Elevating cost-cutting to a whole new level with million-dollar bar bills While the days may seem to be blurring a little at the moment, we can assure you that it is Friday, which means it is time for another in The Register's series of reader experiences at the hands of the On Call telephone. Readers might remember Jeff from last week, and his A-Team-esque paperclip solution for an RS-232 diagnosis …
Not a call-out situation, but years ago the guys at CERN used to see a strange wave pattern superimposed on test data from time to time. Then for a while it stopped. After some headscratching they realized that the stop coincided with a French train strike. Turns out that running a TGV drawing ~9MW of power from overhead wires creates a significant earth current, enough to deviate the beam in the nearby underground accelerator.
Reminds we of a story boardcast on (maybe RI Christmas Lecture when I was the right age... so a while ago!)
Some academic (so really a post-Grad or post-doc) had developed a really sensitive inclinometer in Aberdeen.And were confused by a not-quite 12 hour cycle in the readings when left sitting on a (very stable) surface.
Aberdeen.... built on lots of granite was moving with the tide.
It takes a not inconsiderable amount of RF to corrupt RS-232 signals, but as the driver and receiver have an impedance in the thousands of Ohms, it _is_possible.
Differential RS485 might be a better choice, lower impedance and higher data rates. Converters are relatively inexpensive (much less than $600) from places like Black Box. Or, you can build them yourself for a few bucks/quid.
Run current loop, and I dare you to find any glitches. Though your bandwidth might suffer a bit.
Run fiber and you'll attract the diggers. But you'll be safe from RF.
Article says 80s, so probably not very since RS485 came about in early to mid 80s. Current loop would've been around for sure.
Quite a few implementations in those days supported current loop option, so may not even had necessarily needed converters. The article doesn't mention distance, but sounds long enough for current loop to be the more sensible option especially since lower maximum data rate would not have been an issue.
Differential drivers and receivers such as RS-422 and RS-485 have greater noise immunity than single-ended interfaces such as RS-232. But they won't do diddly for you if there are severe ground loops, or (heaven forbid) a lightning strike. In the '80s I was an engineer at a supplier of wagering systems to horse and dog tracks. The RS-422 line driver and receiver chips were specced at 25 volts common-mode voltage, max. The computer room was typically in the basement of the building. If lightning struck (say) an upper floor of the grandstand, it would let the magic smoke out of serial driver and receiver chips in the terminals and communication processors. The first time it happened, the chip soldered-in at the factory would be replaced in the field by a socket to ease future repairs.
An even more hostile electrical environment was the communication between the infield display board and the computer room. Typically the infield had an artificial lake, so the ground conductivity was very high -- a lightning magnet. A lightning strike to the infield, or to the grandstand, could cause different "ground" voltages between one end of a cable and the other.
Each succeeding generation of display hardware used lower voltage / higher impedance logic than previous ones, so was less and less reliable in hostile weather. 48-volt telephone relays could stand a fair bit of abuse, but the transistors in the computer room that drove them would fail first. Shift registers made of SCRs with 18-volt logic, not as good. TTL logic with RS-422-style line drivers and receivers would fail if there was a storm cloud anywhere in sight. When the Z80 came along and we wanted to respin the display systems, the naysayers in the company said that there was no way we would get a computer to survive in the infield. But fiber optics had become an option. Even though it was pricey, you didn't need many cores. Our stuff Just.Kept.Working in storms that would have taken the older hardware down for days. Fiber saved our bacon; it won't propagate a lightning strike.
I had a call to go down to Silverstone one day from an F1 team who was using some of my lap timing kit; they wondered if the timing beam (a little metal box on a normal camera tripod) would still work after a "little prang".
I found the box and parts of the tripod embedded quite deeply into the side pod of the car and some of the data cable dangling on the floor.
The box was fine, once I had replaced the data connector, but the tripod and cable had to be replaced.
[icon : I designed and sold all sorts of electronics, computing and telecoms kit to most of the F1 teams, some NASCAR (via Nigel Mansell) and other high-end formula around the world, and did the first in car telemetry in the late 1980's and early 1990's]
I was desperately trying to remember the name of the line drivers. They were a bit pricey at the time but after a while we stopped trying to make connections work on the other side of the building without them as trying without then having to go back and ask for cash for the ,ine drivers always provided a fight with the bean counters. A beer for the memory jogger
Though we were entertained once.
There was a furniture shop on the other side of the road to our office - back when you had computers under your office. Anyway, one day some of the local ne'er do wells had 'jacked' a car and were being pursued by the local plod. They crossed the main road at speed and parked it in the doors of the furniture shop.
I don't remember if plod picked them up straight away or whether they managed to reach the posh back streets.
My experiences with RS232 were rather mundane and I won't bore the commentariat with them.
Beer because Friday
Back in t'day, I lived in a victorian house which was in the heart of a large town - less than half a mile from the shopping high street, and on one of the main commuting routes into said town.
(Which handily meant that there was around half a dozen pubs in the immediate vicinity, of which at least one would have live music or a jam session going on a given night. Which was nice, especially when I ended up being paid to not work for a few months. But that's another story...)
And as this was an Victorian townhouse, the front garden was basically a small gravel pit, stuck behind a short and chunky brick wall surmounted by cast-iron railings that had somehow managed to survive the great WW2 melt-all-the-things purge.
