Techie diagnosed hardware fault by checking customer's coffee By the time Friday morning rolls around, starting the day with a stimulating beverage feels like a fine idea. And so does delivering a freshly brewed installment of On Call, The Register's reader-contributed column in which you share tales of tech support triumph and torture. This week, meet a reader we'll Regomize as "Ryan" …
I have had one weird issue with the power supply of a Leica fluorescence microscope in the 1990s. I was developing and supporting an image processing package that supported three different frame-grabber/image-processing boards, either the Matrox PIP1024(A/B variants), which were fairly basic, or the more powerful Matrox MVP/AT-NP boards, equipped with a neighborhood processor (NP), that could perform GPU-like stuff way back in 1992 (video-rate 3x3 convolutions and the like). Both PIP1024 variants run happily on the same software, but the MVP/AT-NP needed a different library linked to the executable. Three systems lived happily in the Department of Medical Microbiology, and another, with an MVP/AT-NP board was installed at the Department of Dermatology. This caused no end of trouble. Code that ran happily at the microbiology department caused crashes on the same hardware at dermatology. I got seriously suspicious when they people at dermatology mentioned that whenever the UV lamp's power supply was switched on, the computer crashed. They developed a protocol that they first switched on the power supply of the microscope, and then booted up the computer. Clearly, the power supply was causing spikes on the mains voltage when switched on. I then surmised that when my code ran on this fast processor, RFI from the power supply was at fault. Indeed, when the power supply was switched off, all my code ran sweetly. On a hunch, I linked the library for the MVT/AT (but not NP) board to the code for dermatology, and all was well. Bit of a bummer we could only use the expensive NP unit when the microscope was not being used.
No problems with coffee (or tea) fortunately.
soon as he said 12 awg could see the problem. think it supports 15 amps at 120 v safely. starting up two large hp color laser printers can cause a circuit breaker to pop on a 20 amp circuit. as it is my keurig can make led house lights flicker only 40 feet from the breaker on a 15 amp circuit with 12 awg wiring.
Re the 'tiny LED displays' - the display on the desktop calculator my team of 5 shared in 1975 at Avon County Council finance dept had green vacuum-fluorescent digits, and it looks like the ones in the linked picture are using Nixie tubes, fashionable now for clocks I believe, sort of retro-steampunk?
There is an area west of Los Angeles (west end of simi valley) that houses a lot of industrial machines (with large motors...inductive loads). They also rent to retail businesses. Apparently the local power company has never seen fit to install the proper capacitors on the power lines to balance the inductive loads and the retail shops will suffer with mystery problems on any computers until someone helpfully points out they will likely need an isolating UPS.
I have learned a long time ago the usefulness of an UPS to keep PCs from getting damaged by power surges or cuts.
Funny, because I live not far from a nuclear power station (about 15km, give or take a few). You'd think that the quality of the power line would be reliable so close to the generator, but you'd be mistaken.
All of my electronic equipment is, at minimum, connected to the mains with a surge protector (TV, stereo, fridge, freezer, etc). In my home office I have a 1600 VA UPS to which all my desktops, my NAS, the connection box and the telephone are connected.
If there is a power cut, I power down the computers and leave just the phone and the box connected to ensure that I can still use my laptop's WiFi if required until the power comes back on.
But I am never again going to trust the mains to give proper juice to my computer equipment. Yes, I know that, in the past 20 years, standards have evolved and computers are more resiliant.
I'm still not taking the risk.
The trick is not to be near the power station, but to be near important consumers. I used to work out of an office that was connected to the same substation as a major hospital, a small military base, and a national broadcaster. We had almost zero power cuts, and on the rare occasions that we did, it was usually back on before the UPS's had to start shedding load.
Absolutely. During the 1970s, when UK power was not at its most reliable, we were on the same substation as the newly built District General Hospital.
Not that we had much computing power in the house to worry about at the time, so UPS wasn't an issue (Lunar Lander on a Sinclair calculator ran off batteries).
In fact, just be near *any* other customers if you want any reliability. My last house was in the middle of nowhere (no mains gas, water or sewage **) and the supply was so bad I ended up with several UPS units throughout the house, including one for the TV & Sky box.
Usually the interruptions were for a few seconds, while the auto-reclosers did their thing. However, if the power didn't coem back in in 3 attempts, we knew that it would take hours before a crew turned up. At that point I had to get the generator started for the duration. The worst experience was during harvest, when I heard the local farmer's tractor crash into one of the poles carrying overhead cables - that took days to sort out, but when they did they also upgraded the cables which improved things considerably.
Thankfully I've since moved, but still have a load of old UPS units festering in the garage, just in case.
