'The capacitors exploded, showering the lab in flaming confetti' Welcome to the very first edition of “Who, me?” a new Reg column we hope will prove as entertaining as our Friday On-Call tales of tech support gigs gone wrong. In Who, me? we’ll celebrate the times techies stuffed up, the lessons learned and the career consequences. To kick things off, meet “Alvin”, who “In the early '90s …
I recently had a similar, though far less violent, incident. I'm working on wiring the garage. There's no breaker panel yet, the whole thing is just run off of an extension cord going to the house (just for now!). With everything nicely lit up from the overhead lights, I was trying to figure out how a line coming from a turned-off switch had 4V-to-ground on it. Hmm. Can't be much current, I wonder if it's "phantom voltage". I switched the multimeter to 10A measurement and put its leads on hot+ground. The building instantly plunged into darkness. I slapped my forehead and went to reset the GFCI in the house.
The original TRS-80 Model I had a reset button on the left rear. Right next to the completely unbuffered expansion port's card edge, which applied inputs pretty much directly to the CPU & DRAM.
So of course when you went feeling around the back for that pesky reset button, you usually ended up groping the expansion port, and if you had any static buildup whatsoever, you could kill the TRS-80.
Yup. Did it. Fortunately it was still in the miniscule 90-day warranty.
When I was a kid, about 13-14, I got my start in electronics at the local radio club and I can still remember when a guy working on an audio amplifier screwed something up driving a dummy load, the EL34s all started glowing bright blue and we were all standing around looking at them when the power supply went TITSUP and a couple of 300V electrolyte caps turned into steam and confetti - YEH!
Young EEs are lucky to be toying with low voltages.
Back in the vacuum tube era, the voltages were much higher. I remember being in middle school, wiring my 65 watt CW transceiver in the bedroom. I learned about paying attention to polarity on electrolytic capacitors one night. Night of the flaming confetti. Dad wasn't impressed.
"I learned about paying attention to polarity on electrolytic capacitors one night."
I forgot to rate the psu capacitor for 1.4 times the transformer's nominal voltage. Measuring the voltage it started rapidly dropping - followed by a bang and confetti sprayed over my bedroom wall.
When I built my Motorola 6800 evaluation board I didn't realise that tantalum capacitors were polarised. Fortunately I used a "pretty" aesthetic of having them all aligned in the same direction - and got the first one right. I was lucky that the other three that were positioned at right angles also were the right way round.
Back in the late 70s, I was a young apprentice.
The company I worked for also imported cheap Taiwanese bench grinders.
One of these had been returned under warranty, so was on the department manager's bench with the base off to check the wiring. When nothing obvious presented itself the switch was turned on and we retreated to the office doorway. These grinders used a large capacitor and a centrifugal switch to energize the motor start winding with the resulting L/C circuit providing a phase shift to ensure the motor started in the right direction. The grinder started up and seemed to be running normally, until with a loud bang, the capacitor blew its lid, (through the small covered pressure relief hole in the end of the cap) blasting a jet of evil smelling vapor directly at all standing in the doorway.
We at least now knew what the fault was. The centrifugal switch "didn't", leaving the start winding connected. The start winding and capacitor weren't designed to run for more than about 10 seconds, so the capacitor overheated, boiling the electrolyte until the weak point cried uncle.
The office became almost uninhabitable for days, and took many months before the smell finally faded to barely perceptible levels.
I still work for the same company some 40 odd years later.
An old boy in the 1960s was the chief electrical engineer for a large company. He used to test for mains voltage by seeing if he could feel it through the hairs on the back of his finger.
In those days some radios and TV's were often AC/DC. That had once been essential when some areas still had local 240vdc mains supply. Even after 240vac was standardised it was economic to omit a large transformer in favour of a large resistor to drop a large part of the voltage for the valve filaments.They only had two wires in the mains cable - and sometimes it ended in a two pin mains plug. It was not unusual to feel the AC on the wooden case if line was connected to the chassis.
As you’d expect of such a student, Alvin “was prototyping a circuit I had put together on a bread board which was connected to a PC clone used by one of my Professors.”