Plus a bus-stop pole sat directly in front of said brick wall. Which was nice (for a very small value of nice), as the bus engine vibrations tended to hit the resonance frequency of the house's windows. Still, it played a part in what was to come...
In the early hours one morning, I was awoken by some strange screeching noises, followed by some rather crunchy noises.
Staggering out of bed, I went to the window, opened the curtains and blearily peered out.
And then I turned and yelled to my housemate "Oi! We've got a car parked in our front garden!"
Some muppet in a sporty little number (A Toyota MR2 or somesuch) had come barrelling down the slight incline of the road, lost control and put his car into a spin, which only ended when they hit the aforementioned chunky wall.
It was actually quite a neat - if accidental - bit of maneouvering - the townhouses were the standard narrow "single room" design, but he somehow managed to completely avoid the houses to either side of mine, and plant his car squarely in our gravel pit. Atop the remnants of said wall and railings, facing backwards, and with the rear bumper of his car barely a foot from the front room's bay window.
This is where the aforementioned bus stop came in handy, as when the car hit it's steel pole, it helped to spin it into the brick wall!
Still, I can think of nicer ways to wake up...
Back in the 60s, a friend who occupied a cottage on a twisty lane was woken early one morning by the sound of a commuter rolling his Mini on to its side in the front garden. The driver got out, called "Sorry, can't stop - see you later", pushed the car back on to its wheels and drove off. He returned to offer compensation for the damage that evening.
Around the same time, I knew someone who swore that the easiest way to underseal* a Mini was to put something soft on the ground and turn it on its side.
* Cars of the day were apparently made of Weetabix compressed to look like steel. If they got even slightly wet they disintegrated.
Turn of the century, I was driving an old VW Passatt for work. Some years old by then, built like a tank, and handled like one. This day, if just left the office and was heading up hill to the main road. An oncoming white van, VW Caddy ISTR, decided it was going to turn across in front of me to get into an office car park. He didn't make it, and I hit him squarely in the side.
The van was written off.. The Passatt needed a couple of new lights..
Which reminded me.. As a kid, someone in a stolen car hit my dad's elderly Datsun estate parked on our front drive, pushing it into the opposite wall, squishing both sets of lights. Rather than spend many times what the car was worth fixing the mess, he just wired up some trailer lights in their place. It looked dreadful, but worked, and the car passed its annual MOT for many years thereafter..
Mid 80s. Bought a second hand Renault 20. A bit of rust on the wing turned out to be following the line of a rubber seal on the inside that trapped moisture. i.e. rusted through.
Never trusted Renault since that piece of crappy design. That and the sealed beam headlights that costed a small ransom to replace when a bulb went.
77? Ah yes well, that was your mistake, wasn't it?
'round about 1977 the RNUR decided that their little R5 was selling quite nicely, thank you very much, so how about cutting some corners then?
One brilliant idea was to replace the original pressed floor with a double-skinned version. Now don't ask me why welding two sheets of steel together to replace one should be a cost-cutting measure, but that's what they told me. Either way the early cars didn't seem to have an issue with the floors, but the 77-onwards ones became thoroughly Flintstone-esque after a few years of winter salt. The early ones did have some issues with the sills (I seem to remember almost all 1970s cars did, including Mercedes) and also a special muck trap above & behind the rear wheels that used to rust out the rear wings.
One of the cheapenings was to replace the speedo for one with a plastic needle. After a few years in the sun the needle would bend up towards the 'glass', eventually jamming the speedo completely, although that might not have been an issue in the TS - always wanted one of them *sigh*
I had a few of them over the years and the pre-77 ones were definitely more substantial.
It wasn't that long ago that certain models of Renault cars had a join in the wiring loom under the carpet in the passenger footwell. All "fine" unless it was repeatedly trodden on by a heavy passenger, crushed by a heavy object or when the car leaked and it got wet (in which case the standard "fix" was to just drill a hole in the floor of the car)
As a kid I remember one day my dad's Peugeot 504 slid on an oily patch (t'was raining) and pranged a double decker bus at 90º
A large steel plate fell from the side of the bus, my dad and the bus driver got out of their respective vehicles, they lifted the plate and put it back in its place, then checked for any damage in either vehicle. Nothing. They just went back to their vehicles and drove off.
With today's cars there would be plastic pieces scattered all around and the car's front hood would probably look like a metallic accordion...
A further irony. In the old days a panel would be pushed out where it was dented. A bit of paint and done.
Now it's repair garage assessor sucks in teeth. "That'll be a whole new sill, - can't push them out they're made with a double skin. That bumper will need replacing, then spray to match and shade in with that side panel. £2000 and I can fit you in two weeks on Wednesday, possibly.
And yes I've been through this a couple of times over the last few years. Daughters and wife scraping cars on concrete bollards or low level islands at petrol stations etc. (To be fair, impossible to see from inside the car).
Each time a bit of a scrape has turned out to be a major engineering task requiring three blokes who normally repair oil platforms and one who paints the Forth Bridge to repair it.
A friend of mine had a very minor argument with a bus in his Tesla, just a small scrape and dent in the front wing.
The bus company were reasonably happy to accept responsibility, up until they got the £20,000(!) repair bill.