** Unbelievably, this particular middle of nowhere was within a 35 radius of central London.
>> ** Unbelievably, this particular middle of nowhere was within a 35 radius of central London.
I'd believe it. We had to install a couple of 1MW Caterpilar flywheel+generator UPSes to keep a site in the Surrey Hills running smotthly
They paid for themselves within 6 weeks of installation thanks to the total elimination of power glitches which allowed critical equipment cerrtification tests to complete in record time
At university in Southampton in the 1980s I worked for a stage light & sound company. Southampton Guildhall had a 3-phase feed on one side of the stage where the only things on that feed were the Guildhall stage and the power station at the other end. The feed was apparently installed during the war for some sensitive radio equipment installed in the Guildhall. A number of sound engineers I worked with said it was the cleanest mains feed they'd ever had.
When Family Outski left KL, we passed one of our baby UPSs to a neighbour/drinking pal. I bumped into him four years later (in the pub, on the trip to unify the family having hurdled the hostile environment, but I digress). He told me that his building, opposite the one we had share before, had recently been hit by a fairly big lightning strike (are there small ones?).
Everyone in the block's equipment failed with no power and/or damage. Except his. Telly on, router still working, laptop fine. Due to a tiny APC UPS that cost me about RM200 (40 quid).
Icon, cos he did do the decent thing and get the beers in
Used to work for a major DCS vendor who also made their own kit whose practice it was to fit UPS's in many of the nodes. We only found out later, after some of our UK customers' systems started bombing with power dips that, no, they were *not* UPS's but battery back-ups with a finite switchover time. On test, some of these took up to a second to switch =:o
IT Manager for a holiday resort in the mifddle of a golf course in Tenerife in the 1990s. There was no mains electricity feed to the resort, we relied on 2 very large generators on site which switched over at 4pm and 4am between the slightly smaller day generator and the larger night one. All our computers had to have UPSs mainly to filter the mains electricity but also to keep everything alive during the switchover which could take a couple of seconds to complete.
In a previous century when I worked for a company with Incredibly Borked Management we had an in-store sales system deployed in a popular high-street electrics retailer chain that is now defunct. Seemingly randomly the store server would go tits-up and would collapse in a e-pile of springs and cogs. No-one could work out what was going on... eventually a customer engineer was assigned the rewarding task of sitting there and watching things all day to see if something environmental was going on such as gross stupidity. It turned out it was the rear electric roll-up delivery door to the store, every time another fridge or whatnot was delivered and they opened their back door the motor sent spikes of death through the mains and the server stopped being one... fixed with some power conditioning.
I went to a site in India and had to visit some people in the backroom - and it was the back room with no windows. They had one power lead into the room, and other power lead was daisy chained off it. If people powered on their machines one at a time everything worked. If they all powered the machines on at once, it drew too much current and blew the fuse. Everything was fine till they had a power cut. When the power was restored all the machines started at once and so blew the fuse.
I also remember visiting India on holiday where the fuse board had a nail instead of a fuse, and the hotel room had live wires adjacent to the light switch.
About 30 years ago, I visited a supplier in Lahore, Pakistan. They stamped out and powder-coated the panels for consumer units.
The factory was in a 4 story unit, with one process per floor, stamping from sheet metal on the ground floor, acid wash on the first, powder coating on the second, finishing and inspection on the top floor, then the finished products slid down a chute to a packing station in the rear yard.
At the time, the area was experiencing rolling blackouts due to a power shortage, so they had 4-hours on, 4-hours off.
The cutting and powder-coating floors used a lot of power, the acid bath and finishing floors not so much, in fact the acid bath guys carried on working in the dark.
Anyway, nobody thought to shut the machines down properly during the power-off times, so every single time the power returned, the fuses blew, and they had to replace them.
Interestingly, while they made the panels for consumer units, they didn't use them, the power cables running to bare wire terminals in a bakelite box, the person replacing the fuse simply donned a pair of rubber gloves, plucked out the dead fuse, popped in a new one then ran away very quickly in case it blew up.
I recall a client (back in the 1990's) who had a small design office with half a dozen or so staff on drawing boards. It was in NE Scotland and it was heated with simple electric panel heaters on the walls, the type with built-in timers and simple thermostats. In winter, the office was often cold and staff invariably had to manually adjust the heaters near them to keep their space at a comfortable temperature. The problem was then that they would forget to readjust them at the end of the day and a weekend with the heaters on full was a wasted expense. The boss decided to have a master time switch installed on the common feed, so everything went off at the end of Friday's work and back on an hour before work started on Monday. What could possibly go wrong? Well - a freezing start to the week. When the timer switched power on that first Monday morning, the main breaker tripped. I did a quick calculation of power draw and the heaters, if all were on, required around 150A; the breaker was 90A. Of course it blew. With the heaters individually controlled, the chance of them all drawing full power at the same time was negligible but, all coming on from cold at the same time...
