The unwritten laws and rules of Electronics

Join the club.

The nearest is "boondocking", but that implies a deliberate act and/or the wink is either still on the mat or can be found on the floor.

Even if you found such an errant critter, how would you know it was the same critter that had just escaped? Or that it hadn't been infected by ESD?

I have now been thoroughly corrected. Thanks for the explanation.

Once it's lost, it's out of the game. I just pop a fresh one off the strip and try again.

On Friday, I chased yet another apparent parasitic oscillation on my circuit. After some microscopic inspection, the problem turned out to be an airlaunched resistor that landed on the board and partially shorted my output. Amazingly, it made a good connection but didn't blow anything out.

Right. Either of the other options would have created less uncertainty.

The effect is real. I think it's because I usually press on a small part with my finger. It sticks, so then I lift it up and release it maybe 6" off the deck. It tends to maintain its orientation.

Picking it up with tweezers, and dropping it, is similar. One has to deliberately invert it and carefully put it down to get a reliable flip. Or maybe drop it from some great height, which has its own issues.

I got a vacuum tweezer thing but never use it. It's not useful for flipping parts.

Den mandag den 4. december 2017 kl. 19.41.44 UTC+1 skrev John Larkin:

put the parts on a piece of paper and tap the paper

That's interesting; maybe some calibrated tap. Enough to flip but not launch them into orbit.

Because the leads are offset far from the centre of mass, chips are much more likely to land upside down. If you picture the solid angle (i.e. area on the unit sphere) of orientations in which the part will tip over the right way vs the wrong way, it's pretty small, actually--probably only 1 or 2 steradians vs 4 pi for the whole sphere.

Cheers

Phil Hobbs

(Thinking mostly of SOT23s and SC70s, which I've been dealing with a lot lately.)

PH

Den mandag den 4. december 2017 kl. 20.08.23 UTC+1 skrev John Larkin:

I have more trouble with resistors. Caps, being unmarked, don't matter how they are oriented.

Things like SOT23's are easy to flip over.

Well, some of the ICs, especially, have a tendency to land upside down. The leads lift the body off the surface when "right side up", so they prefer to be at a lower energy state, with the leads up in the air.

Passives also may have bumps on the top metallization, but are made to be quite planar on the bottom, so they may do the same, but to a lesser extent.

Jon

When you slam your knees together quickly to catch something tiny rolling off the bench, it will turn out to be a ball of molten solder and you'll be wearing shorts.

I once caught a soldering iron in mid-air. Only once.

When you have 50% chance of guessing something correctly, you have 90% chance of getting it wrong.

Communication signals are always labelled "Rx" and "Tx" with no indication if this is from the viewpoint of the board, or the thing connecting to the board.

The Universal Serial Bus isn't very.

If it smells like chicken, you're holding it wrongly.

John Larkin lives in a different universe.

I've solved that problem, for good, by using a consistent naming convention for all signals of the form: owner_function_type. A serial bus would have a name like UC_Debug_Async, with nets UC_Debug_RxD and UC_Debug_TxD. Since the bus is "Owned" by the UC, the directions are with respect to the microcontroller.