(That's more than all the cars I've ever bought put together. Hell, it's close to the brand new price of all the cars I've ever owned, put together)
> The bus company were reasonably happy to accept responsibility, up until they got the £20,000(!) repair bill.
To save weight, because every ounce of energy is precious (being directly related to the range before needing a long charge), BEVs are made almost completely with aluminium.
Aluminium is a PITA to weld, you need to use a TIG welder, and correspondingly needs a higher skilled weldor. As a result, while pretty much every garage can weld steel, very few specialist places can weld aluminium, and the costs + labour are correspondingly higher.
Especially with electric vehicles, as you have to be very careful to not ignite the battery pack while welding, and make sure it is electrically isolated/safe.
Saying that, more and more ICE cars are also going to Aluminium bodies to save weight (and improve economy), so I expect they will have similar repair costs.
"Aluminium is a PITA to weld"
?? Really? I don't find that to be true at all. Now iron, that's a PITA ...
"you need to use a TIG welder"
No, you don't need to use TIG. MIG works just fine.
"and correspondingly needs a higher skilled weldor"
Nah. Just a welder with a different skill set.
Yes, but in those days a lot more people were killed / seriously injured in car accidents. Those crumpling effects are deliberate – they take a lot of energy out of the collisions. For an IT angle they take a lot of computer power to work out how to do.
Many years ago I came upon an accident on a narrow, winding, but busy road. It had happened a short while earlier and the police had arrived but not the ambulance.
A dozy blonde bint (misogynistic? Perhaps. Accurate - definitely) had come around the bend in her fairly new Datsun with her wheels planted evenly on both sides of the solid white line. Travelling in the opposite direction on the correct side of the road was a man in an NSU Prinz - it was vintage even back then.
The Datsun was a mess, driver's side front mangled, car leaking fluids. The woman was sitting on the curb in hysterics. In contrast the NSU was barely damaged and the driver was sitting quietly behind the wheel.
He was quiet because he was dead. The blonde's Datsun had crumpled and absorbed the energy of the impact; the NSU had resisted and transferred the force to the driver, whose neck had broken as a result.
The "plastic pieces scattered all around " is all too true for today's cars. I live in the rural area of the US. When I see plastic on the street in front of my house I know Bambi forgot to look both ways before crossing the street.
A PhD student of my dads had a 1930s cabriolet that was made from plate steel, One day she was pulling out of a side road on the edge of our village which had a massive wall that cut of the view of the road after 100yds or so.Despite the 30mph speed limit she was hit on the drivers door by a motorcyclist doing about 60, The panel didnt bend far enough to hit her and the motorcyclist was in a bad way, his bike was destroyed, and I mean destroyed. It took a panel beater a couple of days to restore the running board and door panel a lick of paint and it was as good as new. Most modern cars would have been written off!
Back in the 1960s I was at a service stop on the RAC Rally in North Wales. A BMC Mini came in and needed work on the exhaust so the crew got a mattress off the roof of the BMC 1800 service barge and put it alongside the Mini then tipped it (the Mini) on its side onto it. Quick and simple!
In a similar vein; living as a child in a Victorian vicarage halfway down a fairly steep hill, I woke one morning to a loud bang. Finding the bed still level I apparently went back to sleep. Meanwhile, the grown-ups were investigating the 30-ton artic that had jack-knifed on the hill and planted its cab into next door's kitchen.
I recall a lot of acro jacks holding the wall up when they took the lorry out...
There is a tiny road off the main town through road (also small in comparison to most) snuck between the tiny co-ops car park and a bus stop in our town center. The road is one way, and hits 2 other roads in a 3 way junction, then another 20 foot and you get another 3 way T junction onto an off road that heads out of town down a country route (not main route).
Somehow. Impossibly. In the space of 20 foot, someone came off the main road at speed, around the corner *fine* but planted the car into the corner of someones house. Even weirder they went through a metal railing and a brick wall, and have near demolished an entire cornet of this house (rental, so thankfully those living there are now rehomed).
I can see they did it. They crashed at speed. But unless dive bombing from the air after jumping out a helicopter, I have no idea how they managed it. There is just not the space or cornering to get the car in that positions. They deserve a medal for just doing the impossible!
Ah, takes me back to the early hours of January 1st sometime in the early nineties when my “first-foot” was a stolen Vauxhall Corsa. The driver had lost it on black ice and ended parked up in my front garden almost touching the house completely demolishing the garden wall. The driver and passengers managed to get out, carry-out intact, and legged it leaving me dealing with the fallout and a very tearful owner in the morning. Not easy when I had a monumental hangover myself.
In my old village there was a moderately steep hill with a humpback bridge at the bottom and a slight kink to the left after the bridge. The wall in front of the house on the right looked like it was part of some endless stop motion animation where as soon as it was rebuilt someone would knock it down again. Points were given for how high up you managed to hit it.
We get similar here with the big old farm house on the corner. The town is now built around it, but the farmhouse arrangement is still there. So there is a tight turning from the cul de-sac. The bin lorry and gardner with a trailer do fine. But if anyone ever gets gravel delivered, or bricks off a big truck with a trailer, the farm house loses it's gated entrance wall again when the trucks turn to leave out road.
I think in the first 5 years I was here it got knocked down 3 times. I do wonder if those knocking it down, then returned with the same truck to deliver the bricks that were used to put it back up!