The time switch was quickly removed and the heaters returned to individual control; switched off overnight and at weekends, with a rota for someone to come in early on cold mornings to switch them on (usually the boss - as it was his money he was saving)!
One place I worked we had a main incoming fuse blow on one phase of the 3 phase supply - it was fairly obvious it had come slightly lose and simply overheated. It took 3 visits to get the thing replaced:
The first lot who were tasked with replacing the fuse couldn't get the old one out.
Three guys from the heavy mob arrived, having been sent from another job, to remove the fuse, which they did, but they weren't authorised to work on live equipment beyond that.
The senior engineer who'd been sent to check if there was a phase imbalance that had caused the blow was authorised for live work so he donned his high voltage rubber gloves and replaced the fuse, did his current checks and claimed that the phase which had blown was drawing a significantly higher load and we should get it checked out ASAP. One of our engineers pointed out that all the toilet hand-driers were on that phase, and had all come on when power was restored! If the chap from the Grid had waited ten minutes, as he should have, he'd have found the phases in balance.
The only IT angle is that fortunately the server (an HP 486 SCO Unix box) was on one of the other phases and stayed up throughout.
Reminds me of the RAF motto - "Per Ardua ad Astra" - through adversity to the stars. The RAF used to have on their bases a cinema chain, aptly called "Astra". This led to some wags (usually from the Army or the Navy) mistranslating their motto as "It's a hard walk to the cinema"
Similar, also India, about 2 hours south of Chennai. 15 years ago I think.
A local MSP had installed a nice server, firewall, the whole shebang. But there was only power for about 12 hours a day, and not even consecutively…
Each time power went out someone had to run out to manually start an ageing generator. The UPS installed was woefully underpowered to even perform an automatic shutdown. So everything crashed at least once a day. And was set to autostart, which caused fuse to blow if it happened. *Every* night that hardware fell over, *every* night the fuse went out.
Solution: added about 8x the UPS battery capacity (so at least a clean shutdown was possible), and got a local electrician in and showed them via pictures what his installation was supposed to look like.
Summarized: company sent a system engineer on a 5 day trip, of which 2.5 days on site, to show an electrician and an MSP how to actually do their job.
A couple of years later that site was moved, and the local manager told me that both the electrician and the MSP went around to all other local businesses with that site as a reference case. While we should have charged thém TBH.
The bad part: in that area (30m drive radius from the site) there was just 1 hotel that actually served beer.
The good part : they had a room available.
From my experience of that part of the world, a 30 minute drive, if that's what he means, is about a 45 minute walk, the roads outside of cities being very, very poor.
We were once on a train between Karachi and Lahore in Pakistan, whilst passing Hydrobad at little more than walking pace, the train was held up by bandits on horseback. They were held off by several passengers armed with rifles that probably first saw service during the early years of the Raj...
We had a QA/test lab full of workstations, each of which had an associated monitor. Being back in the days of 24" CRTs these had the usual degaussing coil which made a pronounced "thunnnggg" noise on power-on. No UPS on the lab.
We soon learned to leave all the monitors switched off when not actually viewing the screen, because if a power outage happened the total degaussing surge when they all came back on at once would trip the main breakers.
Power on or off... It was the late 90's and I had twin 21-inch Samsung CRT monitors at the office. I learned the hard way not to power them on or off at the same time. The few times I did that, I managed to trip the whole floor. And each of those times several angry looks followed my walk of shame from my post to the electrical board to switch it on again.
At one point in the mid-1930s my father was the staff electrician in NY residence hotel. One person there wouldn't pay his electric bill, so they'd pull the fuses. He replaced them with pennies. All my father would say about his solution was that the next time the guy pulled that he'd "get a handful of fire." After that...he paid his electric bill on time.
They are called "Fusestats" or type S fuses.
To remove the penny you had to pull the main disconnect directly above the fuse body.
Most of these boxes were rated at 100 amps so had two main disconnects, both rated @50 amps. One for the electric stove and the other for the rest of the house.
I still have machines from the 1950's in use that have Fusestats.
We had a monastery as one of our customers (Not really a customer, as we had donated our software and supplied the hardware at cost). The "IT" was in a small two-room annex with a single power line to one of the rooms. The line terminated at a single socket that had another line with a switch going to a single 100W bare bulb. The mains socket had an adapter going to the PC and its CRT, and a small LaserJet (5L?) printer. They could only run any two devices before the circuit breaker (in another building) tripped out. There was no heating or cooling (an austere order) - I suggested that they move the equipment to another building, but apparently the layperson who was driving it liked the "peace" and "solitude". They solved the problem by only using the equipment on clement days with natural light.