And I thought by the 1990's we'd already stopped using the phrase "PC clone" ... as if IBM still led the way at that point.
Many many moons ago, when pc's had cluncky on off switches and techs still went out on the piss the previous night, i managed to electrocute myself so well it blew me off the seat.
A desktop i was assembling had power going to its front switch, the was in the days of 240v not wimpy 3v or whatnot to a soft button. They had little boots on the connectors, now normally i made flipping sure these covered anything "bad" but one day, slightly inebriated from a session the night before one of them was not doing its job properly.
As i grabbed the sides of the box to move it forward i touched the contact, as it was plugged in at that point there was a bit of a bang, i flew off the stool and all the lights (and everything else) went out.
I managed to the not only the workshop but the whole buildings breakers, once i managed to get up and stagger about a bit i felt well proud of my accomplishment.
Not long after one of the boys found out that the CRT screen he was attempting to dissipate was still powered producing a much louder bang and sparks accompanied by another lights out event - again i wonder if the previous nights alcohol might have been a reason for this foopar.
Nowadays i get a pfy to check things that might produce sparks and bangs, they are easier to replace.
In the early 1990s I was playing with MOSFETs in an attempt to build an extremely loud audio amplifier on the cheap. I built 4 Class AB amps on a +/- 45V power supply that could manage 8A momentarily. The circuit design was crude and involved twisted hookup wires between the MOSFETs and the main breadboard. The final build suffered from a bit of distortion and my old oscilloscope showed bits of the waveform missing, which was a typical symptom of intermittent oscillation. I soldered some picofarad caps on the MOSFETs as an easy fix. Now it was weird. Every time I turned on the amp, the MOSFETs hissed, my radio went dead, my CD player spun BACKWARDS, and my oscilloscope went totally blank. I disconnected the oscilloscope probes and it still wouldn't work just being there near the amp. The amplifier felt oddly hot where it shouldn't.
Moved the oscilloscope and powered on the amp. The scope a drew what looked like spring viewed from an angle, which means that the trace was partially going backwards too. I measured the loops and got 170 MHz on a 'scope rated for 20MHz while its probe was not connected. It turns out that the pins on a TO-220 MOSFET make a fine RF transformer. You're supposed to put series resistors on the pins to stop oscillations; never capacitors. I was energizing all the speaker wires in my room with 170 MHz of high voltage AC. The amp felt strangely warm because it was RF heating my fingers when I touched it.
I eventually got it working and it was worth every bit of effort.
When I was a teenager back in the mid 70s, i accidentally blew up a medium sized elctolytic capacitor in my bedroom.
I had a home made bench PSU capable of supplying upto 25Volts at around 5Amps.
I accidentally connected the capacitor the wrong way round. It was fun to see the needles on the Voltmeter and Ammeter slowly rising until they both shot up to FSD, then BANG. lots of flying fluff and ears ringing.
Did that stop me? No. The rest of the bag of about 25 caps, were dangled out of my bedroom window and hanging on Crock Clips. Great fun was had. Wow, the echo around the neighbourhood.
It did discolour the wall just below my window though.
Ah, happy days!!
In the 1960s I had a bright idea that a pair of SCRs with capacitors could be used as a cross-coupled oscillator in the 12-0-12 volt side of a power transformer. Unfortunately while the theory was probably viable - I guessed at the size of capacitors. Connected to a car battery it failed to oscillate and the insulation on the wires started to smoke and melt.
Before becoming an engineer I worked as an electronics technician bringing up boards on ASIC test systems.
I had a new board on a riser card. On first powerup, I flipped the switch, I heard this hiss start as it rapidly rose in pitch ... then a loud snap. A puff of smoke then a streak of heat past my eye. The body of the capacity struck the wall behind me and feel to the floor where it sizzled and spun around on the floor for a few seconds. It burned a nice brown circle into the tile.
I have grabbed 110 volt wires before. But, nothing scared me quite like the viciousness of an exploding tantalum capacitor.
Last year I bought some small voltage up-converter modules off ebay - up to about 30v in and up to 35v out. They looked much the same as the buck down-converters I had used in several designs - a small can capacitor at each end and the transformer gubbins in the middle. Voltage adjustment was by a multi-turn pot.