Back in the '80s we didn't need to get up to discover it was snowing. We could lie in bed and hear cars crashing into the gate-post across the road. We were on a site on the inside of the corner. It took scarcely any snow for some drivers to fail to make it round the corner properly.
But one day I did see a Fenwick crash an entire rack of servers.
I was parking at a customer site when I saw a Fenwick barreling down the loading dock with a rack of servers on its enormous tongs. I remember thinking "this guy is going a bit fast with that", then I saw two guys running behind the Fenwick, shouting. I stopped the car and got out when I heard an enormous crash. When I turned around, the Fenwick was stopped, the rack was lying on the ground, and there were three guys arguing.
Turns out the two guys who were shouting made the driver think there was someone in front he couldn't see, so he slammed the brakes, which had the inevitable result.
The rack was fine, if a bit scratched here and there, and it turns out that only a few hard disks had bought the farm. The kit was new, so I guess insurance paid for the disk replacements. They had one hell of a time getting the rack back up again, though. I don't know what happened after that.
I'll never forget the almost cartoon-like quality of those two guys running behind that Fenwick.
"For the Yanks in the audience, a "Fenwick" in this context is an approximation of a Yale Forklift (as built by the French), not a fishing rod."
Explanation not only needed for the Yanks, the only Fenwick i had heard of before was the department store.
https://en.wikipedia.org/wiki/Fenwick_(department_store)
I've never heard anyone call a forklift a Fenwick.
i guess its like us calling a digger a JCB or a vacuum cleaner a Hoover, but Fenwick is hardly a household name in the UK.
Indeed. As a native of Newcastle I have only ever known Fenwick as the department store (on Northumberland Street) and by derivation as a place where one might expose oneself if proven wrong, as in "if the Vega guys ever ship any product, I'll bare my a**e in Fenwick's window"
JCBs have made it all the way down under.
In an environment where average daytime temperatures are in the low 40s for half the year (not going to convert that to some horse-blood based approximation of a scale), JCBs are noteworthy for having really good heaters, and utter crap air conditioners. If only it would snow, they would be fantastic...
"Servers on the tongs" was enough for this Yank to figure it out. Even though I had to guess that "tong" was the forks on a forklift.
I never drove a Fenwick. We only had Hyster's, White's and a Clark at the lumber yard that paid for most of my University degree while working minimum wage.
Picture a data center in the basement of a tall building in San Francisco's financial district. Card punch up against a wall, near the ancient Otis heavy goods lift. Every now and again, at seemingly random times, the punch generated errors for a couple characters. Nobody could figure out why, not even IBM's field circus dudes.
Until IBM was traipsing in and out one fine weekend, upgrading who knows what hardware, as only IBM could. Someone (ahem) noticed that the gibberish was being generated about ten seconds before the elevator doors opened.
Turned out that the motor for the lift was drawing so much current when it first started that it was inducing errors in the punch on the other side of the wall. Nobody put two and two together prior to this because the lift rarely went into the basement (that level was key-protected) ... until IBM was in and out that morning.
"Nobody put two and two together prior to this because the lift rarely went into the basement"
The number of "lifts and IT equipment" stories i've heard over the years always made me check for shielded cable and correct use of RS485 instead of RS232 for long runs in such installations as a matter of course when problems were reported.
I've got a set of my own EMF interference stories but they tend to involve radio transmitters rather than electric motors
My days were mid 80s onwards. Yes you'd use RS422 or current loop in olden times.
I actually did build and maintain a few 20mA current loop setups when playing with RTTY as a spotty youth - mainly to tame the prodigious EMF kickbacks you get out of old style Creed model 7 teletype.
I did a bit more with loops and optoisolators in later years interfacing various ancient technology in university laboratories to "modern" computers
"the lift rarely went into the basement"
I wonder what the connection with the lift was then, because the motors for the lifts are stationary, and usually at the top of the shaft. Even if the motors were in the basement, you'd expect them to be causing the same amount of interference no matter where the lift is moving in the shaft.
Perhaps it was the small motor that opens the doors?
The motor was in the basement. Behind the wall that the punch was up against. The lift wasn't a standard passenger lift, it was the heavy goods lift, and rarely used. The card punch was used as necessary, but wasn't always in use when the lift was in operation. Once I figured it out, and could reproduce the problem at will, a little shielding (spec'd, provided and installed by IBM, gratis!) made it go away permanently.
"Perhaps it was the small motor that opens the doors?"
It's invariably a combination of the motor operating the doors, switch arcs and long unshielded runs from the car to the relay logic in the controllers in the hoist room
Dry risers and building comms risers invariably run beside or inside the elevator shafts so you learn very quickly that anything with high impedance ends sees a lot of capacitive/inductive coupling and that no matter how many times a building might have steel covers specified for the trays to achieve electromagnetic isolation, they're either never installed (cost savings!) or never replaced (lazy arsed electricians who don't see the point of them - which brings up a point of one sparky firm who'd been contracted from day one on a building and were forced to pay for replacing every single missing one they'd "lost"....)
Many years ago I got called out to a mysterious issue that was bringing down a workshop's servers every Monday morning at about 08:30.
Systems were checked; cables replaced; even one of the power supplies changed, all to no effect. The environment was checked and it was established that nothing out of the ordinary was happening at that time.
