Poor or Unreadable text on Some Electronic Components

[jonovid]

just an observation when looking under the microscope, maybe poor or unreadable printed text is done on purpose.   
you would think with laser engraving this would not happen.
laser engraving microscopic font size is possible even if space is limited on the component.

[Monkeh]

Then why would they bother marking at all? Why spend money to achieve your conspiracy theory when you can do it for free?

[thm_w]

Why make this thread without any photos?

[SiliconWizard]

Proper lighting (intensity and angle) is key to be able to see the small markings on components with a microscope.

[Siwastaja]

Then why would they bother marking at all? Why spend money to achieve your conspiracy theory when you can do it for free?

And quite indeed, there are components without any markings at all; many small SMD parts that could accommodate 3-4 letters that would be easily visible under microscope, but they do not.

Threads like this pop up frequently. The answer to the all related questions:

• Why unclear markings?
• Why abbreviated markings instead of full part numbers?
• Why no markings at all?

is the same: the audience which uses the parts, professional electronics manufacturing (and to lesser degree, prototyping and repair (where approved schematics and BOM is available)), do not need these markings. If they are there, nice, maybe they add a little bit of value. If not, then fine too, we do well without. We buy parts in supplier packages with printing on it; we pick/place carefully not to mix parts. Added value from markings is detection of some pick/place accidents (by manual or automatic optical inspection), but clearly such accidents are not common enough so that engineers would demand better markings from the manufacturers. For example, I have nothing to complain about the current situation. If I can get markings for free, fine; but if I could replace markings with a coffee and cookie, I would take the latter and enjoy my work more that way.

[tom66]

Chalk can help reveal engraved markings on parts which aren't otherwise visible.

[nctnico]

If I can get markings for free, fine; but if I could replace markings with a coffee and cookie, I would take the latter and enjoy my work more that way.

Until somebody loads the wrong reel in a pick & place machine and starts producing a couple of thousand boards with the wrong part. AOI doesn't pick it up because the part looks OK. You'll be royally screwed if something like that happens. IOW: markings are important. Even if they are some obscure logo and product code.

One of my recent designs has tiny semiconductor parts which can be placed correct both ways around. The marking on these components are typically printed twice but mirrored. My assumption is that this is done to aid AOI.

[Siwastaja]

If I can get markings for free, fine; but if I could replace markings with a coffee and cookie, I would take the latter and enjoy my work more that way.

Until somebody loads the wrong reel in a pick & place machine and starts producing a couple of thousand boards with the wrong part. AOI doesn't pick it up because the part looks OK. You'll be royally screwed if something like that happens. IOW: markings are important. Even if they are some obscure logo and product code.

One of my recent designs has tiny semiconductor parts which can be placed correct both ways around. The marking on these components are typically printed twice but mirrored. My assumption is that this is done to aid AOI.

You are probably right that if all markings went away, we would have a bigger problem. Now if one IC from the BOM is unmarked and rest are not, AOI picks mixups correctly. But once you have two unmarked, then there is theoretical mixup risk but it's still small. If you had 50 unmarked ICs on BOM, all unmarked, the risk is indeed real.

On the other hand, with the continuing trend of miniaturization, nearly all SMD resistors and capacitors are unmarked anyway. I mean, even if 99% of the ICs are marked, still < 50% of the whole BOM is marked. And actually a wrong resistor or capacitor can be worse than wrong IC; wrong IC usually leads to obvious failure in test, while swapping capacitors/resistors could cause weird signal integrity or EM compliance issues, and very rarely we go so far in our test fixtures to e.g. measure rise time of on-board I2C bus, or go to EMC lab with random samples from every batch produced.

And yet somehow we survive with unmarked resistors and capacitors; somehow contract manufacturers are really good in not mixing the reels up. Actually I'm quite surprised how they manage to do that because to err is human.