jobs had been pushed from the machine getting the upgrade, he had thrown the wrong breaker!
A much smaller DEC, in a filing-cabinet-sized rack, was installed in an attorney's (barrister's) office suite. This was a new business series intended for such uses. The fans were quite effective at keeping it running, but did not fit in with the character of the surroundings. Normal abatement efforts were unsuccessful, so the senior attorney contacted an executive at DEC. The executive, oddly enough, went down to the shop floor, listened to an example of the unit in question, and replied "That's unacceptable. I'll have a tech there to replace them with quieter fans very soon."
In the fullness of time, a tech with a couple of brown boxes appears and installation of said new fans begins. Unfortunately, a Murphy demon also arrived and a wire slipped from its usual place to brush the CPU board. The other end of the wire, as you have guessed, carried voltage and currents suitable for destroying business computers. Imagine upsetting sound effects and smells. The Murphy demon's friends in the DEC warehouse then hid all the matching CPU boards so what had been a nice bit of slightly-noisy kit remained expensively silent until a distant assembly line came round to building that board again.
The secretaries transcribing dictation tapes appreciated the quiet, at least.
I have a couple of server rooms that have ethernet thermometers installed, that are set to email all the important people in the business if the temperature exceeds 40 degrees C, the idea being at least one of them can investigate and solve the issue, before the server room melts down.
On 3 occasions now, someone has been working in the server room, myself included, and brushed against the metallic surface of the thermometer, discharging static electricity in to it, resulting in over 20 people receiving an urgent email to tell them the temperature of the server room has reached 3000 degrees.
Mushroom
I have a couple of server rooms that have ethernet thermometers installed, that are set to email all the important people in the business if the temperature exceeds 40 degrees C
would you not want that alert at ~30c or more likely 25c?
operating temps typically only go to 45c for network and 35c for servers.
servers shutting down is often enough to cool a room so that 40c might never be seen whilst your servers have expired.
The 40 degrees email goes out when all other alerts have been exhausted, it literally emails all directors and managers and not just the IT dept because something must have gone drastically wrong to reach this point. We're thinking possible fire. The room is set at 19C. We also have a caretaker function that starts to slow the CPU and GPUs down if the temperature keeps rising.
The 40 degrees email goes out when all other alerts have been exhausted, it literally emails all directors and managers and not just the IT dept because something must have gone drastically wrong to reach this point. We're thinking possible fire. The room is set at 19C. We also have a caretaker function that starts to slow the CPU and GPUs down if the temperature keeps rising.
you have a plan then
have they ever triggered apart form the static shocks?
last times I was in a situation with a hot server room was during power events & the aircon failed, once during a power failure where the ups and genies kicked in but genies didn't power the aircon, ironically was in winter with snow on the ground and we had a keen heldesk guy walk in which took him ~ an hour and the power came up 20 minutes before he arrived.
another previous time was when they where ultra sound testing the mains and we powered everything off, when given the all clear the aircon (can't remember the brand but the rack size units that blew cold air through the raised floor) failed to start but we where assured it would work soon and where instructed to power everything up. took 4 days for that to be fixed and we had portable units blowing cold air in the room, server temps where around 30c.
Network kit is typically the last to die in hot rooms and can be long after other kit has expired. All Kit is usually a bit warmer than ambient temperature.
I think some of our network kit would be faulted at ambient 40c
"have they ever triggered apart form the static shocks?"
Not outside of testing no. We got close in the 4-6 weeks of summer, but that was taken care of when one of us had the idea of a process that polled the temperatures once a minute, and if the temperature started to creep up past 25 we'd slow down the processors to compensate. Without that the temperatures would skyrocket in summer. A heat feedback process was taking place, hot air coming out the back, would circulate round, go through the machine again and come out hotter. We had dual air-con systems, the idea being one was a backup for the other, but during summer, even having them both turned on we would still get the heat runaway, so for 4-6 weeks each year we just processed slower, and lived with it.
The funny thing is the thermometers were rated for -25 to 145 degrees, but they always reported 3000 with the electric shocks. Seemed a weirdly arbitrary value for it's dev to put a hard limit on. If it had been 32767 degrees etc I'd have accepted it with a smirk.