Eventually it was decided that I had to go and spend a Monday morning on site to try and catch it red handed. For the first hour everything worked normally and then silence. After a few moments the servers came back to life and started rebooting.
It turned out that the workshop used a lot of air tools and had a huge compressor tucked away in a shed outside. Usually this cut in and out throughout the day with no difficulty to keep the air reservoirs topped up.
Over the weekend the ancient grease in the compressor's bearings essentially set, stalling its electric motor until it overcame this resistance, and stirred the grease back to a normal thickness - which it maintained until the next weekend.
During this period the supply voltage on that one, somewhat flaky, circuit dropped to about 30V, and needless to say, the servers were on the same electrical circuit.
My boss quoted for a complex solution involving UPS and the like, but the workshop simply oiled the compressor instead ...
"Not just a case of of curing the fault, not the symptoms, but also a case of not every problem needs a high tech solution."
There's still a problem there (three problems actually(*)). They just punted the problem down the road
(*)
1: servers being knocked out by voltage drops
2: The VERY REAL RISK OF FIRE if voltages are dropping that far due to current draw without blowing breakers - meaning there's bad wiring somewhere
3: very poor maintenance of the equipment
Probably not yearly, if ever. The maintenance manuals on some of my old kit recommend liberally greasing the fittings regularly, and specifically recommend NOT disassembling them to replace the old grease. Seems that taking it apart causes more trouble than wiping off the excess grease squeezed out between the parts, which had the effect of replacing the old stuff over time. They do recommend repacking the cordage on a regular basis. (Insert locker room humo(u)r about packing glands, stuffing nuts and grease nipples here).
Note! Do NOT do this with modern equipment! If you pump enough grease through the zerks to cause it to blow past the seals, you have defeated the purpose of those seals ... and now you are allowing the introduction of dust and dirt into the bearing surfaces. Dust and dirt + grease has a name: Grinding Compound. RTFM for info on maintaining your specific kit.
No, really, I mean it. Read The Fucking Manual. Do it today. Don't say I didn't warn you.
Back in the Dark Ages when I was in the RAF the maintenance manuals for vehicles contained phrases like "Every 3 months give this nipple 2 squirts of a grease gun". This was quickly changed to "Every 1,000 miles give this nipple 2 squirts of a grease gun" when a large number of vehicles developed blown seals because the vehicles were in storage and only travelled brief distances every month or were even stored up on jacks.
before the big printing press was turned on, as it disrupted the RS232 signals between the control panel and the actual climate control unit.
Replacing the cable with a length of fibre and a converter on each end was the most effective solution, though a bit pricey, but also reduced the amount of environment-related problems of the printing press.
What is all this talk of "line drivers" and fibre converters, when RS485 was designed for this task, has been around just about as long as RS232 and can actually be quite cheap? Neater solutions are available, but this would work as a quick fix, even on standard serial cable; proper RS485 cable is different, but if a system "nearly works" on RS232, converting to RS485 would probably do the job just fine.
M.
First server in the building (Z8000/Unix v7). TPTB decided that it could be located in a cleaner's cupboard small room next to the lift shaft. Experience gained led to its being rapidly located to the library.
The other computer and lift shaft episode was much, much earlier - the computer was probably an Elliot - from the company where my Dad worked. It had to be located on an upper floor of the building so it was decided to haul it up the lift shaft. The rope failed. According to Dad the man who'd tied it on* fainted which might have run in the family; his daughter was in my class at school and fell off a biology lab stool when "blood" was mentioned.
* For added irony his surname was Crane.
Our people fitted a new central process control unit in a plastic film factory. This had just about every test imaginable before being fitted including being hit by brownouts and static discharges. However, once installed at the location demanded by their engineers all sorts of random errors were occurring. This was in the middle of the factory on a partition attached to the back wall and at the top of a flight of steps. It was months before we identified the problem, and only then because one of our engineers took a laptop there to make notes of readings and it kept crashing. One of their production workers commiserated and said that their mobile phones would sometimes freeze if they went there. In a moment of inspiration our guy asked what the partition was there for, and was told the 3 phase supply for the entire factory complex was behind it.
I did massive amount of Cable and switch replacements in a small school to try and fix an intermittent drop in the network. It happened mostly at lunch times but at random through out the day, it was only luck I was in the staff room one day when it happened. Some one used the ancient microwave in the kitchen area, seriously it must have come off the arc, and that was when it all dropped. As I said this was a small school and the server room was off the staff room, most of the cables were trunked behind the microwave in question. Thankfully due to its closeness to the server room, I was able to pull though enough slack to move the trunking up about a foot and fix the issue. We also added tinfoil to the inside of the trunking and added a wire to earth it as a little extra protection. To be honest it would have been better to replace the faulty microwave, as it must have been slowly cooking their staff too, but they were happy we had fixed the issue...
"To be honest it would have been better to replace the faulty microwave, as it must have been slowly cooking their staff too, but they were happy we had fixed the issue..."
Was this before or after the tinfoil hatters started blaming mobile phone masts for all their ills? I'm amazed those staff were not falling like nine pins with debilitating headaches or worse!!
Had a similar head scratch situation whereby a customer would be in the process of buying something when the stock would suddenly disappear from the inventory mid transaction.
