Most weird-assed Elect. problems I'll probably not see again ??

[GlennSprigg]

In a Factory once, I found a bank of Fluoros that would ONLY work, if another 'bank'
of lights was turned on too....   (All or nothing). Found the 'fault' in the main SwichBoard.
TWO lighting circuits, on different Phases, had their NEUTRALS twisted together, and in
the same TUNNEL (hole) in the Neutral bar. However, this 'termination' was LOOSE !!!
So it only worked when 415v was applied across BOTH 240v circuits in series, at once !!
I said.... "F^&$ING C$#% Sh@#$", but wrote something different in the report....   

[Brumby]

I have my entry.....

One day I was visiting a friend in his home and he showed me a really bizarre result of a number of light switches being operated in various combinations which produced different brightness in 3 different light fittings.  The combinations and symptoms aren't simple to describe (and it was a few years ago), so please excuse me for not detailing them.  After a few cycles of trying various combinations, I worked out that there was likely a problem with a particular connection - but the question was: Where, exactly, was this connection made?

Fortunately, in trying the more easily reachable light fittings, I found the problem on the second fitting I checked.  It turned out that one of the points on this light fitting had been used as a common connection for 3 wires - but these had badly corroded.  This seemed strange for a light fitting in an under-stair storage cupboard - until I noticed that my friend had stored the dry chlorine used for his pool.

Once cleaned up, all lights worked perfectly - and he relocated the chlorine.

[TerraHertz]

This hardly qualifies as 'weird', but it was yesterday, so...
I was recently given a big _old_ 240V to 120V conversion and isolation transformer. Someone had replaced the output socket, and anyway it's a good idea to check mains stuff. First test: ground continuity between input cord and output socket. OK. Then ground to case.... nope, open circuit.

Opened it up. Huh, ground point on one of the transformer compression bolts. Which are sometimes insulated from the transformer laminations, to prevent eddy currents. I visually checked, can't see any insulation at all on the bolt. It _should_ be electrically one with the L-brackets and the case. But it isn't.
There's no other original grounding point to be seen. So since 1953 this thing has been operating with the metalwork not grounded.

In the last pic the ground lead with the yellow lug is my addition. And yes, the paint is scraped off at the hole, and connection checked.

[duak]

In North America we have single phase for residential and 3 phase for commercial/industrial.  In Canada, 3 phase is most commonly available as either 208/120Y  or 600/347Y (delta/wye voltages).  A friend moved his cabinet shop from a building with 600/347Y to 208/120Y.  Some of his machines operated on 600/347Y and so the electrician connected some of my friend's step-down transformers as step-up transformers.

One transformer was quite hot even when unloaded and one day it blew its primary fuses - 100 A and north of $100 each.  I measured the line voltages and currents and saw that the current was a lot higher than it should have been.  On a hunch, I measured the neutral current at 50+ A when the transformer was unloaded.  I disconnected the neutral and the line currents dropped down to a few amps and the transformer cooled right down.

3-phase distribution transformers are usually delta primary & wye secondary.  It allows the secondaries to be re-referenced to neutral to prevent a circuit from floating.  The electrician, true to code, tied the neutral of the transformer secondary to system neutral, but that's a no-no in this case.  The delta-wye transformer configuration also eliminates negative sequence currents - this is where 3rd harmonic currents produce currents within the transformer that run opposite to the fundamental current.  It not only detracts from the power delivered to the load but it causes I^2R losses.  When reversed, the transformer doesn't block this current.  It seems too that any voltage imbalance will also manifest as a neutral current.

The electrician probably knew better but this is what his customer wanted so he did what he was paid to do.  I don't know the code well enough to say if he violated it but, in the end, you canna deny the laws of physics.  I suppose the wiring was safe, it just wasted power and fuses.

I measured the secondary voltages (600/347Y) to earth/ground and saw they were about 350 V.  I figure that if there's primary to secondary leakage, the maximum voltage won't be all that much higher than usual, ie. maybe 800 V instead of 600 V

Cheers,

[ajb]

First test: ground continuity between input cord and output socket. OK. Then ground to case.... nope, open circuit.