Many years ago, back in the 8086 days, I managed to fry a multi-io board in a machine simply by releasing the static charge by touching the mouse. The hard drive led went solid red and the machine became unresponsive. New multi-io board required.
Even more years ago, back in the physics class, it was our classroom show-offs turn to have sticky-up hair with his hand on the Van De Graff generator whilst stood on the insulating box. He thought it might be a laugh if he reached out with his finger to touch the teacher but as he leaned forward, his wedding tackle got too close to the gas tap on the bench....
Worked for Sequent in the US back in the day and had a customer with a vertical stack of 6 NUMA-Q compute units.
These were located close to a large UPS and the NUMA-Q system would randomly crash.
We did all we could: monitored the incoming power, replaced boards, memory etc.
Not sure what triggered the eureka moment but after exhausting everything else we focused on what might cause external EMF/EMI interference and with the UPS being so close, it was an obvious candidate.
Now the UPS was a big bugger and its chassis was thick as a brick so no obvious EMF/EMI but right at the bottom was a 2 inch high perforated vent that went all around it with numerous vertical cutouts.
So we setup a monitor and low and behold, at certain times (for reasons not explained but maybe related to the batteries) a large amount of EMF/EMI would blast out.
While the radiation pattern was wide it wasn't very high but it was in the perfect place to corrupt the active memory in the lowest NUMA-Q compute unit and cause the crash.
Solution was simple....move the rack a few feet away and never another issue.
Bluck
"Has static electricity ruined your day?"
No, but I worked in a place where it had the potential (sorry!) to do more than that. A lab where one branch handled explosives and detonators and another handled live ammunition. I think somebody in that area did have a det go off due to static (I was about as far from them as I could be and still be on-site so didn't know everything that happened there.).
The solution was an earthed brass plate outside the door to any sensitive area. Anyone going in was required to touch it first. I recommend that to the design of any installation where static could ruin your day (or more).
I'm a bit puzzled by all this.
It's at least 3 decades ago that I worked with statically sensitive devices and there's a whole raft of things you can do to prevent it, wrist straps with 1Mohm resistors, special mats or even tabel surfaces and even shoes and a conductive floor so I cannot see this being a problem in a modern environment, especially since quite a lot of this is not dirt cheap compared to when I was working in that field. Heck, I even set up a work area for myself in case I need to take a PC apart which set me back for less than 100 quid which included a solder resistive mat, wrist strap and grounding measures (which I duly measured for certainty - you either do this well or you're wasting your time).
Outside, yes, which is why you can get a grounding strap for cars (I recall some people being very sensitive to this), and indeed some gadgetry which usually simply contains resistors in the Mohm range to get rid of the charge in a slightly less shocking way but none of this is even remotely new.
BTW, the high resistance (+ 1MOhm) in these gadgets is also to protect you if the end bit somehow gets connected to mains.
This post has been deleted by its author
Many years ago I was told about the old BBC outside broadcast lorries, the one's you used to see at events like the grand national. These articulated live broadcast rigs created a huge static charge during operation. Needless to stay standard practice was to earth the van as soon as you arrived on site. However one day the BBC arrived late at Aintree, and in the hurry to set up, neglected the earthing. Needless to say when you are at the far side of the track (Beecher's brook) there are no "facilities", therefore the common way of releasing oneself of the suffice of tea that had been drunk was to duck behind the wagon, out of sight of the public, and widdle against the van. The people in the van heard the scream of the unfortunate first engineer who had hopped out of the van to relieve himself and had to pick him up off the ground where he was holding onto his shocked manhood.
In 1972 I was on a youth training course learning GlassFibre boatbuilding. We had an order for a 44ft hull so its mould needed to be brought from storage into the working area. The mould was supported by a wooden structure braced by scaffold poles and had two axles which steered making it easy to move around. It was hooked up to a car and the first attempt to move it failed. I could see that a heavy chain attached to the keel was trapped under something and rather than free it I disconnected the chain from the mould. Once free we got the mould out and into the workshop. A gantry ladder was brought over and I climbed up and lowered the middle ladder into the mould so I could climb into it.
I was given a tin of wax and a load of polishing cloths. Shoes would scratch the surface and socks were too slippery so I went in barefoot. Starting up front I waxed a two foot square area and polished it then moved along a bit. Wax and polish, wax and polish. About five hours later I was done. I called for the ladder and it was brought over. One guy held the gantry while another climbed up and lowered the middle ladder. I went to hand him the cloths and tin of wax but there was a huge spark between the tin and the ladder accompanied by a deafening bang.