At UEA in the late 70s, if they had a power cut the power would come back on and immediately throw the breakers, so after a few tries a land rover would drive across the greenway outside Suffolk and Norfolk terraces with a bloke yelling into a megaphone "PLEASE TURN OFF YOUR ELECTRIC DEVICES!"
It was all most entertaining.
Me neither.
It's a septic thing a four core three phase (115VAC)+neutral cable manufactured by Southwire.
No idea what the equivalent 240 VAC cable would be called in AU or UK.
With a drop of 50V I would have thought E2/R might have made the Romex cable fairly hot if not the coffee.
When I was a kid some my contemporaries had a rock band whose "sparkie†" (the drummer I think) used to run 240VAC to the light show over 2 core (solid) Bell wire for cheapness and if a particular set of lights were illuminated for more than 15-30 seconds the insulation on the wire started smoking along its length. Quite spectacular but fortunately I never could tolerate loud music and always withdrew to be incinerated another day. :)
† totally unqualified - "Electricity: it's just like plumbing" - not that he was a plumber.
.
Or even simpler as 3&E or T&E
In the USA it is usually 2-phase split supply so 120V - 0V - 120V plus earth/ground. Big loads use the 240V from the outer set, smaller loads are 120V off either half.
In the UK it is rare for a domestic property to have 3-phase, and when it does it is usually either a very big place with a few single-phase boards on different floors/wings, or there are a couple of big loads like fast EV charger, or a big water heater or heat-pump. So the 3&E cable is almost always seen in small sizes (1mm^2 or 1.5mm^2) and used for two-way light switches, even though it is rated for 400V/230V 3-phase use
With few exceptions (simple heaters or motors, for example) 3P equipment also needs a neutral as it will have either an imbalanced load or a few single phase loads in addition to a big 3P one (for example controller logic on 230V and a 3P motor system using the 400V delta) so when you do see it wired it is normally 4 core plus an earth (could be a 5th conductor, or the steel armour of some cables).
"In the UK it is rare for a domestic property to have 3-phase"
Unfortunately there's one just down the road. Back in late 2020 we had one of our regular power failures. Two phases of the main cable (buried no more than a foot or so below the surface because nobody wanted to trench into solid rock) started arcing - the vibrations could be felt standing on the spot.
They also damaged the gas main in the same shallow trench. I discovered that because I could smell gas coming up through the drain about 100 metres further up.
The electricity team wouldn't start work unto the gas was made safe and the gas wouldn't start without temporary traffic lights (the spakies weren't fussed about that). The transport bringing the TTLs broke down and by the time replacements arrived it was starting to get dark, 6 hours after the initial 10.am fault.
The electricity people had a temporary generator but it was only single phase. They were going to strap all the phases together on our side of the fault but couldn't because of the house with the 3-phse supply; the owners were away so they couldn't get in to turn it off. It was finally fixed about 3 am.
Here in Germany, domestic 3 phase is common. Electric stoves are generally connected to all phases (but can run on one), larger water heaters (18 or 21 kW) almost universally are. My house gets 63 amps per phase, the most common main fuse (you have to be a licensed electrician to hange those).
It's handy when you want to install a high-powered wall box for an EV, for example.
My house gets 63 amps per phase, the most common main fuse (you have to be a licensed electrician to hange those).
IINM, that type of work needs to be signed-off by a master electrician if done by a standard (journeyman) electrician.
Yes, Germany still has a medieval style guild system for many careers.
In the UK it is rare for a domestic property to have 3-phase,
It's not rare in France, but it's still uncommon...
My parents, when they got our house built went 3-phases ( it was probably imposed at that time by EDF as the house was all electric (heating, hot water, ... )...
And while the industrial side ( heating, hot water, and a few other things ) had the phase balanced, the guy that planned the domestic side didn't really thought things out... which lead to tripping the main disjunctor on a regular basis in the begining until we sorted out that we couldn't use the dishwasher, the dryer and put the water heater on forced heating at the same time ( and a few other similar combinations also including the washing machine, and the various radiators ) as it was unbalancing the phases and tripping the main disjunctor. after a year or so all the tripping combination were known and avoided.
Mine still is.
It’s a hang over from the 70s when everything was electric because the French nuclear program was going to make electricity almost free.
See how that turned out.
Though because of the phasing and the probably insufficient power delivery there are combinations to be avoided
Always have a torch in a well known place with working batteries.