I little while ago I was asked to look at some old hand-me-down Lightronics dimmer packs and see if it was worth getting them working again.  These have four dimmed outputs, and either two input pigtails (giving up to 4800W per pack) or a single input inlet (giving up to 2400W per pack).  All the connectors are standard North American NEMA 5-15 plugs/receptacles, and any fixture plugged into an output is theoretically grounded through the pack and back to the input.  Usual repairs for these types of things are fuse holders and maybe an SCR here or there, no big deal.

Except.  There's no direct connection from the input ground to the output ground.  The inputs are mounted in an anodized panel, and the receptacles are mounted in a powdercoated panel.  So the ground pathway is theoretically output receptacle -> screw -> powdercoated panel -> screws -> anodized panel -> screw -> inlet.  So after 20+ years of being kicked around and sitting there oxidizing in the heat of a bunch of incandescent lamps the LOWEST resistance I measured from input ground to output ground was like 100 Ohms.

I contemplated doing a complete rewiring and fitting a new control board that supports DMX instead of the old Lightronics protocol no one uses anymore, but they're really not worth that much effort to fix these days.

[CopperCone]

cracked socket in a motherboard that made the computer work only when it was sitting flat. When I put it back upright in its desk hole it would stop working. Took me a while to figure out when I was in high school.

Eventually I got the bright idea to  power up the computer laying down and measure which angle it failed at (something like 60 degrees). It would fail slightly differently each time too, so I thought I was damaging it with ESD, thermals/interference from other cabling (in its cubby hole) or some other outlandish shit. 

[TimFox]

A good Hewlett-Packard 6209B 300 V lab supply (from eBay).  I unpacked it and connected to a DMM and found normal behavior.  However, when using it in a test setup, with one side grounded, it made very low voltage.
For some reason, -hp- had an old-fashioned rubber grommet between the circuit board and the metal support structure that contacted a trace on the board.  It had charred to charcoal, and effectively shorted that trace to the case.  When the supply was not grounded, it worked fine.
Replacing the grommet with a non-rubber plastic washer solved the problem.

[Red Squirrel]

I only heard the story and was not around to witness it, but we got a 611 trouble come in and the customer said every time their phone rang, the furnace started.  Baffled, and figuring it was some crazy person, we sent a tech.  Tech gets there and sure enough it's really what the customer says.  Turns out the phone line had somehow been connected to the thermostat wiring.   What I'm surprised is that it did not blow the furnace relay or even the whole board.  Those take 24 volts while phone ring voltage is like 170 volts.

[NiHaoMike]

I only heard the story and was not around to witness it, but we got a 611 trouble come in and the customer said every time their phone rang, the furnace started.  Baffled, and figuring it was some crazy person, we sent a tech.  Tech gets there and sure enough it's really what the customer says.  Turns out the phone line had somehow been connected to the thermostat wiring.   What I'm surprised is that it did not blow the furnace relay or even the whole board.  Those take 24 volts while phone ring voltage is like 170 volts.

There's also the case of the dog that would bark every time the phone ringed because it was chained to a bad ground. Might have just been an urban legend, however.

[Brumby]

A true story - not mine.

Someone had put together a computer rig in a tower case and whenever they tried to run it, it would glitch on far too frequent occasions.  Handed to my friend, he went about checking a few things out and found this problem went away completely with the side cover removed - and would reappear when replaced.

After some further testing and getting past the "It couldn't be" wall, his final assessment was that with this particular arrangement, there was some sort of RF tuned circuit that was feeding signals back into the motherboard and simply disrupting normal operation.  With the side panel off, this vanished.  If he wasn't into antenna design and construction, I think he would have been fighting with it for a lot longer.

The rig builder wasn't convinced and it required a couple of demonstrations to show that my friend wasn't bullshitting.

[Nominal Animal]

I once had a motherboard blow one of its CPU voltage regulators.  Not just burn out; it literally exploded, half of it blown clear off, with a clear audible bang, after almost a year of reliable operation (and a week or two of flakyness).

