Your stupidest mistakes

[obrien]

Hi,

 Can you please take a while and laugh at my stupidity? Thanks  Of course I realized it, when it was already being make.




 Feel free to submit your mistakes so I do not feel alone 

[obrien]

BTW I did a search to find a similar topic but I could not find it. But I am new here, it if exists please point me there

[filssavi]

I don't know if it's my stupidest but it's pretty up there...

A couple of years ago I was working to diagnose a glitch in the electrical system of a formula SAE car that randomly turned off the car that turned out it was a faulty relay connector (the relay itself was fine), since it was late at night after a day of work/study I was quite tired and didn't realize that I had connected the car backwards, goes without saying that a control unit I made the previous year (actually the first board I designed) didn't have reverse polarity protection 

fortunately I had a Ti DC/DC step down (don't remember exactly which one) that I wasn't using that year, that sacrified itself (the soic-8 package had nicely exploded) failing short circuit and protecting the usefull things on that rail

[Asim]

Ouch, my stupidest mistake would be connecting 7805 regultor & lm317 to the same heat sink with no insulation.
So basically shorting the tabs together ( lm317 tab is vout while 7805 tab is ground )

[dekra54]

My stupidest mistake was to mix up 2 Strings from a PV system in a connection box on the roof. It was all fine when i connected the first inverter in the basement. But when i got to the second connected and starting things started to go wrong. The 2. inverter showed an overvoltage warning but unfortunately the voltage was so high due to my wiring mistake that in the few second it took me to shut down the inverter a few hundreds of milifarad 500V caps inside the inverter started to explode. I measured it after i got it all sorted out and the inverter was getting 1650V DC because the first inverter was connecting 2 strings in series 

[gnavigator1007]

All of my stupidest mistakes involve late nights, too much caffeine, and forgetting to unplug something before digging around. Makes all of my other mistakes seem quite reasonable.

[Howardlong]

Joining EEVBlog and finding I now had 20 oscilloscopes when I'd lived perfectly happily with one or two for a couple of decades.

[ZeTeX]

All of my stupidest mistakes involve late nights

Can agree a lot, every-time something that should work, doesn't work, and its late night I, try to debug it for a few hours then just say to myself "I will fix it in 5 minutes when I will wake up" and in 100% of the cases I was right.

[kcbanner]

Always review your board design the next morning instead of ordering them at 2am.

[GreyWoolfe]

About 4 years ago someone gave me a couple of old Tek scopes.  I plugged them in and turned them on without the benefit of a variac or dim bulb tester ( I have both now).  The first scope, which didn't have a cover on it, powered up fine and I saw traces.  Plugged the second one in and nothing happened.  I saw the modules in both scopes were the same so I thought that I would test the modules from the second scope in the first scope.  The first module worked.  After powering the scope with the second module, I heard a loud pop and a tongue of flame came out of the scope.  Naturally, the smell lingered in the house for a couple of days and SWMBO was not a happy camper.  She made me put the scopes out on the curb with a stern warning not to bring them back in.  As she has half the money and all the vagina, I listened.

[slurry]

Spent three hours fault finding on a headphone amplifier i built,  the audio on one channel was extremely weak.
I traced the audio all the way to the last output transistor pair where the audio went missing,
finally i replace the transistors with no luck 
I was tearing my hair until.... i realized i had connected the headphone jack with the missing channel to ground pin and the ground pin to the missing channel 

[Benta]

My worst mistake ever...

throwing away all my old databooks (Motorola, Texas, National, Linear, Harris, you name it), because with CDs and the Interwebs they'll no longer be needed.

Boy, was I mistaken. I'm still crying.

[David Hess]

I once swapped the vias where the output capacitor was routing power and ground from one side of a board to the other so power and ground were reversed on one side.  At least since the mistake was in one spot, just cutting the traces and adding a pair of jumpers was enough to fix it.