This was back in the early nineties, we were a catalogue store not unlike Argos, the customer would select an item, the sales person would enter it into the till, process the order and the item would be picked from the stock room, but occasionally the item would disappear from the inventory between the software finding the item in stock and then processing the sale.
We had the IT people in several times, they couldn't locate the glitch, and the disappearing item would always reappear the next day after the End of Day routines were run.
Then, one day the software guys were on site with all kinds of gear when they at last spotted the stock disappear from inventory, and they saw just why it happened, it was when the store manager walked past the server room when on a call on her analogue cordless phone.
They swapped her phone for one of the new fangled DECT phones and the problem never reoccured...
... and point the finger of guilt at who ordered the wine and which greedy so-and-so had three starters.
when a cheap and simple SEP Field generator would have made the problem disappear.
I hear a 5G mast (the more visible the better) works wonders to generate a SEP field...
...it doesn't even have to do 5G. A decoy is much cheaper for the same effect!
Not quite computer equipment but when I spent a few years in silicon valley at end of the 90s I used to look at the USGS web pages showing all recorded eartquakes and noticed that there seemed to be small tremors on a regular basis near where we lived in Cupertino on the map. After a bit I went to the more detailed pages to find the records of this "quakes" and discovered that happened on almost every weekday at 12 noon with the USGS comment saying "probably quarry blast" ... and there was a quarry just outside Cupertino. N.b. these "quakes" were 2.0 magnitude and one of these on a daily basis from a quarry seemed to cause no qualms which, for me, has always put the alarmism over occasional 1.0 tremors (10x les powerfull) being sufficient to close down fracking into perspective!
https://www.google.com/maps/@37.3121889,-122.0945717,5670m/data=!3m1!1e3
The Steven's Creek Quarry just south of the larger Permanente quarry also used to have blasts on the USGS page, but at more random intervals. I worked on the original USGS backend for the web page that your used to eyeball, contracting in the Menlo Park offices. It was a very early example of what the WWW was good for. A newer site is available here.
Note Apple's Infinite Loop campus in the upper right hand corner of the go ogle map.
I'd love to be able to measure the tremors that happen every time a bus passes my parents' house. They're strong enough to make the monitor wobble on its stand, CD cases rattle in the storage rack and dust to fall from the 5mm crack that has opened up between the wall and the coving since the bus route opened about 5 years ago. (For the previous 25 or so years that we/they lived there, there were no buses along that road.)
Where I currently live, the nearest I get is bits of plaster occasionally falling from around the window frames when the Army are playing with their artillery on the nearby range, but I think that's caused by atmospheric pressure waves (which I can sometimes feel if the windows are open) rather than ground vibrations.
" They're strong enough to make the monitor wobble on its stand, CD cases rattle in the storage rack and dust to fall from the 5mm crack that has opened up between the wall and the coving since the bus route opened about 5 years ago."
FYI: UK councils are _required_ to ingestivate complainst about this kind of thing and deal with it. Whilst traffic noise issues are exempted from environmental protection legislation, traffic _vibration_ shaking buildings is not - and the liabilities if they ignore complaints can be stupendous - as Croydon council found out after traffic vibrations caused the frontage of a building to fall off a few years back.
I remember a school science class involving a lash-together rig complete with one of those tuning knobs that connected to a "variable capacitor", i.e. metal plates moving in between other plates with air as the dielectric. Try as we might we only got the "Third Programme" as it then was. I guess that's the price you pay for living a few tens of miles from one of the main transmitters in the region.
IME, living maybe even closer to the main AM transmitter, Third was more or less inaudible below the howls from the line oscillator harmonics of any TV set in the vicinity. What was worse, Third wasn't even transmitted from that site. It was, according to my cousin-in-law who worked there, transmitted from the even closer TV station on a low power with an aerial suspended from one of the stays holding the mast up.
it was all in the diode you used, and the length of copper wire you used as aerial. Back in the 1960's the first radio I built was one valve, a solid state glass diode (type not recalled) headphones and i had 200 yards of 24 gauge enameled copper wire running out the bedroom window and out over the fields as aerial. I recall i could get about ten stations on long wave. More interestingly, i made a Gold leaf electroscope (actually an aluminium foil electroscope), out of a 200ml glass beaker, which when connected to the aerial was fantastically sensitive when black clouds passed overhead. The possibility of lightning actually striking the aerial never occurred to me............
"Years ago, I was checking a cabling infra in Paris, only to hear a perfect radio signal in the controller."
One of the more annoying problems around landmobile repeater stations is that the wire fences of the surrounding farm boundaries fences act as antennas.
Which wouldn't be so bad, but fence wires tend to be nailed into fence battens using galvanised staples and when things rust (as all things left in the weather eventually do), some of these turn into perfect diodes.
If you have 100-plus transmitters all turning on and off, combining in antennas and being rectifiied in diodes, then you end up with all sorts of nasty intermodulation products and sooner or later some of those intermodulation products show up at the same frequencies of the receivers. At that point you can end up with a cascade effect.
At some point someone has to wander around, hitting said fence battens with a hammer to dislodge the rust, and encourage farmers not to use galvanised wire fencing.
in order to itemise a customer's bill and point the finger of guilt at who ordered the wine and which greedy so-and-so had three starters.
I cant wait for the day this happens.
