Strangest electronics repairs

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
That senior chap had the right idea. Sometimes you just need to lay aside whatever project is frustrating you for a while and work on something else. This is a luxury usually not available except to us DIY types, as bosses want things done to schedule (or before).

Yup. Sometimes you just get too close to see something simple. Walking away and getting a cup of coffee (or tea) helps shift your perspective.
Doc
 
My supervisor had great confidence in me after my first two digital design projects and assigned me, a physicist, to design a 10 mhz data logging memory address calculation board. This was to analyze data coming in on the NASCOM network. (Internet was not yet ) It was 1976 and Schottky 74S logic was new and hot. I calculated and drew out time charts. I settled in an AMD bit slice 2901 ALU set to calculate the next address and figure when to circulate back to the beginning of the buffer, etc. It was built up on wirewrap board. When I got the prototype, I hooked up the HP logic analyzer to the output to check my "microprogram" was running through the addresses as planned. I was getting doubles of some instructions, kind of randomly occuring. Not the expected sequence. I postulated the clock line was ringing, and terminated it with suitable ohm resistors as the line driver application book suggested. Same result. I changed the clock line driver IC to something else. Same result. I doubled up the number of clock driver IC's and halved the length of the clock wire run, with terminating resistors. Same result. I had the entire clock line pulled up and re-wrapped with twisted pair, with one end of each dead half grounded. Same result.
I pulled the government audited calibration stickers off the logic analyzer pod (horrors) and opened the pod up. I intended to look at the clock in there with a scope, but the ground wires were burned in two inside. The whole thing had been an artifact of defective test equipment. eight days of checkout down, three weeks to go. The system eventually tested out at 13 mhz, when my calculations had said it would go to 16. Darn! Met spec, though.
 
Last edited:
This one is a little tame compared to some of the stories here. I am not a repair tech, just a DIY nut and an EE by trade (but that itself means nothing; most of my EE classmates and coworkers would not know which end of the soldering iron to hold).

My automatic garage door opener would open the door but would not close it. It would start down, then bounce back up and the lights would flash. This usually happens when something interrupts the optical sensor across the bottom of the door. I checked the connections to no avail, then replaced all of the solid (!?) 18 ga zip cord wiring with new stuff, also without success. Then I took the optical modules down to my workbench and pried them open. Tested the electrolytics and one was in fact bad so I replaced the lot. Re-installed the modules and it seemed to work. After a few days, the problem returned. Opened up the modules again. Inspected very carefully ... eventually I found the problem. There was a missing solder joint on one leg of an SMT transistor connected to the output of the module. Not a cracked or cold joint; no solder joint whatsover. The thing had worked for years on nothing but incidental contact between PCB and component. Eventually, vibration and/or corrosion made that connection go bad. One touch of the soldering iron later and it has worked for years since.

A friend gave me an older surround receiver that was going into protect mode. He said that it had worked intermittently but was now going into protect immediately every time. I opened it up and saw that several grounds on the PCB (and RCAs) connected to the common ground (i.e. the chassis) only via mounting screws. So I backed off every screw I could see about a 1/4 turn, then re-tightened them. Fixed.
 
This one is a little tame compared to some of the stories here. I am not a repair tech, just a DIY nut and an EE by trade (but that itself means nothing; most of my EE classmates and coworkers would not know which end of the soldering iron to hold).

I started this thread and so far every story has been most welcome, and enjoyed. A long time friend and I have remarked many times about conversations we had that not only our fathers, but danged few people on earth would have a clue of understanding about. Just knowing that there are others on earth who know some of the same things is a comfort at times.
That applies to Audio, electronics in general, physics and even computer programming.

One of the joys of working in Silicon Valley in the eighties was the sheer diversity of technical people that you encountered. At the local bar I played foosball with a doctor of chemistry. One of the guys worked at the hypersonic wind tunnel at Moffet field. A contingent of the Norwegian Air Force would drop in (on assignment to Moffet, got a scar on my chin from the Screaming Nazis drink that they introduced me to.... involved a ten speed and a dog...afterward.) One night I was talking to someone about lasers and remarked the only ones I didn't as yet comprehend was Ring Dye... lo and behold he's one of the worlds leading authorities about Ring Dye lasers and proceeds to explain them to me on a bunch of bar napkins. One of the treasured mementos of the valley was a distributor cup that had a Db / SWR chart printed on the side. Like the sands of time these are the days of our lives.....


