What's the weirdest fault you ever fixed?

[star882]

Originally posted by woody
I had a dead computer that wouldn't turn on because the
hard drive had a bad power supply bypass cap. The computer
power supply was detecting this short on the power bus and
refused to power on. I connected the hard drive to an older
power supply that didn't detect the bad cap it did hovever blow
the bad cap up in a small puff of smoke! Put the hard drive back
in the computer and it worked fine without that bypass cap!

The newer PC power supplies, due to their size, also will not shut down on overload unless it is a really severe overload. I remember reading about an unfused neon lamp ballast (used in a case mod) shorting out but not tripping the overload protection. The ballast simply burned up spectacularly. (Given the small size of the wires and the large power supply, the fault current was probably well below the current rating of the supply!)

[unclejed613]

most frustrating (non-)repair, rebuilding an Acoustic Controls model 370 bass amp, 3 times, replacing every piece of silicon all 3 times, replacing questionable caps also the second time, and replacing all the silicon, all the electrolytic caps, and a few resistors that had drifted the 3rd time. carefully brought it up on a variac all 3 times. at 60V the ammeter would slam, both fuses in the variac as well as the line fuse in the amp would instantaneously vaporize. never figured it out, ran out of stock of output transistors and the amp owner said "never mind". years later i was working at a music store with another tech that's been repairing guitar amps as long as i've been working in electronics (i've had a bit more varied experience than just guitar amps) he asked me once what the worst repair i'd ever had was and i began describing what happened with that amp, but before i could tell him what type of amp it was he says "it was an Acoustic 370, wasn't it?".... i was surprised, and asked "how did you know?, he said he's tried many times to fix blown up 370's and got nothing but smoke and fire, even with a variac.

most interesting repair: would have to be when i was the amp tech at Apt Corp, and amps began coming off the assembly line that would fail a test where we would fee a 100khz input into the amp with no load and watch the variac ammeter to make sure it didn't go above 3A. this was to make sure the amp would survive the user mistake of hitting the FF/CUE button on a cassette deck with the volume turned up . the amps had never been known to fail this test, but it was there for a reason. on a particular production run, the amps began failing.... a lot. before the first day of the production run was over i had 10 amps stacked up in the failure rack. the boss halted production until i could figure out what the problem was. as a control group i grabbed the last 10 amps to pass final test, and began taking apart a good one and a bad one side by side, and started inspecting both to see if there were misplaced ires, wron cable routing, etc.... i didn't see anything obvious, so i began running the test on both amps and looking at the output waveform at 100khz (i used a lower signal level that didn't blow the amp's line fuse). the good amp, while the sine wave wasn't perfect, was about what you'd expect to see out of a decent amp. the output waveform on the bad amp looked horrible. since the problem was mainly one of high current at 100khz, i put a current probe on the emitter resistors. the good amp had a slight current spike at crossover, but the current spikes on tha bad amp were much larger and wider. so something was keeping the output transistors turned on too long. i had access to some of the engineering notes for the amp, so i began looking through them, and i found exactly the same waveform i had seen from the current probe. the notes had different waveforms for differing beta of the output devices, so the company had paid Motorola to pre-select our transistors. the published beta range for the transistors was 15-75. at betas over40 or 45, the miller effect increased the base capacitance to the point where the drivers could not turn the transistor off fast enough to keep up with a 100khz waveform. so the outputs would go into common mode conduction and dump huge amounts of current directly across the rails. so i began looking at the output devices on both amps. all the ones on the good amp had APT stamped on them by motorola. about 1/2 of the transistors on the bad amp didn't. so i went to the parts room and began asking why the change. our new parts supervisor said "oh, i found we could get the transistors from a distributor for 1/3 the price that we got them from Motorola". i told him what i had found, and he said "we can save money by doing the selection process ourselves then" so i grabbed a flat of new transistors and sat down with a beta tester. about 1/3 of the transistors passed spec. we ended up doing our own selection from then on. but it didn't save any money, since once the 1/3 of the transistors was selected the parts cost was the same, but we had to pay someone wages to do the selecting.

[amc184]

Not a particularly weird repair, but a stupid one:

I was putting a Marantz CD52 back together late at night, and when screwing the mainboard to the chassis I made sure screws went in every hole.

Including the holes for adjusting the laser current and focus offset pots from the underside of the board.

One pot was just misadjusted, the other was wrecked. Lucky I had spares. At least it didn't take me long to figure out what I had done.

[rdf]

Not my solution but my early career was spent at a Music of Your Life AM station serving Detroit. It's a music format catering to the elderly, which twenty-plus years ago meant Benny Goodman, Nat King Cole, etc..

Late one morning a listener called in driven to distraction by a constant, repetitive beeping on our station. Naturally first thing is to run for the off-air monitors and listen. No beeps. Crank it up, wait for commercials and cold voice, use headphones, still no beeps. She's gets adamant when I hear nothing with her phone to the radio. The disc jockey on air and receptionist report no complaints. The elderly can have trouble with hearing and localization so I ask if she's absolutely sure it's coming from the radio. Yes. Understandably she's growing frustrated and impatient with the incompetence between her and her music. In an act of desperation/cleverness, working feverishly through logical possibilities with all the tact of young engineer I risk the question:

"Does anyone else hear it?"
"Of course others can , my daughter hears it too!"
...voice in the background....
"Mom, I told you, IT'S YOUR CLOCK RADIO!!!"