"There was a channel across the ceiling in the polystyrene insulation (no, really), at the bottom of which was a 1.5mm2 Cale to the cooker. "
Obscure fact for your next pub quiz: you're not supposed to put PVC-sheathed cables directly onto polystyrene foam insulation. If you do that the plasticiser from the plasticised PVC cable will migrate to the polystyrene. As a result the cable sheath turns into unplasticised PVC which is brittle and will crack when the cable is disturbed.
Time of the week -->
the plasticiser from the plasticised PVC cable will migrate to the polystyrene
Which also "melts" the polystyrene. So the OP's might not have melted through being too hot but by a similar chemical reaction to the one you get if you dribble superglue onto expanded polystyrene.
Some cable clips used with rubber cables were polystyrene (the not-expanded type, obvs!) and when used with newer PVC cables had the exact same problem. I've come across PVC cables seemingly hot-melt glued to nails. In actual fact, polystyrene cable clips melted into a glob with the cable. Of course, rubber cables leached their plasticiser without any third-party help, which is why they had service lives generally no more than 20 years whereas properly-installed PVC cables are often good for 50 years or more.
M.
Romex (aka NM-B), in the US, is what's usually used inside the walls of houses. It comes in different wire gauges, with and without an uninsulated ground (earth) conductor. Most commonly seen as 14-2 and 12-2 (first number is the AWG, second is the number of current-carrying conductors, excluding the ground). Come to think of it, I have never seen it without the ground.
When we wired in my son's Tesla charger, we wired it on a 60A dual breaker (240V) and used 4-3 "tray cable" to support the 48A charge current. Tray cable comes without an uninsulated ground, and though the appropriate gauge wire for 48A is 6 AWG, we went thicker because it was a 60 foot run from the panel to the garage.
> No idea what the equivalent 240 VAC cable would be called in AU or UK.
Three phase, but 230 V. Four cables. Three three phases and neutral/earth. (You can have earth and neutral separated and run five cales, but they come together at the equipotential bonding rail somewhere in your house or flat).
This is all over Europe this way, not just UK, AU or DE (latter being me).
According to the US wiring code, you are not supposed to connect Neutral and Ground (Earth), *except* at the panel.
Some bright bulbs jumper the Neutral connection to the Ground connection on 3-terminal (grounded) sockets when the house is wired with the old ungrounded cable from the 50s. This is not considered good practice, because any IR loss on the neutral wire can show up as a volage on the (supposedly) grounded metal case.
Grounding...
>>>"According to the US wiring code, you are not supposed to connect Neutral and Ground (Earth), *except* at the panel."
Well not really. The only place they can be connected together is the FIRST panel, which may be the circuit breaker next to the meter. After that, they MUST be separate. The idea is that there is only ONE place where they are common, and it is early in grand scheme of things. The idea is that the safety ground should carry NO current in normal operation.
Of couorse, your milage will vary depending on the knowledge of the local sparky.
NEC says 'first point of disconnect' which as you say, may be a switchfuse or breaker next to the meter.
Point of order to the ,DE poster, UK isn't ALL TN-C-S (with a combined earth/neutral). Round here we have a lot of TN-S from the 50s and 60s still, and rural areas are often TT (no earth at all supplied by the power company, you rely on your own rod/earth electrode, and need a time delay RCD on the main for fault protection
In the UK it is not permissible (domestically) to run a combined neutral and earth beyond the service cutout (the main fuse). There are various reasons for this, but possibly the most obvious one is that with three different earthing systems domestically in the UK*, it's safest, particularly from a DIY perspective, if the actual consumer installations are all to the same standards.
M.
*known as:
Actually neutral and earth are ONLY bonded together on the supplier side of the cut out fuse, on the the customer side they are always separated (TNC-S - Terra Neutral Combined - Separated)
The rules for the DNO (district network operators aka electricity boards of old) are different to electricians, who work to BS7671 - for example the DNO can and do direct bury split concentric cabling but BS7671 (IIRC it's been a while) does not permit that.
Also the DNO have a lot more leeway in their current carrying capacity rules, whereas BS7671 has a far higher safety margin built in.
There are so many responses with "how earth and neutral are done the ONLY RIGHT (USA) WAY !!1!1!11oneleven!!11!1" that I would love to go to all fuse boxes for each house and flat in my neigbourhood (im Germany) so show: Everywhere, literally everywhere, the equipotential bonding rail is in the fuse box, and the rule is that an equipotential bonding rail is always AFTER the power meter, but in the first fuse box after the power meter. In some cases two rails to have enough connections (like mine), but still in the first fuse box after the meter. It is the same for Switzerland and Austria, where Switzerland is even more picky than we are.
Beyond that some EU countries rather see this as a "recommendation to do" :D, causing various funny effects (Hello France!)..
> Never heard of Romex cables? ...a four core ....cable manufactured by Southwire.