I ended up on my back and the guy holding the gantry was writhing around on the ground in a fit. The one on the gantry didn't feel a thing. An ambulance was called and we went to hospital. I was fairly sweaty so all I had was burns to both heels plus track marks up my legs and back and down my right arm. My collegue was kept in overnight and was off work for a week.
That chain I removed was the earth, designed to prevent static buildup. I didn't know this and nobody else checked it either. 1970s health and safety - didn't exist.
First time I met this; a new peripatetic teacher was added to our team ( this was in the days when education had money to support kids in schools- so a long time ago). On days when she was in the base (and I was away in schools) she needed to use our computer. And kept telling me it was always crashing. I had no problem at all with it.
Until the day she came in while I was working there.As soon as she came close, crash!
I rebooted and we experimented.- No giggles in the back there! But her fluffy jumper was pretty sexy. And yes I could feel the electricity.
Much nastier. A few years later we were moved to a building with nylon carpets. The computers were OK- perhaps they were better protected by then. But even going close to a radiator, door knob, filing cabinet ( all the things you'd find in an office) and wham! A jolt you'd really know about.
We have random issues at the office with multiple monitor setups randomly losing a monitor (going full dark).
The fix can go anything from pressing the monitor's power button to having to unplug some other monitor's video cable, re-detect monitors in Display Properties, reenable the offending monitor, extend the desktop to it and re-plug the poor thing we unplugged.
This was quite a mystery, and happening completely at random.
A couple of months ago I got up from my office chair (not touching my desk in any way) and a spark flew out as I touched my chair's handle.
At the same time, one of my monitors went dead.
I'm appalled that it's 2025, and static electricity has not been abolished yet. This is unacceptable.
Not static problems here but RF
the monitors on my desk (3 x Asus IPS screens) all use touch sensors for the controls, unfortunately they must have an internal track that is resonant at 433-439 mhz, any transmission locally produces all sorts of menu flashing before the controls lock up completely (only thing left to do is pull the plug), even more unfortunately I run a 70cm amateur radio repeater with it's output frequency slap bang in the middle of the range that causes the issue.
There are many places I want real buttons back... Control for the monitors is one where they are coming back on customer feedback, for other tings as well. It is a big advantage to feel the button you can then press, so you don't have to take your eyes off more important things. (Tesla anyone?)
It seems like that's in advance of NCAP toughening up their testing standards in relation to touch screens.
More info here from the RAC but the TL;DR is that safety controls must have physical switches, and points will be deducted from vehicles if basic controls are not separate and intuitive
Not static electricity, but the real thing. During the 90s I was the UK based IT manager for a holiday company with resorts on the Costa del Sol and Canary Islands. I had flown out to a resort on the Costa and on arrival was informed there was a problem with the computers in the accounts office that had started about 24 hours earlier. We were running a 10base2 Novell network on the resort at the time, so both machines in the office were networked together and shared a printer which was attched to one machine.
I was informed that when either computer was switched on there was a horrible buzzing on the phones, and when the one computer tried printing to the printer attached to the other machine, the printer switched itself off (HP Deskjet). These phenomena were demonstrated to me so I could confirm it was true. Knowing the quality of the wiring in electrical installations in Spanish developments, I called the head of maintenance to come and check the supply, I suspected we had some sort of earth problem.
The head of maintenance duly turned up with a special plug that when inserted into a mains socket lit up different colured lights depending on what the problem was. He plugged it in, watched it light up and then took a very quick step backwards.
It transpired tat the local electricity distribution company had been doing some maintenance work in a substation attached to the resort and somehow had managed to cross connect 2 phases, so we had 415V coming from 220V sockets, not just in the offices, but also in 12 apartments in the same block occupied by paying guests. Strangely nothing that was plugged into any of the sockets blew. The electricity company came out to correct the situation the same day.
"how on Earth" - no, don't touch earth in such situations as well, could have 414V there are well. Like all three phases connected to Live-Neutral-Earth. Any real electrician will know that "Earth" is not without potential, and you can measure it - usually the above mentioned 220-235V can be sourced there, until the FI triggers... Except when the installation is old. House-wide FI (plural) is relatively new.