So - I set one up in the same way as I did the down-converters. Hooked up 12vdc to the input and put a multimeter on the output with no load . Being lazy I didn't connect the meter permanently - assuming I could twiddle the pot and stop to take readings occasionally.
Quite what went wrong I don't know. While fiddling with the leads to get the meter to show a reading there was a sudden bang and the output capacitor's can shot into the air - ricocheting off my finger. The paper innards were left behind connected to the board. It was fortunate that the can had not hit my face.
My presumption was that the module was capable of producing more than the stated 35 volt maximum output. Both capacitors were marked as "35v". Looking on ebay I found that other suppliers' apparently identical modules were populated with a "50v" output capacitor. Sent the seller a picture of the damage and pointed out the other sellers' difference - immediate full refund.
Back in 1981 or so when I was learning RPG, TI-Basic, and COBOL on a TI-990 minis one of our fellow students forgot a semi-colon in a program he wrote. This cause two problems. First, the lack of a colon caused the program to go into an endless loop. Problem two was the code was sending the results to the printer. Unfortunately, the grad assistant was out on lunch. One full super-case of greenbar went thru the printer, which almost fried itself.
The ICL2903 was an interesting computer to program at system level. You were programming with microcode to handle its built-in devices.
One of these devices was a card reader - which had a powerful solenoid that was designed to be activated in short bursts.
It was reputed that one of the development sequences gained the name "select solenoid, stop, and catch fire". Now when I build anything with a solenoid rated for intermittent use - I always include an independent safety device to cut the current after a set delay.
A tractor-feed line printer with a duty-cycle not damn near 100%?
I'd have asked for me money back. Our old 1900 was sending to the printer any time there was enough memory to load XKZE.
On multi-part paper (which meant crankin' 'ammer force up to 11).
Up 'ill.
Buth wez.
In winta.
Back in the early 1970's, wandered one slow evening into the main computer room to chew the cud with the operators on duty. I looked for a place to put down the 3 cm thick fanfold printout I was carrying and plopped it onto the top of the IBM mainframe, which had the unfortunate result of making a shaky circuit breaker inside the CPU trip and the mainframe then shut down.
No real harm done. Open a panel, flip the CB upright, close panel, power on, re-IPL the system and restart the running jobs.
Weekend working on our mainframe computer in the 1960s was a "catch-up" exercise to clear any backlog from the week.
The last job was a long one and finished with lots of real time printing. By mid-afternoon the end was in sight - and the shift leader started to power down the other peripherals. Electrical noise from the papertape reader caused the system to crash. That meant running the whole job again.
I can't remember how many different unexpected mishaps there were that day on that job. On the final successful run - late in the evening - everyone stayed outside the room until the printing had finished.
Long, long ago, before the dawn of most things digital for mere mortals, I was an undergraduate at a liberal arts college run Cambridge, Massachussetts that was and remains far too full of itself. Let's call it the Eastern Massachusetts School of Mines. Downriver, there was a more technically oriented university, whose graduates liked to remain among the fleshpots of Boston and Cambridge after graduation, and in the largely pre-digital age what better way to exercise their entrepreneurial muscles than to build and sell high-powered stereo equipment. A freshman roommate purchased a pair of ~ 50 lb. speakers from one (imagine the size of the permanent magnets), and an amplifier from another, a brand-new startup. (I believe the amp had serial number 00001.) One afternoon, as my roommate was enjoying a fine, recreational pharmaceutical and entertaining most of the freshman class through our open windows with his massive (45 W per channel, a big deal in those dark days) system, a transistor in the power supply blew.
And when I say, "blew," I also mean "blue," which was the color of the perfect, vertical jet of flame, immediately followed by a miniature mushroom cloud exiting one of the vent louvers on the top of the steel enclosure. I didn't know about the persistence of vortices in those days, but I was fairly certain that nuclear fission was not involved.
Fortunately, the young entrepreneurs were deeply apologetic and gave my roommate a replacement unit (S/N 00007 or so), which survived several more EMSM serenades.