Moving on with stories.....

Doc
 
Off topic I guess. But I once spent a very (VERY) drunk evening (thanks to an unplanned weather event at JFK) with two fellow travelers (I met in the line at the AA desk) - one was a Norwegian fish magnate (on his way to Alaska) and a guided-missile salesman (on his way home from the Paris Airshow). Those stories were toe-curling.
 
The Wang Incident

Back in the day we had a RF transmitter of sorts, where the resonator had to sit a very high vacuum due to the large amounts of power we were pumping.

Once in a great while the vacuum would burp and things would go very bad, very quickly, lots of melted copper and glass Jennings caps, scorched wave guides. The residual gas annie would always show a clear water signature.

Only glitch is we couldn't find nary a water leak with 100's of hours of He leak checking.

Fast forward 20 years into the Past. One night in that distant past a grad student (Mr. Wang) was manning the controls. Automatic controls shut everything off due to vacuum issues.

Mr. Wang sat by the hour forcing things to run..... till they wouldn't. Next morning the day crew came in and heard the story.

Looking into the RF tank through video or sight glass showed none of the tell tail signs.... just clean copper, SS and glass. Maybe a little too much sparkle!

Opening inspection hatches let out ohhhh 600 - 1000 gal of polished water.... the RF tank was full to the top, ditto the vacuum pumps, roughing pumps, chases, catch basins.

Everything came apart got cleaned, dried and put back together, 3 months of double shifts...... everything but one capscrew.... one hidden out of the way capscrew.

Twenty years later that one capscrew was still burping water, from what we called a virtual leak... which is why it never showed in the normal leak testing.

Most of the time that water just sat nice and polite in the bottom of a threaded hole as ice. Institutional memory through change of site managers had all but forgotten Mr. Wang.

Cyclotronguy
 
Last edited:
Benc: Yes, airports used to be a fact of life for me. Had a tag line that read "If you are too good at your job they put you on airplanes". Got a message back from one guy who said he read it at 30,000 feet and stewardess was wondering why he busted out laughing.

Met a guy in the bar at the San Francisco airport from a south American country -which shall remain nameless- who got REAL INTERESTED when he learned I was working for a used test equipment place. He was actually a minister of technology procurement for x-country. Wanted me to sell him stuff on the export restricted list. I said we'd be happy to sell him anything it was LEGAL to sell him... conversation wained.


More stories:
Basic rules forgot.
Working in the late seventies as a job shopper in cal lab at Sperry Univac. Had these ATE (automatic test equipment) power supply testers that were pretty simple. A switch selected ROM output would turn on X transistors leading to load banks to verify that the PS boards were doing their job. A whole bunch of these things kept eating the switch transistors for seemingly no reason. The devices were well heat sinked and well within current and voltage specs. One day I walked up to this one engineer with a masters degree and told him he could cure the problem if he just put some back emf protection diodes on the switch transistors. Yup, he hadn't taken into account the load banks were big wire wound resistors. The back emf spike was punching the switch transistors. Diodes got put on each device and I don't think they lost another transistor. He started bringing me coffee in the morning in exchange for learning electronics.

Doc

Seems I was posting at same time as last message.
 
Last edited:
Cyclotronguy. Once got a private tour of SLAC. Pretty impressive. At the time they had one 15KW klystron ever fifty feet or so pumping more RF umph into the line (phased push to the accellerated particles.) They were upping that to 45KW every fifty feet. The guy giving the tour told a story about the beam taking an unexpected turn one day and started boring a hole into the adjacent mountain at 20 feet per second.

Doc
 
The story above
.......this one engineer with a masters degree.........
reminded me of an incident in the land of the pharoahs.

It was a requirement that the company hired one local for every expat employed.
The only problem was that the "graduates" only wanted to drive a desk and not do any field work.

One day, we had to make up some new lighting cables, -- plug, 20m twin flex, light socket -- nothing difficult.

So we set the local guy the task and went for our lunch (it was Ramadan).

We came back 2hrs later and found him still sat at the workbench with not one complete cable.