Reminds me of what happened to me here in the UK some decades ago. Working from home (luckily) when suddenly there was a problem with the electric supply. Not a power cut, very much the opposite thanks to water ingress out in the street. I'm guessing it was something like getting next door's phase of the mains hitting the neutral of our supply and so bypassing the fuses. Every light on every appliance was *very* bright and the whole house was humming! Cue a sprint to the fuse box to throw the main switch. If noone had been home then I very much doubt if the wiring, or indeed the house, would have survived.
(told before...)
Back in the 80s I worked in the same building as the local phone engineers. One day 'Big Bob', the jovial giant, the life and soul of the party, appeared almost in tears...
He had been given a job to reprogram some extensions at a big ad agency at their big swanky open plan offices. The PBX (a Herald?) stood in the middle of this vast office, surrounded by all this opulence. He had thrown open the doors of the cabinet and was working away when he sensed a presence behind him.
A voice said "look at all those flashing lights" and a spark jumped from a pointing finger and fried the CMOS circuitry! It instantly went from 'flashing lights' to darkness
They had spent well on the fancy furnishings and artwork on the walls, but that, and the PBX, all sat on nasty cheapo nylon carpet
The Ex-Mrs Oncoming Scorn, found herself frequently getting shocks when we emigrated to Alberta Canada due to the dryness of the air, such was her fear of the shocks she would always gingerly dab & point with her fingertips (Rather than just spread her hand & slam her hand flat) to touch metal doors to "Test" & shockingly would usually receive the positive test result she wasn't hoping for.
One day while in a furniture store, with lots of new furnishings, nylons, plastics, carpet displays she pointed out a over the top cooker extractor fan with the comment of
"Ohhhh! Thats just like the one we have in our currently rented (temp) home"
I wasn't actually looking, but I saw the bright flash as she pointed to the extractor in the stainless steel finish of a fridge freezer I was looking at & heard the scream at the same time!
So the TRS-80 Model I had a reset button in the back left.... next to the bare "expansion interface" card edge connector... that went directly to the Z-80 w/o any buffers because TRS-80s were done on the cheap like almost every other '80s computer.
You can see where this is going if you have a good static charge and need to hit reset. Suddenly the screen is garbage and power-cycling does nothing.
This is how I found out that warranty repair for TRS-80s came directly out of the Radio Shack store manager's budget.
There's an old story about the days of above ground nuclear testing in the Nevada desert.
Everybody was in place, Far Far Away, for a test. When the time came, nothing happened (but it happened suddenly!).
After much wailing, gnashing and drawing of lots, a couple of intrepid engineers drove to the test site, aware that they could be ionized vapour in an instant, to find out what went wrong. Telemetry was not a thing then, so nobody had a clue as to the condition of the Gadget.
Long story short, they found that, after thoroughly testing all the electronics and pronouncing them sound, the engineers had disconnected all their test equipment, which had in fact been putting a load on the system. Prompt removal of the load caused some kind of spike, and a breaker popped.
I lost an extremely expen$ive USB stick once to static. Came home, got out of my car and went to open the door. Felt a tingle. Went across the room, saw that my USB had finished copying so reached out to unplug it. Next tihing, bright blue spark at least 1.5" long jumps from my finger to the drive. Di-DUM. Yes, fried to a crisp. Left a burn mark on my finger for good measure!
Turns out that my car was missing a grounding strap so I'd been charged up to probably 100+ KV.
Strangely enough that's the first time a PC component has actually failed this way, have probably built dozens of machines and never zapped one. Lost a motherboard once but it had a chip torn off so likely that the extra stress of putting a K6-III 550 in there was too much for its already stressed regulators. Put the old chip back in, also not working.
Re checking scripts
The ideal script
- has 'set -euxo pipefail' (that should be default anyways, imho), so basic bugs cause hard fail, and it prints the commands (as a basic logging step, not to replace true logging).