Before you were born, it was "The Rome Wire Company". They "developed Romex to be a flexible, easy-to-install alternative to other wiring methods. By encasing individual conductors in a non-metallic jacket,..." Used to be just white, or grey for underground. Then #12 20A was assigned yellow. Now we are at black and grey, orange and pink and purple.... In the UK, yes, twin&earth, nearly.
Contemporary alternatives would be (bare wire on) Knob & Tube, and Armored "AC" Cable which is metal spiral. Also pipe and thinwall conduit but never in residential except at fire-scared Chicago and Boston (NYC?).
In a US house, the vast majority of circuits are 120V 15/20A, so 2-wire ungrounded or 3 wire with ground. Stoves, ovens, hot water, large air conditioners, anything over 3,800 Watts steady, gets dedicated 240V or 208V on 3 or 4 wires (there's a whole history back of this).
Early Romex was cloth braid, sometimes tarred, frequently 2-wire (I've had a house full of that). Modern stuff has been colored PVC. Rome Wire Co is long gone, Southwire (started 1950, a latecomer) bought many-many 100 year old tradenames a few decades back.
https://wesbellwireandcable.com/blog/the-history-of-romex-cable-and-the-evolution-of-color-codes-according-to-the-nec/
https://www.southwire.com/newsroom/archive/history-of-southwire
Depends on the length, if the loss of several kW is over 100m or so you would hardly notice the heating at any one point. For short cables the usable capacity is normally set by the thermal heating effects, but for long cables it is the voltage drop that usually* does it.
[*] the 3rd limit is being able to disconnect under fault conditions fast enough for safety, that is more complicated as it depends on the characteristics of the protection device (fuse, circuit breaker, etc) and the supply (fault impedance / earthing type).
Then there was the urban legend of the psychic dog that would always know when the phone was about to ring; an engineer was dispatched and eventually discovered that the dog was chained to a pole to which the overhead phone wire was also connected. Apparently the insulation had worn through on the phone wire. When a call came in, the line would energise the pole, which would give the dog a jolt, who in turn barked and let out a startled wee, thus completing the circuit and letting the bell ring!
Biff was a dog who hung out with some of the computer geeks working on the prototypical BSD. Rumor had it he barked his fool head off whenever the mailman showed up, and so we named one of the first email notification programs biff after him. It was "officially" released with 4BSD, in late 1980 ... in reality, he was a quiet dog who rarely barked at all, but was really good at chasing frisbees and helping young coders get a little exercise.
Today, you can try $ man biff on Linux.
That was back in the days of shared service working, a.k.a "party lines". Two subscribers were connected to the same copper pair, and each ringing circuit was formed between one side of the pair and Earth, so a call to either one would not ring the other's phone. There was a clear perspex button in front of the receiver rest, which you had to press to get a dial tone and start a call; each subscriber would only be billed for calls they began. It was almost like having separate lines, except only one of them could use the line at any time -- they could not call each other. (But you could hear the other party if they were in a call when you picked up your receiver. And you could shout at them to stop hogging it .....)
Also in those days, it was an open secret that the closer you were to a military base, the more reliable the local phone service would be.
I worked at a site in Saudi Arabia where there was a large collection of sensitive test equipment that would frequently have trouble taking low level measurements because the earthing rod for the building was not making a solid connection to earth due to the extremely dry ground. When it rained, all would be well until the ground dried out again.
Icon, because despite the local prohibition of alcohol, luckily the expat compound we stayed in had a couple of 'private clubs' where we could get a surprisingly decent home brewed beer.
They also (allegedly) had a policy for a while of answering the phone with the name of the office - so "good morning, Wang London" or "good afternoon, Wang Los Angeles" etc. You get the idea. Very quickly, their Cologne office was exempted from this policy.
I used to work for a company which when I started was based on the fifth floor of an office block in City Road (London). Several times during roadworks outside we'd lose all power, although offices on the other side of the block were on a different circuit and still had power.
One day this happened, my boss went into the empty office opposite which still had power, switched the power on, and plugged a four socket extension cable in, then ran it through into our office, and plugged four more four socket extension cables into the first cable. Then four more cables into the second set of cables. and so on until all 'mission critical' systems in the office were running off one power socket in another office.
Someone tried to move the first cable a bit. It was hot to the touch.
We did this a couple of times and it was a wonder the cable didn't catch fire.
After a few power cuts the boss decided to have a backup generator instead. But that's a whole different horrifying story.
It was the top floor and the meeting room had a skylight. SO the boss bought a petrol (or diesel?) generator and built a box around it to try and cut the noise down, and had a big pipe attached to the exhaust with the idea being this would hang out the window. What he'd forgotten was that engines need air intake to run and he'd not allowed. So the box had to come off. The pipe didn't fit that well either, and even with the meeting room door closed as much as it could be with a wire running out I'm not entirely sure whether it was the noise or the carbon monoxide that meant we all went home feeling horrible with splitting headaches but probably both.