Restaurants are supposed to be about service , so why cant they do something about the awkward scrabbling around to get the bill together, even if you are splitting it evenly its difficult.
Ideally everyone would swipe a cashcard on the way IN then , eat whatever they want, get billed appropriately and piss off at the end without waiting around for the bill and then discussing it for 30 minutes whilst missing last orders at the pub across the road
I was installing some software on the receptionist's PC at a client in an office block in the 90's. The screen glitched sideways severely while I was doing it and I mentioned it to her. She said "it does that all the time. They've replaced my screen 3 times and it never goes away."
I sat there finishing off the install taking in the environment and then said - "let's try something". I moved the screen to the other end of her desk and the glitching stopped. She was overjoyed and wanted to know how I fixed it.
I told her the screen was right up against the wall and the glitch happened everytime the elevator went past.
Pretty sad their hardware guys didn't figure it out.
Zilog had their own range of servers which were free-standing boxes, none of this rack-mounting malarkey. We had several standing in the server room, each with its console standing on top. The displays of the big servers were stable, those on the small servers had a permanent shimmy. I assume there was some stray field from the servers that didn't quite match the frame rate of the displays.
All second hand but, long time ago in Sydney, an expensive minicomputer was being unloaded by specialist computer/fragile equipment specialists from truck loading lane, outside the premises where precious computer was to spend next decade. Sydney taxi in a hurry rounded corner in left lane and high speed. Collected computer and half reloaded it back into truck. Write off natch.
"Long time ago in Sydney, an expensive minicomputer was being unloaded by specialist computer/fragile equipment specialists from truck loading lane,"
Electrronics Australia recounted the story many years ago of a regional newspaper in NSW which got a colour printing press and new computer (secondhand). The delivery company transporting it from Sydney put it all on an _open_ delivery truck which was also carrying fertilizer to save costs and didn't liberally encase the fragile bits in plastic wrap.
1500km later, AU$2million of computer and about the same of printing press ended up being a complete write off, as was the transport firm when it got the bill and its insurers refused to pay up.
Many years ago working for a department with overseas branches, the customers at very remote site complained that sometimes their local unix computer was slow.
Dialed in and checked sar. No load to speak of, just system processes and an idle database. I did note screen refreshes were sometimes slow. Thinking network noise I tested network with ping, steadily increasing packet sizes from 64 bytes up to 4K. At 1 Kb size 50% of all packets were being dropped. Bigger and almost all dropped. A nearby colleague had visited the site in previous years, and mentioned it was down a long dark corridor in deepest subbasement basement. I rang the site admin and suggested they check the corridor fluorescent lights. On a whim I suggested about 75% of the corridor length. Two hours later a phone came in asking how the hell I knew a light was blinking on and off in seldom used corridor. Ah, the joys of deduction and luck
In a different life I used to pollute the Chemistry labs a lot, which is where I found an innocent young undergrad hooking slender wires to an oscilloscope. The object was to measure the micro-resistance in a small liquid filled cell undergoing transformations of various kinds, but the trace on the screen was pure noise. Recognizing the threat posed by ubiquitous fluorescent lights, not to mention megawatt radio transmitters just across the border in Mexico, I helped with some judicious tinfoil wrapping of the wires (hat optional) and saved the day for this young man.
However, not all threats come at high frequency. Here at home we were having strange problems with that new-fangled digital TV coming in over your basic co-axial cable from the outside. One entire local channel would pixelate and drop off the map, along with all its sub-channels (which tend to get used for cheap and cheerful re-runs of classic shows). Internet was not wholly reliable either, and even our POTS line to the outside world had a mains hum on it. I suppose that last bit was the clue, because I suspected the outside mains power line that had been upgraded. Sure enough, both POTS cable and co-ax were pinned against the house by the new power line, which was carrying two-phase 120V and would have all the noise that you get these days thanks to electronics, dimmers etc. A little rearrangement, and all was well. Social distancing is necessary for intelligences of all kinds.
I posted this last year:-
In the 1990s one of our customers moved to new larger premises as their business expanded. They had started with a single computer, and then a small 10BASE-2 network with the cable carefully snaked around their main office (tied to ceiling panels, run under carpet edges etc.). Our software ran on a small server in the corner. Fortunately we had persuaded them to install network cards with both 10BASE-2 and 10BASE-T connectors as we knew that they had bought land to build new premises. A couple of years later we were asked to help move the network to the new building that they had built themselves (they were in the building trade).
They had build a small comms room under the stairs with a couple of 19" rack slots and square section steel drainpipe conduits down the wall for the cabling. We bought them a 16 port switch to temporarily connect two computers in the reception areas, and a new server placed under the receptionists desk. We moved the software to the new server - It worked. The owner said that the "proper" cabling was being installed over the weekend next week (Yes, AFTER the building was built), and asked us to move the server and switch to the comms room on the Friday after they closed.
On Monday they phoned us and said that nothing was working and we needed to be there. When we got there they said that it must have been us moving the server that broke everything as no-one could log-on. After faffing about for a while we realised that it must be the cabling. I disconnected all of the wiring to the switch except for the server and connected the nearest reception computer with a 10m ethernet cable - The receptionist could log in. After experimenting we found that a couple of users in nearby rooms could also log in, but when we connected up the others everything stopped working. We got blamed for recommending the fancy new networking when the old coax stuff "had worked fine". I made up a ~30m cable and ran it from the switch, up the stairwell, to the bosses office in the upstairs corner of the building - He could log on, when we connected "their" cabling he couldn't.