[Vgkid]

The one I laugh about:
Diagnosing why my cars battery kept dying. Put the battery on the charger overnight. Had the battery tested, it was good. Put the meter in series with the battery and the  (+) connector. I blew that fuse. CRAP, grabbed the meter, made up a quick test to see if the 10A fuse was good, it was.  Went back at it, the car drew a few amps. The diode bridge had a few open diodes(I wish they actually stocked those). With the meter still bolted to the terminal set the meter down on the battery. A few seconds later POP.  The other probe managed to hit the opposing terminal,instantly destroyed. Still haven't replaced the fuses.
The one I cry about:
I got a broken genrad 1689, had to sneak it out in a duffel bag* interesting story about that. 
Got it home, gave me every error message except bad memory. It was really late, pop the covers off, thinking what is the worst that can happen. Start checking voltages, they are good. Unplug the power connector. They were smart with, you can flip the connector around and it will still work.  Plugged it in,put the covers loosly on(I was going to resume in the morning).  Next morning
 Plug in the power cord, press the power button. Lights flash, fan runs, then the display goes black. Take the covers off, realize that when i plugged in the ribbon cable from the psu. I had shifted it by one pin. One rail was left floating, one connected to ground, the +/- rails were inverted.
Sigh...

[dexters_lab]

like me putting in a PAL20V8 IC backwards? that was pretty stupid

[Highvoltagefox]

Replaced the cord on a clothes dryer and didn't know my fingers were touching the contacts when plugging in. Woke up on the other side of the room. Good times..

[rs20]

Probably stupidest mistake was touching my oscilloscope ground to the wrong tab on the output of my LED strip controller; which created some complex fault loop including my PC's 5V USB supply, isolated supply's 24V output, and oscilloscope earth. Had to replace most of the ICs on the board, but at least I did manage to fix it. The thing that makes it worse is that I actually only wanted to measure DC voltage, but couldn't find the multimeter or leads; so if I had been careful enough with the oscilloscope ground, OR organized enough to have the multimeter to hand, nothing bad would have happened.

In a similar kind of story, I was once heating some copper hydroxide in a beaker to convert it to copper oxide to make it better for disposal. I wanted to loosely cover the top to keep the heat in, without creating a seal. But I couldn't find my watch glass (a small piece of concave glass that is designed for exactly this purpose), so I stupidly thought it'd be alright to just loosely place a normal rubber stopper on the beaker, without pushing it down. A few minutes later, I hear a slight "whistling" sound, see some gas working its way past the rubber stopper in one particular place. To my brain's credit, I realised I should sprint out of the room immediately. The stopper had somehow worked its way down to form a seal under gravity, but now the seal was failing. About a second later, the stopper pops off, and the pressure drop causes a huge boilover, with scalding copper hydroxide spraying in every direction, including on the wall exactly behind where I was standing a few second earlier. Luckily I was out of the room by then. As if that wasn't stupid enough, I later discovered that the watch glass (that I couldn't find) was in front of me the whole time.

And for my "tip of the day": I designed a PCB late last year, but held off submitting it because PCB shop was closed for holidays. So I decided to go ahead and write the firmware for the MCU on board, and in the process of working through that, discovered a couple of mistakes in the PCB (e.g. LEDs connected from open-drain pins down to ground.)  I might make a point of doing firmware in advance more often in future, as it's much easier to fix these issues in the digital world!

[M4trix]

Joining EEVBlog and finding I now had 20 oscilloscopes when I'd lived perfectly happily with one or two for a couple of decades.

I'm still resisting ! I have one analog oscilloscope and since I'm only fiddling with audio amplifiers, I need it for observing sine waves
and the clipping level. Digital ones still can't beat the analog in this field. 

[retrolefty]

In the early 80s, while I had read that all unused input pins on a simple CMOS logic chip should be pulled up or down I failed to do it and touched the chip on a solderless breadboard. Damn blister took months to heal, but lesson stuck. 