You should have seen what has happened when a Schlumberger S9K was brought up after a California Rolling Blackout in 2005. The TIU (Loadboard) couldn't be released because of the design of the interlock. The maintenance manager (contrary to his own SOP) threw the switch. Now the VCC was only like 1.15v even in those days, but the power capability was 5kw which is 5k AMPS! ... well the power supplies had a habit of coming on full power from a dead wakeup, but the relay would have been in the open position only in a safe power down which we didn't have. Max voltage for the SLB HV PS is 12.5v for the SLB digital HV PS. Needless to say when 5kW hits 50 or so under rated caps sparks fly.
This was another one of those days when I yelled the F-I-R-E word at work ... and meant it.
Im not sure if this is 'entertaining', but it did contain a 'Who, me?' moment. I said 'Yes'
The excellent University of Sheffield designed us an antimatter production rod, a target for the conversion of incoming pulses of GeV protons, into high energy pbars. The antimatter fell out of the back of the target, in a Pratchetian magic way, as statistically it was able to. I daren't use the word quantum, as most sentences which use the 'Q' word are wrong. We of course also got a lot of electrons, positrons, neutrons and positive and negative pions, kaons and muons, which was lovely, but not wanted in the pbar phase space.
The pbar target then was basically a solid rod of copper (size of a pencil), using magnetic focussing to self-contain the GeVp to allow maximum matter/energy/matter interaction magic, and a bit of cooling.
This focussing current was around 320000A pulsed(*), at some hundreds of volts. It worked well, low p/-p efficiency , but we were able to make and store more antimatter than had existed since the big one, quite a while ago. (*other Lithium targets take 1 MA pulses, @ ~6Hz)
One day, something didn't go very well, I never found out what, maybe a UPS tripped?, the Norsk Data computerised interlock system did manage to dump the incoming 26GeV pulses and switch everything off, but still the room and all the activated engineering inside it burned. Some heroes, as they do, ran towards it & put the 'nuke fire' out.
I was 'volunteered' into the second response team, as were the other twenty engineers on the un-named project, we were lined up by age - those childless were moved to the lower bias, and then we went in for an annual rad dose in my case I was painting everything with super sticky paint for around 6 minutes to get my dose. Senior colleagues took much more. Our WTF alarms were screaming all the time. We stabilised the errant activated BeO etc dust/soot. Think Wii-U Splatoon, we were the glow-in-the-dark squids.
Science started again after about a month. I salute Eifionydd who led the team and made the show go on.
Ah, that brings back memories of the odd high voltage exploding electrolytic PSU capacitor which would bomb out of its case with a large explosion and leave a rail of aluminium and paper as it followed a random trajectory putting life and limb at risk. There's probably a definite potential for weapon development there...
I remember visiting a PSU factory in the early 80s and looking around their test/rework area. The ceiling was unusually interesting - the usual fibre tiles above the operator’s bench were peppered with hundreds of aluminium cans (from exploding electrolytic capacitors) lodged firmly in them. This was a well-known phenomenon in the early days, which is why modern electrolytics have an "X" pattern formed in the top of the can to allow them to rupture and gases escape without any ballistic effect.
I can also personally attest to "cratered chip" syndrome. I experienced this when applying power incorrectly to some 74LS TTL logic when I was a teenager. Likewise a black piece of hard epoxy became ballistic in the process and a smouldering and glowing piece of silicon instantly became visible beneath!
Nick
Years ago when I worked on avionics we had a shiny new box delivered to the lab which needed to be tested before it was installed on an aircraft. It was bespoke and very expensive and so shiny that it attracted the attention of a project manager. He decided to start it up while we were at lunch and got busy sorting out power and data connections. We arrived back just as he switched on the power. There was a really loud bang and the room filled with magic smoke. The PM ran past us looking like the coyote trying to catch the road runner.
We were puzzled by the big bang, then someone checked the laboratory PSU. Voltage correctly set to 115V, but oh dear frequency set to 40Hz not 400Hz. Someone didn't understand the use of the decade switch. The manufacturer refused to replace it under warranty.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