In front of him was a pile of white insulation pieces and another pile of copper wire cuttings.

"This reel of wire is no good !"
"Why"
"It doesn't matter how much I cut it back, it's still rusty!"

Closer questioning revealed he had done no practical electronics in getting his degree and also had never seen flex with bare copper wire only tinned!

He didn't last much longer.


.
 
Last edited:
I once tasked an 'engineer' with a 2:1 in Electronics from a good university with building a simple PSU (toroidal, bridge, electrolytic) - gave him the contents of that mornings delivery from RS Components.

Came back later on to find him unwrapping the clear plastic insulation around the toroidal, it didn't occur to him as *odd* that the *packaging* was so difficult to unwind..
 
Last edited:
I once tasked an 'engineer' with a 2:1 in Electronics from a good university with building a simple PSU (toroidal, bridge, electrolytic) - gave him the contents of that mornings delivery from RS Components.

Came back later on to find him unwrapping the clear plastic insulation around the toroidal, it didn't occur to him as *odd* that the *packaging* was so difficult to unwind..

Around 1978 a friend of mine turned up with a Marshall valve guitar amp. He had been given it by a BBC sound engineer with a degree in electronics. The sound engineer thought it was a good idea to dip the whole chassis in some type of strong solvent, maybe trichloroethylene. This damaged the transformers, possibly stripping some of the varnish off the windings, leaving them short circuit. I got it working by using some parts from another Marshall amp that he had also given to my friend, he could not understand that a diy person such as myself with no formal qualifications had managed to build up a working amp from his failed experiments.
 
The most horrid thing I have ever repaired was a lighting controller for a fairground ride.

It was designed to switch 6 channels at 115V at over 200A per channel.

At 200A Triacs become funny things, involving latching curremts (the things don't turn OFF when the gate voltage is removed).

This thing had 100W light bulbs in it which were shorted across the Triacs to get them to shut down when the drive voltage was removed.

Anyway, the drive electronics had been toasted.

Simple solution was to provide an ASCI interface. All the guys needed to do was to write a simple program on their Spectrum (yes a few years ago) that output the pattern of lights that they wanted.
 
Oh, I'm so happy. My Hifi is fixed! :)

It's been sounding AWFUL for ages. Like loose speaker wires or something. Today I had the 10-y-o Rotel RA-931 30 Wpc amp apart, cleaned it and checked the output stage bias, which was 4mV as expected. A 63mV offset on both channels at the speakers looked worth investigating though...

While probing the 37V rails, I took my eye of what I was doing to look at the multimeter and (we've all done it, no?) FLASH! BANG! I shorted the two rails with the probe. Bit worrying! Power down and check the supply fuses which were strangely stubborn to make any contact at all with the ohmmeter. So I pulled out a speaker fuse to compare. Seems they were all OK.

Put it back together, and it was working fine. Even sounding better...:confused:

The penny dropped. The speaker fuses were corroded and making bad contact. Pulled out the other one, wiped it and replaced it. My Hifi is sounding TERRIFIC! Either that or the bang frightened away the gremlins in the amp. :D
 
Accelerator

I had a water cooled beam stop loose cooling once. In a twinkling a 70 Mev at some significant current proton beam auger through the now un-cooled carbon block, punch through an aluminum gate valve and then do battle with a box of kem-wipes.

More amusing than anything.

What you need to know about the Stanford Linear Accelerator is that any hole it cut into the hillside was pretty tiny. Most time were looking at MW in and mW out.

Most of the energy goes to distracting electrons while you heard packets of protons along their merry way. They're always wanting to go wandering off and it takes heaps of energy to keep them "on the narrow path"

Cyclotronguy
 
I had a water cooled beam stop loose cooling once. In a twinkling a 70 Mev at some significant current proton beam auger through the now un-cooled carbon block, punch through an aluminum gate valve and then do battle with a box of kem-wipes.

More amusing than anything.

What you need to know about the Stanford Linear Accelerator is that any hole it cut into the hillside was pretty tiny. Most time were looking at MW in and mW out.