The other line goes 'click'. Problem solved.

[unclejed613]

i have seen similar ones to this on the internet, but i took this call in the summer of 2000.....
i was wotking for a company that repaired banking equipment. and as an FYI point a Proof Machine is a desk-shaped machines that a bank pre-processes checks with. the operator reads the amount from the check, typed it in and drops the check on the left side of the machine. the check goes across the machine and gets the bank endorsement stamp stamped on the back, and the amount printed on the MICR line of the check along with a transaction code, and the document goes to a stacking pocket and the machine records the transaction.

so as the bench tech at this company, one of my duties was to talk customers through basic troubleshooting to determine whether they actually needed a field tech. i get a call from a large check processing center of a local bank one hot summer afternoon from a proof machine operator:

op: my proof machine isn't working....
me: ok, what model is it?
op: it's an NCR 7766.
me: ok, do you have any LED's lit on the front panel?
op: no
me: do you hear any motors running inside?
op:no
me: is the power switch turned on (the power switch was under the front lip of the machine, and easily bumped by a knee.)?
op: yes it is, and if i turn it off and on again, the machine doesn't beep like it's supposed to.
me: ok, well you're in a processing center, and it's easy for somebody trip over the power cable and unplug it, is the machine plugged in?
op: i don't know, i can't see the outlet box, even if i stand and look over the top of the machine.
me: why not it should be right there between your machine and the machine next to you.
op: i know, but the lights went out.

[DCPreamp]

Back in my days working at Linear Power, while doing some customer repairs, I came across an interesting one. An amp came in from New York with one channel out. I fired it up - sure enough, one side out. I popped off the cover and noticed part of the PCB and bar (the interface bar that the TO-3 transistors mounted to that then bolted to the heatsink) were bent. I pulled the amp out of the sink and then spotted the problem: A bullet was puncturing the heatsink and deforming the PCB and transistor bar. Once the amp was no longer shorted by the bullet, it worked fine. Apparently the car had been a shootout and a bullet fired into the trunk hit the amp.

We kept the sink in the shop jokingly referring to it as the "bullet-proof amp. . ." We also reported it to a car stereo magazine who printed the story in a little "news of the weird" kind of blurb saying that Linear amps were indeed bullet-proof.

Another time, we had a customer call with a troubleshooting problem that had baffled them for days. When they mounted the amp to the speaker box, the amp wouldn't turn on. Off the box, the amp worked fine and several amps did the exact same thing when mounted. Grounding? Vibration? We did some serious head scratching over that one. Turns out the mounting of the amp on the sub box was in just the correct place and orientation for the speaker magnet's field to cancel out the magnetic field in the speaker relay opening up the contacts. They repositioned the amp on the box and it worked fine. Problem solved.

[Glowbug]

Cockroach in a dual Eimac 3500-Z RF deck.

They might be the only survivors of a nuclear war, but they don't survive 2500V of B+ voltage.

And now I've learned to check everything I buy at a hamfest before turning it on...

[cbcassell]

Back when I was more youthful than I am now, I was given the job of equalizing remotely located sound systems for the first "Who" farewell tour that was broadcast to 12 theaters from the Chicago Amphitheater. The audio and video signals were microwaved to each location.

All went well until I arrived at the Uptown Theater on the north side. It soon became apparent that the stereo four way PA system brought in by a subcontractor no longer had any functional low mid speakers. Needless to say with one hour to go before the broadcast and no spare speakers available, I was somewhat upset.

After recriminations about how poorly they had prepped their PA, we worked out our options. Luckily, there were 5 RCA theater speakers behind the projection screen. We moved 2 to each side and plugged them into the low mid amp channels, balanced the xover outputs and adjusted the system equalization by ear during the performance as there were over 1500 people in attendance. The result was very acceptable and no one noticed the substitution.

Needless to say, I only made it to 9 of the theaters.

Chris

[unclejed613]

i just remembered this one while i was answering another thread, and figure it would work well here too.......

i once worked at a stereo shop in NC. we had bought a case of dual RCA cables from Sony. they were some of Sony's low capacitance phono cables, and thought they might be useful, since low capacitance cables were required for the then somewhat-popular Quad encoded records, which had a 38khz (i think) subcarrier encoded on the disk, and low cap cables were required to present as little loss to the subcarrier info as possible. we got these cables at a bargain basement price, and as time went by, we found out why. it seemed that the right channel cables had a tendency to get noisy and eventually short. we couldn't figure out why only the right channel cables were doing this. it didn't make sense. so i decided to find out where the cables were shorting, and maybe i could then figure out why. so i took a cable that was shorted, and as with most intermittent or bad cables it had an area just above the connector where the cable felt limp. so i cut it open right there, and a bunch of red powder poured out., then i cut open the left side to find out why it hadn't failed. in these cable sets Sony used a plastic foam insulation on the inner wires. the white (no colorant added) insulation was still strong and pliable, the red coloring in the insulation on the right had caused a chemical reaction that caused the plastic foam to crumble into red dust. if you flexed the cable enough you could make the whole inner insulation of the cable crumble and pull out the inner bare copper wire... since sony had sold the cables as a "discontinued item" we couldn't return them for credit, and now had a box of single channel cables with matching shorting plugs.

[Enzo]

"Matching shorting plpugs." I like that...