I had a similar case this year, involving a 0603 SMD resistor which was purchased as 27 Ohms. THe label on the reel, and the marking on the top, both said 27 Ohms. But the resistors were within 1% of 270k, which made quite a difference in circuit performance.
Laughing at Mr88Cet's story . . . All 3 scenarios striking a familair chord! Not sure I ever got the EE team to hope for a puff-o-smoke, but we did see a few without trying before . . . and Yes, the painstaiking trial & error of checking chip by chip by chip brings to mind the vision of Horton the Elephant, looking for that dust-spec in the field of dandelions . . .
About 20 years ago, when I was working for a small telecom startup, we got a run of prototype boards back, and they all had a power-to-ground short. That sort of thing happens sometimes - a solder bridge between two connector pins, etc., but it's a bit odd for the entire run of boards to have such a short.
We engineers checked the artwork in detail, while the technicians studied the boards very carefully for any place where there could be a short. Everything looked right though. So we graduated to a few "classic but more desperate" techniques.
I think the first "desperate measure" we tried was ... not exactly a subtle technique, but quick: Fusing the short: We hooked up a sample, probably sacrificial, board to monster power supply to in hopes that we'll see a small puff of smoke erupt across a hairline trace that's causing the short. No help there: The whole board got warm, but nothing "went pop," so to speak.
Next, we hooked up a beefy, but low-voltage, power supply to it, and did a whole lot of power-to-ground voltage measurements, trying to find a geographic area on the board where the voltage is lowest. Conceptually, that spot should be closer to the short that a visual inspection could locate it. No such luck though: It was almost perfectly consistent throughout the entire board.
Perhaps there was a bad chip on the board? So we started removing chips from the board two or three at a time, until the entire top side of the board was cleared off. Still totally shorted.
So the technicians started clearing off the back side of the board, which mostly contained passives. Upon taking off one of the decoupling capacitors, he suddenly realized: "Wait, that 33pF decoupling cap ... that looks like a 33ohm resistor! Hey, so is that one!"
Then came the grand forehead slap: Every single 33pF decoupling capacitor - a couple-hundred on that board - was populated with a 33ohm resistor!
My last design used 0201 caps and resistors. And lots of them. No room for values marked on the parts, nor silkscreen on the PWB. I convinced the boss I needed a tweezer RLC meter (best tool we ever bought, you can meter one of these 0201 parts with one hand while flipping thru the CCA drawing with the other). Still need the microscope to check the other parts, though.
Frustrating to be sure, like chasing the wind; its hard to determine what you’re looking for while de-bugging a failing prototype circuit. Your Mantra is solid: "assume (& trust) nothing." -- At least the color banding offered you a small clue.
Compare that to T&R packaged discrete SMD’s which look like specs of pepper. Hard enough were the 20x40’s; then came the 10x20’s and now placing 5x10’s. For low volume proto runs (under 25) this type of component is often hand-placed by operators using tweezers. Opportunity for mis-placement is high, and component verifications are performed using a multi-meter with needle probes, under the scope. Painstaking but necessary. Is there a better way-?
In an age of globalization and rapid changes through scientific progress, two of our societies' (and economies') main concerns are to satisfy the needs and wishes of the individual and to save precious resources. Cloud computing caters to both of these.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.