[retrolefty]

Joining EEVBlog and finding I now had 20 oscilloscopes when I'd lived perfectly happily with one or two for a couple of decades.

I'm still resisting ! I have one analog oscilloscope and since I'm only fiddling with audio amplifiers, I need it for observing sine waves
and the clipping level. Digital ones still can't beat the analog in this field. 

 Many here will disagree with you, but I too have been happy and well served with my vintage Tek 2213. 

[Alex Eisenhut]

Have made many. My first tech job I was using OrCad SDT, this was 20 years ago. I made my first schematic and PCB, and when I powered up the board it worked. As I was working on it I noticed the current consumption crawl up slowly (nice old meter movement). I was too excited to work on my first board to really pay attention to the simple power supply.
Eventually the current consumption went way up and the tantalum caps on the board caught fire.
Turns out the polarized capacitor symbol had swapped pin numbers compared to the footprint. I didn't check any of that, I assumed it was all good.
Since then I always check every symbol and footprint.

Other mistakes include taking the DC rating of a scope probe and not checking frequency derating. I was working on a high voltage (KV) AC power supply, I just grabbed the "HV" scope probe and slapped it on the test point and smoke started coming out the other end of the probe cable... that's where the compensation network lived...

[Mr.B]

My biggest goof in the last couple of years...

One off project for a family friend.
16 x 16 LED matrix, all SMD, controlled by shift registers. My friend was going to do the MCU side himself.
16 n Fets on the columns, 16 p Fets on the rows.

I got the p Fets upside down in my schematic.
All the rows were always on because of the body diode.

Unsoldered all p Fets.
Re-soldered all at a strange angle.
Added 16 little bodge wires.

When my friend saw it, he asked "whats going on here?".
My reply - "Bad day syndrome. It happens."

Lesson learned: Triple check your schematic and board layout before sending the gerbers to China. 

[eugenenine]

Hi,

 Can you please take a while and laugh at my stupidity? Thanks  Of course I realized it, when it was already being make.




 Feel free to submit your mistakes so I do not feel alone 

It depends on if your doing conventional current flow or electron flow

The old PIC microcontrollers that were UV erasable, if you connect them backwards they emit light

Thinned down my collection of books too much.  Gave away a bunch of parts I wish I'd kept.  Got rid of a bunch of computer stuff I wish I still had.

[David Hess]

I'm still resisting ! I have one analog oscilloscope and since I'm only fiddling with audio amplifiers, I need it for observing sine waves
and the clipping level. Digital ones still can't beat the analog in this field. 

 Many here will disagree with you, but I too have been happy and well served with my vintage Tek 2213. 

You guys are so far behind the times; move into the digital age; my go to oscilloscope is a 2232.  You should always have at least two oscilloscopes so you can use one to repair the other; two of several different oscilloscopes is even better.

Mistakes to go along with swapping power and ground on half of a printed circuit board layout that I mentioned earlier:

1. Designed and built a 1000 volt DC stroboscope power supply with way too much output capacitance and no safety enclosure.  Shorted it between my index finger and elbow but at least my whole arm only hurt about a week after getting thrown off of the ladder, across the garage, and onto the concrete.

2. Picked up a powered 25kV neon sign transformer by the insulators, but at least I set it down instead of dropping or throwing it.  A neon indicator wired to the primary would have been handy.

3. Managed to create a ground loop through the first DSO I used with those neat and tiny micrograbber ground clips; at least the ground clip only turned white hot and melted.

4. Was not paying enough attention and fumbled a glass bottle of concentrated nitric acid trying to set it back on the shelf while wearing shorts in the chemistry lab pouring it all down my leg.  The skin on my leg turned grey instead of the usual yellow from nitric acid before pealing off and the hair only took a couple years to grow back.

[JacquesBBB]

Maybe this thread would be better placed in the General Chat section.