[Rerouter]

Context:
Most vehicles have a hall effect sensor to provide a speed signal for the vehicle, when it measures metal, it pulls its output low

In this weird case, I was asked to look into why a Buses radio stopped working the moment it went under 8kmph,

Turns out that particular model of speed sensor inverted its output, and some water had gotten inside and shorted across the inverter, so after about half a second of it not seeing metal, it would oscillate up and saturate, working as a wide-band FM jammer, broadcast by the 3+ meters of unsheilded wiring carrying the output signal.

The only reason I even though that might be the case is the workshop radio went out during one of our tests, when we moved it forward enough to align with a gap on the sensor gear.

[CJay]

After several visits to a site and workshop investigations on a PC with repeat claims of random characters appearing in a rather comely lady's word processor documents we figured out that moving her copy holder closer to her made the problem disappear as her ample chest wasn't hitting the keyboard when she turned the page.

[X]

Not sure if I could really class this as being "weird" but once I was testing a prototype high current LDO regulator circuit and it seemed to work OK the first time. Then the next time  I plugged it in, one of the two tantalum bead caps I was using almost transformed me into a pirate.

After clearing the solidified tantalum chunk that landed on my face, I realised the polarity was wrong on both caps. Corrected the polarity of the other cap, and replaced the Kentucky Fried Capacitor with a new one. This time, with both caps in the right polarity, I plugged it in yet again, and then the other capacitor tried to transform me into a pirate.

I didn't plan the localised festivities that occurred, but I had a fire extinguisher ready in case the battery or one of the connecting leads to catch fire due to over-current or an accidental short before the fuse.

[Berni]

We had a interesting situation during construction around the house.

A table saw with a 3 phase motor was brought in along with a smaller single phase chop saw. To get power to it a extension cord was brought along, we plugged it all in and started with the work, however we soon noticed that we have a problem as turning on the chop saw caused the ground fault breaker to trip. We thought the saw developed some short internally, started suspecting the extension cord too, yet it just would not run. Starting to run out of ideas we start unplugging all the cords and the problem went away. Then we discovered that whenever the 3 phase extension cord was plugged in the single phase chop saw would trip the ground fault. What does the 3 phase cord  have to do with it when the saw at fault had its own separate single phase cord going to a seperate mains outlet.

Eventually a neutral to earth short was discovered in that 3 phase cord. The reason this was never discovered before is that my grandparents who own that cord have a old school voltage based ground fault broker that does not care for a neural to earth short.

If you have not already figured it out here is an explanation why this happened. Both the single and 3 phase outlet was close to the electrical distribution box and on a branch that had no heavy loads on it. As a result the neutral wire the socket was on didn't have enough current to leak into earth to trip the GFI. Then the 3 phase saw does not need a neutral as it takes identical current from the 3 phases so the Delta/Star switch never bothers to connect neutral to anything even. This means no current in neutral so the saw runs without a problem. Then when you start the single phase saw the current flows from one phase into neutral and joins the neutral in the house, however this neutral had enough common copper in its path to cause a slight rise in voltage of the 3 phase outlets neutral terminal. This allowed the returning current from the single phase saw to flow into the 3 phase extension cord where it went to earth and back into the sockets earth. This made for not all live current returni back to the GFI trough its neural terminals and caused it to trip. Phew this could really give an inexperienced electrician a good run for there money trying to figure out the problem

[NiHaoMike]

After several visits to a site and workshop investigations on a PC with repeat claims of random characters appearing in a rather comely lady's word processor documents we figured out that moving her copy holder closer to her made the problem disappear as her ample chest wasn't hitting the keyboard when she turned the page.

Sounds just like a common legend. But I'm under the impression that area is very sensitive so it would be hard to miss the accidental contact? Closely related to that was a funny story a friend of mine told me about the time in computer repair class when she accidentally knocked over a box of screws...

[rsjsouza]

I was told this story from my Power Electronics teacher at the university.

He was working with other folks at a small remote power station that was experiencing an uncommon higher power consumption even under minimum load conditions. The situation went on for many months and. during an unscheduled disconnect for repair, they noticed that a large metallic plate (part of the structure) that surrounded the output isolators was terribly hot. This uncovered the mystery that went on undetected: the wires going out of the station were the primary and the metallic plate has become the secondary of a power transformer with its outputs shorted. Replacing the structure with a wooden one solved the problem.