I asked who had done the cabling - It was his brother in law, who "knows what he is doing, he's an electrician". Oh dear, the conduiting went past the wiring and motor for the lift and the main air-con unit. We pulled one cable and saw that the sheathing was damaged where he had pulled it through the metal conduit. The business owner got a mate's brother, who's business was actually cabling, to rewire it properly. They earthed the metal conduit and ran lengths of ABS piping down it from the top and put labelled patch panels in at the top and bottom. I suspect that the brother in law didn't get paid.
I had a similar experience with elevators.
Back in the day, MCI was trying to expand their business from long-distance into local markets.
The networks, and therefore the network management centers, legally had to be separate, so the local Network management center was set up in the same building, but with a different physical setup than the LD side. Data (alarms, Performance data, etc) was collected from around the country via x.25 circuits, into the basement, and then consolidated and sent upstairs (via unshielded, untwisted pair cables) to the element managers servers and workstations in the local service NOC.
I was sent to North Carolina to find out why the connectivity was lost for long periods of the day. The NOC was totally blind to the network for hours at a time.
I noticed that the problems started at shift change. The elevators were inducing noise on the lines between the basement and the third floor, causing the connectivity to break, and backing up the data at various networks around the country. It would take several hours for the lines to clear,
Worked with a bloke whose first job was doing Crystal Reports (or something similar) to generate annual safety certificates for lift inspections. After a few years, he realised there was a bug in his code which automatically approved *every* inspection. He quietly fixed the issue and was *very* relieved there were no incidents.
Rs422 would be better as it uses a pair of wires in each direction with a larger voltage difference so is more suited for long distances. The protocol clearly needed a checksum as well... I know things were 'simpler' in those days but we new about checksums and comms (hell I was even taught it in college way way back in prehistory when comms were achieved by cutting marks in stone). If you do the right type of checksums you can even correct the message when it is scrambled. Even without some form of data validation is quite a good idea... if the number is beyond plausible then throw the communication away. And then there is a need for acknowledgement and confirmation. Seems to me that the problem here is rs232 - a perfectly good standard - but the use made of it. Its a bit like saying matches are bad because some idiot tries to use the stick instead of the head to light a fire
Installed a computer in an office and a controller terminal and printers in a factory. The terminal and printers communicated with rs232 as they were nice and close, the terminal and computer with 422 as they were a long way apart. The comms of course had checksums etc so we never printed anything we didnt want to.
It worked well from when it was installed all the way through winter and into the spring, then in summer it stopped. I went up to investigate, put on the debug mode and the computer was seeing the messages it sent, which it didnt like and rejected. The terminal was similar. I checked out the connectors (ip57 in the factory and very tough), nothing seemed to be wrong. I buzzed out the wire, very very odd, I seemed to be getting short circuits. All of this took most of the day (it was a good long walk back and forth). Eventually I spoke to some of the guys and said we would have to trace the wire all the way, something seemed wrong. They did, right up to the factory roof where they found that an automatic vent opener had activated in the summer heat, opened and cut the cable.
I remember about 5 years ago, a dot matrix Oki printer in an airport was relocated by the airline from under the check-in desk to a counter on the back wall so that multiple agents could easily get to it.
Unfortunately when they moved it, the guys who recabled it bought the cheapest and nastiest serial extension cable they could find from a local store, and helpfully tucked it out of the way underneath the baggage belt next to that check-in desk, inside the metal shielding.
We spent weeks afterwards trying to diagnose sporadic printing problems remotely, before someone at the airport actually admitted that they'd moved it since we originally installed and tested it. Once they sent us photos of the cabling job, it didn't take a genius to figure it out - turns out if someone was sending a print job when someone else dispatched a bag on the belt, you'd get - at best - pages full of garbage; at worst, lines missing in an otherwise seemingly normal print, which might not be noticed until after a plane had been dispatched and was airborne...
This post has been deleted by its author
Two quarries in South Wales. One with a System/36 and the other with a cluster controller as a satellite. Leased line and sync modems linking the two, as this was back in the days when a kilostream came with a monthly bill best described as "eye-watering". After initial setup it was as unreliable as fuck.
You know the standard BT drill: 1) "Line's OK, must be your modems"; 2) "They're your modems"; 3) "We'll test it again"; 4) GOTO 1.
I eventually got to speak to the bloke who did the testing who, it turned out, worked nights <LIGHTBULB ON>. After much escalation, BT eventually dragged him in during daylight hours and once it was shown he didn't burst into flames, we tested it.
Very Grumpy Engineer: "There, it's perfect."
Me: "Yes, I have good carrier.....hang on....it's dropped."
VGE: "Well it's all showing as good h.......it's gone......now it's back."
Me: "Carrier's back up".
Rinse and repeat.
So, this produced the Holy Grail, an on site visit from BT. On investigation locally, it turned out that the engineer who installed the line didn't have a long enough drill to go through the stud wall between the back office (modem and cluster controller) and the weighbridge where the line came in. The office had a phone in it, so he'd borrowed a couple of pairs off that cable. One was iffy and every time a truck went over the weighbridge outside, the line dropped...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