- has an option to do a dry-run
- is transactional, so if things do catch fire, you don't end up with intermediate results, and it rolls back
On my todo list is figuring out how to call btrfs API to make a directory snapshot of any data that needs to be modified, but I lack the requisite expertise and snapper (for me) needs a bit of tweaking to get it to work on Fedora.
if anyone has resources on how to do transactional changes in zsh/bash, would love to see those
I have never once used one, either in working on my personal computers dating back to the late 80s performing surgery on a Mac Plus to upgrade it to a faster CPU or in my work life dating back to the 90s. Over the years I'm sure I've installed several hundred parts from CPUs to DIMMs to hard drives to add-in cards, and built 7 or 8 PCs for myself. Never had a problem.
I remember getting shocked a lot as a kid, but we had shag carpets in that house. Once we left the shag carpets behind in a new house static electricity shocks became something extremely rare. In the past decade the only time I remember having one is a previous car would often give me a shock when I touched the door handle in my garage to open it - and only when it was in my garage. Which is weird because I'm wearing rubber soled shoes, the car has rubber tires, I'm not walking on any carpet on the way to the car, and the garage is just ordinary bare concrete. It wasn't one of the newer models with the door handle that senses the presence of your key and automatically unlocks the car, either, so shouldn't be any electrical connection to the door handle at all. So not sure what was going on there, but I don't have that car anymore.
Just a warning that even though you didn't have issues without a wrist-strap in the past, ESD tolerances on modern electronic components have been reducing in recent years so prior immunity does not guarantee future success. You should wear a wrist strap and/or take other precautions nowadays.
As for getting static discharge when touching the car when your shoes and the car tyres are both insulated -- there can still be a static discharge when charge redistributes between you and the car to allow you and the car to be at the same voltage compared to the surrounding garage. This discharge will be short, but there's still a discharge.
I once attended a BBC Micro that refused to boot. The rig was set up in a bedroom next to a ham radio set, sat on top of the feeder.
I tried over and over to get the bbc booted but no joy. I took the lid off, wearing a static strap, and started pointing out the components to the owner. As I touched the top of the CPU, there was a big crack, a spark, and a boing. The surface of the CPU package must have been holding a charge probably picked up from the radio transmitter which I discharged with my grounded finger.
In a computer room with racks of minicomputers (and a couple of 900 line/min printers), the operator kept a spray bottle with Downy fabric softener to spritz the carpet and help keep the static dissipated.
I've worked in a place with well-grounded equipment in a room with epoxy-painted floors. My sneakers (trainers, for the East Pondians) would get me static-charged as I walked across the floor. I learned to carry my keys in my hand to act as a lightning rod as I approached the equipment.
I've noticed that my denim jeans will rub up a static charge as I slide off the car seat. Best to discharge away from the gasoline filler neck!
I am surprised that wrist straps were not mandated at the Alpine observatory cited by the first commenter.
They used to sell (and probably still do) conductive rubber grounding straps that you could attach to your car's (metal bumper, in those days) to discarge static that built up whilst driving.
Toll booth lanes, back when we handed cash to a collector, had a flexible metal rod that would rub the cars' underbodies to prevent shocking the attendant.
> conductive rubber grounding straps that you could attach to your car's (metal bumper, in those days) to discharge static
weren’t they shown to be ineffective ? I remember seeing them on cars and all of them were over time bent at an angle from the airflow and never touched the ground anyway!
"bent at an angle from the airflow and never touched the ground anyway!"
Presumably when the car stops they'll straighten and then carry the charge away to earth, in an instant.
But haven't needed them myself.
But in relation to cars, your clothing rubbing against the upholstery when you get out of the car might also impart static electricity on your body, and such a strap wouldn't help against that.
In the 1980s, the main radio studios in the BBC had a lot of gram decks made by a company called EMT. Wonderful machines, incredible quick start, push button reverse cue for setting the start point, time to run counter in mins & secs etc. all you could ever want in a deck designed for a major broadcaster.
However, all the transport controls were back lit tactile buttons which had a tendency to trigger erroneously when you carried any static. You only had to get within a cm or so for the switch to trigger the CMOS circuitry and all hell would break loose. Not just limiting the playback to normal 33 or 45 rpm you now had an infinite variation including reverse at top speed. So fast in fact that the cartridge and tonearm would be ejected off the disc with a loud splat to accompany the catastrophe.
To add to the fun, most of the carpets were hard wearing nylon. If you were working with a 'personality' that was unpopular (some were!) a surprising number of production staff would be seen rubbing their feet vigorously on the floor on their walk into the studio - I wonder why?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