Apart from the occasional ISA card that was being lightly disconnected with the motherboard angle, there are some other experiences of mine reported at a different thread, but none of them are particularly weird.

[CJay]

After several visits to a site and workshop investigations on a PC with repeat claims of random characters appearing in a rather comely lady's word processor documents we figured out that moving her copy holder closer to her made the problem disappear as her ample chest wasn't hitting the keyboard when she turned the page.

Sounds just like a common legend. But I'm under the impression that area is very sensitive so it would be hard to miss the accidental contact? Closely related to that was a funny story a friend of mine told me about the time in computer repair class when she accidentally knocked over a box of screws...

If it hadn't been me and a colleague who diagnosed the problem I'd say it was.

I also believe the area is sensitive but I don't think it's quite sensitive enough to be able to tell if it's hitting desk or keyboard, plus the scaffolding holding it all up and in must have been impressive.

It was rather embarrassing having to explain but to her credit she also turned bright red and later sent us a thank you card for our 'patience' which I *may* still have somewhere.

[jmelson]

Not sure if I could really class this as being "weird" but once I was testing a prototype high current LDO regulator circuit and it seemed to work OK the first time. Then the next time  I plugged it in, one of the two tantalum bead caps I was using almost transformed me into a pirate.

After clearing the solidified tantalum chunk that landed on my face, I realised the polarity was wrong on both caps. Corrected the polarity of the other cap, and replaced the Kentucky Fried Capacitor with a new one. This time, with both caps in the right polarity, I plugged it in yet again, and then the other capacitor tried to transform me into a pirate.

I didn't plan the localised festivities that occurred, but I had a fire extinguisher ready in case the battery or one of the connecting leads to catch fire due to over-current or an accidental short before the fuse.

OH, big rule violation here!  If a tantalum capacitor is reverse-polarized, even with just ONE Volt, even for a millisecond, ALWAYS replace it!  If connected to an ample power source, either immediately, or sometime soon, it will either burn or explode.  Even the momentary reverse polarization causes permanent damage.

Jon

[dmills]

Wirewrapped industrial control system back in the day, big card cage full of 4000 series cmos doing what you would do on a micro the size of a credit card today.
The thing was a mess of random logic, multiplexers, flipflops and counters doing state machine type industrial control.

The complaint? Every few days the thing would loose its state and a (separate) safety system would bring production to a crashing halt.

Much digging ensued, as the thing worked perfectly on my bench.

The fault?

Failed power supply connection, but as long as a sufficient proportion of the inputs were a solid logic 1 the thing was powering itself via the input protection diodes! Sweary, just a bit!

Regards, Dan.

[Gregg]

Three phase power but each phase routed through a different metallic conduit.  Things were getting warm but the client didn't want to spend the money or down time to fix this problem.   

[bsudbrink]

Three phase power but each phase routed through a different metallic conduit.  Things were getting warm but the client didn't want to spend the money or down time to fix this problem.   

Ok.  High power AC electrics aren't really my thing.  Would you mind explaining the problem with this?  All of the conduits are "bonded" at the panel, right?  Where's the trouble?

[dmills]

Magnetic fields and an effective shorted turn.... Lot of current flowing in the conduit walls.

Regards, Dan.

[BrianHG]

Here is a computer related one.

The Amiga SCA (computer virus).  This one caught me off guard when I was young.  Most likely the first Amiga Virus, 1987, if not, the first one to gain notoriety...

When rebooting, you expect the computer to reset, pause, cycle through a few shades or grey for around 5 seconds, then the insert boot Disk logo, or auto boot from the HD.  However, if you had the virus, a random coincidence once in around 20 boots, no 5 second reboot pause, instead, instantly, this colored bars appeared, then, the following eerily timed message:

Something wonderful has happened....
Your AMIGA is alive !!!



History: https://en.wikipedia.org/wiki/SCA_(computer_virus)

[bsudbrink]

Magnetic fields and an effective shorted turn.... Lot of current flowing in the conduit walls.

All right.  I think I see.  It's not really the conduits themselves, it the multiple holes in the cabinet?  You could effectively "fold" the holes together and end up with two interlocked rings, one with the AC flow and the other with the induced current, have I got that right?