We only did that the once.
What I'm learning from this thread, is...Never let the Boss do (even slightly) technical stuff.
Aside; I'm surprised by the number of mechanical engineers, who think that simply adding a fan, will cool the equipment. You also need a fresh air intake and an exhaust (this seems to get forgotten), of appropriate size, and an ambient temp that's cooler that what's inside the box. Same goes for a heatsink...you can't just throw one on. It needs airflow, a cooler ambient, etc. Also, hot air rises, so bestto put your fan down low and your exhaust up higher, if possible.
Hot Air Rises and Heat Sinks is the title of an excellent book on the subject.
"Aside; I'm surprised by the number of mechanical engineers, who think that simply adding a fan, will cool the equipment”
Indeed, I assume that they think that it works in the same way it obviously works for us. But we sweat, some of it vaporises which moving air removes; so to keep equilibrium, we sweat more and this removes heat from us (see latent heat). Now computer equipment doesn’t sweat so that won’t work, at best the hot components heat up the air next to them which if it can be removed by a fan then fine. But the hot air needs to go somewhere and be replaced by colder air. And so on.
Why don’t supposedly clever people get this?
Long, long ago my University hall of residence had a large dance hall (it had originally been built as an hotel) and our 3-weekly hops (that dates it) were part of the social life of a good chunk of the south London student population. We had a home made lighting control panel that was powered from wall sockets; we weren't daft - sections of it were powered from different sockets. It had a number of switched circuits with a 2-pin socket across each switch. We had a few dimmers which could be plugged ino the sockets - switch on, light, switch off, no dimmer, no light, switch off and dimmer, light controlled by dimmer. One day a friend looking out of his window say a roll of cable fall off the back of a passing lorry (yes, really, the A23 was a busy road) and ran down to grab it. That gave us extra scope for our arrangements....
Back then musicians were all a bit low key but I heard that after I left one of the new lot with lots of flashing lights & what not were booked and blew the whole thing. The College electrician condemned the lot and had a professional lighting setiup installed, something that would never have happened otherwise.
It was a much better place than the dreary newer halls but now it's been flogged off and replaced by a block of flats on the same footprint, Bloody philistines in the College management.
(have told this before)
Fibre optic data and voice mix installed on an empty floor had a temporary power feed from a hole knocked in the wall terminated on a rubber socket feeding the rack. The kit had then been joined by a second rack containing modems and 2 others containing 'servers' (Z80 or 8086 processors feeding green screens)... all feeding from the previous rack.
I was asked to investigate intermittent problems and noticed the rubber socket had burned a mark onto the (thankfully) concrete floor... the socket was naturally red hot
A call to the power guys to fix the problem initially got a P3 rating for attendance, grudgingly uprated to P1 when I pointed out that most of the comms for the 6-floor building rested on that cable
"plugged in a kettle"
We had a multi-hour scheduled outage on the incoming 3-phase supply to the SMB I labored for. Petrol/Gas generator was borrowed the afternoon before and didn't work (Wouldn't start without starting fluid, and only produced a very lumpy 70VAC!). So, no 'backup' supply was available and the plan descended into a graceful shutdown beforehand...leading to total disruption of business. Nicht Gut,
However, Boss's new car was a full EV, complete with a socket providing up to 3KW of inverter-supplied 240VAC. We used the same 'lots of extension cables' trick to power up the essential loads (VOIP Phones, Internet router, Admin/Sales PCs, Networking, one single printer etc.). Things worked fine after a managed transition from 'Shore' to 'Vehicle' power.....
....right up until someone senior[*] wanted a brew and plugged in the kettle. Like most kettles in the UK these days, it was a 3KW 'Quick Boil' one which of course overload the EV Inverter output and it tripped out. I heard the UPS alarms kick off, and went to find out why...and found the kettle still plugged in. I professionally explained that that wasn't possible until we'd moved back to 'Shore power', reset things and re-checked business functions and customer contact systems were OK..
Pretty similar story to @herman and I suspect lots of others.... however ... what makes this different is that the same person deliberately tried using the kettle AGAIN as they didn't believe me and thought it should have worked. This time, my suggestion re leaving our working lash-up alone - or at least asking before plugging things in - was louder, more public and way, way less 'professional' :-)
[*] And who very definitely should have known better!
"....those Wang calculators occupied ~ 1/4 the desktop space that their predecessors, the electromechanical, clackety-clack calculators from Friden and Monroe, did. And the Wangs were silent as well."