Most of the energy goes to distracting electrons while you heard packets of protons along their merry way. They're always wanting to go wandering off and it takes heaps of energy to keep them "on the narrow path"

Cyclotronguy
OMG, the world is a safer place while I'm not given 70 MeV Proton beams to play with. But cyclotronguy, you might have zapped Stanford String Theory Professor Leonard Susskind by accident! He's my HERO! :eek:
 
Them electrons is slippery little devils. Think that's why they invented wire nuts. To keep them from forming dangerous puddles under bare wires..... Or so I was told.

One of the things I remember about the SLAC tour (early '80s) was all the pallets of one foot square steel blocks stacked all along the 2 miles between ends. Marked with how hot they were. Guess they swap them out every few years.


Doc
 
Last edited:
Stanford and String Theory.

Well there be a rivalry between Stanford and Cal.. My stepson when he was very little hollered out of the car window (Stanford Sucks) before he understood that is was a friendly rivalry...... still I wished he hadn't done it in the middle of Palo Alto.;)

Most of that steel shielding at the SLAC was old WWII Ryerson Armor plate from the local navy stores. Not the best shielding material, but it was however free, back in the day. Free was good, cause you needed lots! We had a breech from an Atomic Annie cannon in about 50T of concrete / magnetite doing the same thing. Why? Because we could and it was free.

Cyclotronguy
 
In the midwest, our tools are less exotic. The government very astutely sites most of the expensive research equipment either in an earthquake zone, under a volcano, or on the coast in the hurricane-tornado belt.
I was the test equipment engineer during a changeover of the refrigerator factory test line, whose mission was to run the new units for some minutes and see if they got cold. The electrical engineer replaced the now unavailable GE bus bar with rolling tap boxes, with an available new plastic bus with much smaller copper bus bars. As the short circuit current of the new bars was much smaller than the old huge bars, he had split the run segment into two circuits each with a transformer with a smaller short circuit current. Melting the bus copper and plastic when a box wrecked and shorted across would be very annoying, causing maybe a 3 day production interruption. Because refrigerator compressors, if they see interruption of power, overload and stop running, the electrical engineer had put an isolated segment of bus between bus 1 segment and bus 2 segment, that would be connected to bus 1 by a crydom SS relay as the brush bridged bus 1 to isolator, then would shut off and allow a second SS relay to connect the isolated segment to bus 2 as the power box advanced.
The relays kept burning out, evidenced by all the refrigerators overloading at the isolator instead of running on. They fiddled and fiddled with the photocells, delays in the software program of the AC controller, and finally ordered and installed bigger rating solid state relays.
Friday afternoon before Monday production start, it did it again, one of the new 30 A 240 V relays burned up, and all the test refrigerators were hot and low watts at the exit test system.
I suggested the electrical engineer install 1 ohm 225 W ohmite type resistors in series with each relay to cut the surge current when the brush was bridging the isolator to run bus. On Friday afternoon, there is nothing like that available in Louisville, and Newark in Chicago was closing even if we had time to get a ship today order done with all the signatures.
I looked at the Belden catalog, and checked two 1000' reels of 16 ga THHN wire out of the tool crib. (Running parts supply for maintenance). They measured a little over 1 ohm each end to end. The electrical engineer agreed to try it, and had the electricians install the rolls of wire in series with the solid state relays. No more problems. I think the official 1 ohm logs came in on Wednesday, and were installed after three days of production.
 
Last edited:
I know this thread is about fixing things, but sometimes that can be a BAD thing to do. I was in the Bedford College Physics lab one afternoon running a graph plotter connected to a Mossbauer effect experiment, which is something to do with gamma rays and moving Iron nuclei and really quite amazingly precise.

Anyway, things were just Mossbauering along and I was a bit bored and spotted an ancient Van Der Graaf high voltage generator lying about. Replaced the perished rubber belt with some sellotape improvisation and had that cool thing sparking about 2" long and punching holes in paper, making my hair stand on end, and stuff for a happy hour. Like you do...:cool:

Then I noticed that every time it sparked, the Mossbauer effect plotter registered an earthquake of a spike, so gave it a rest. Seems I was making an electromagnetic pulse on a big scale.

On going for coffee later, the coffee bar had the entire College computer department sitting around. "What's going on?", I asked. "The computer has been crashing all afternoon, we've quit work for the day, it was hopeless !" they told me.

Happens the computer department was next floor up from the Physics Lab, er, I think it was me...:eek:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.