Somewhere, I still have mums Comptometer School certificate (99% pass!!). There's a picture of the rather large electromechanical Comptometer on it and it is a sight to behold.
At $lastjob we had a production machine with fairly standard industrial relay controls for the main power (start and stop buttons, power light, E-stop circuit, etc.). On occasion, the "start" button was wonky, you may need to press it multiple times or hold it for an unusually long time to get it to start. This was odd, especially since the machine was rather new.
Several rounds of troubleshooting didn't reveal anything obvious, until one particular day. Opening the front cover of the NEMA controll cabinet required a "coffin key" to unlock the latch. The key normally hung on a string from a hook on the side of the cabinet.
On this particular day, as I went to retrieve the key to begin a surely fruitless troubleshooting round, I fumbled the key which, not yet unhooked from its storage position, swung back and thwacked the side of the enclosure. The green "runcl" light promptly illuminated.
I took great joy in using the Fonzie method to start the machine for a few more weeks before we found the root cause.
The equipment was fairly small, so it was fed from a 20A 120V circuit (US here, in case not obvious). We learned that the outlet feeding the machine was on the same circuit that also supplied several workbenches full of lights, test equipment, soldering irons, and heat guns. When enough stuff was running, the machine was only seeing about 90V. That was enough to run just fine, but the 24VAC control power in the system was non-regulated, and came from a simple transformer. If the 24vac dipped below about 18v, the control relays could hold, but would not pull in when commanded.
Damn, those "wonky US power grid" stories are pouring in today! A flood of unheard proportions! Being off be 20% or more seems to be the norm... Here in Germany +-6% sets off alarm cascades, more can end up in national news, questioning our infrastructure at large, and US don't even shrug at +-10%...
“Damn, those "wonky US power grid" stories are pouring in today! A flood of unheard proportions! Being off be 20% or more seems to be the norm... Here in Germany +-6% sets off alarm cascades, more can end up in national news, questioning our infrastructure at large, and US don't even shrug at +-10%...”
I know what you mean, I listen to a couple of tech related podcasts (US based) and sometimes they discuss emergency generators etc. And I always think, ‘just what the heck is happening to the power systems in the US that you need these things’? I’ve forgotten the last time we hd a power outage in my part of the UK, I think was a substation issue and it literally lasted five hours. Now I understand that in the US you have hurricanes and the occasional tornado to deal with. Fine why don't you mitigate against that, bury the cables underground*
* OK yes I am not a qualified electrical engineer and there may well be very good reasons why it can't be done (technical reasons; not too expensive).
> * OK yes I am not a qualified electrical engineer and there may well be very good reasons why it can't be done (technical reasons; not too expensive).
Nope. Only the money. No other reason. Citing from that video: "That groove (in the power cable hook) took about 98 years to wear in."
For the HV cables I understand that capacitance is an issue when burying them. Wouldn't make any odds to the local LV stuff obviously, but I suspect that the argument went along the lines of "you can bury your local cables all you like but when the HV ones are on the ground you still ain't gettin' no power, boy!"
The cables, at least those mounted in 5 to 10 meters height, are medium voltage. Before the hit the transformer, for your house, between 2000 and 7000 volt if wikipedia US grid information serves right. You don't have that capacitance problem at those voltages, a few finger distance is enough if unshielded, but you won't do it unshielded under ground.
However, if the grid would be reworked with putting cables under ground, they would get rid of the per-house transformer too, crate a big transformer station every few km, a tiny but tall house. Like they did in Germany after WW2. In US it would transform the ~20000V to 115 V (three phase), and then distribute that. I doubt US would switch to 230 V, despite enourmos efficiency gains - unless a future president makes an act. But three phase as default would be a big step up, and getting rid of the often free standing transformer in front of each (older) house too. The fire brigades would thank for that.
Larger customers (i.e. skyscrapers and industries) will still have their own transformer(s) for efficiency reasons, even if they rework the grid.
For comparison: Look at Manhattan, do you se any cables hanging high to supply the skyscrapers? The knowledge is there, the will to do (on a state or national scale) is missing.
We installed fingerprint readers in grocery store checkout lanes and one lane always had trouble. Checking it’s voltage in the middle of the night (when lane upgrades happen) it was fine - but would be flaky during the work day.
Finally we put a logging voltmeter on the lane, and found that it would drop precipitously whenever the nearby electronic door would open or close.
I don’t know if they pulled more power before the startup went bust.
One morning call during my mandatory service (10 months, germany), the guy in front announced that the facility is going to test it's emergency power generators and all unnecessary loads, especially calling out coffe makers, shall be switched off.
There had been two test events. Yours truly named the first event "NATO-Disco". Guess why. The second event went much better.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
