Why do SMT capacitors have no markings?

[lazarusr]

N00b question:

Why do 0603 MLC capacitors have no markings? If the strips become separated from their packaging/reel there is no easy way to tell what value they are. There doesn't seem to be any difficulty putting markings on 0603 resistors.

[tooki]

I've wondered this, too!

[T3sl4co1l]

Because it's the default option?

Lots of things in life are that way, simply because they are that way.  There needn't be a reason, let alone a good one, why something is the way it is.

Marked caps do exist, I've seen them before!  They're very rare though.

Equally so, one might ask -- why bother marking resistors at all?

Some 0603s and most 0402s are unmarked, but the larger ones usually are marked.

Haven't seen marked ferrite beads or other multilayer inductors.  Those should be equally plausible as capacitors are, but probably just aren't as common, so, even if they're marked as often, I've probably just not seen one yet.  (Wound parts usually are marked.)

Tim

[lazarusr]

Actually, the more I think about it, the real question is probably why bother marking resistors? Plainly, component manufacturers are more interested in industrial clients who will use multiple reels at a time and aren’t really interested in what the components look like once on the board and certainly aren’t interested in retaining tiny strips of components.

[Ian.M]

Serviceability.   Resistors commonly fail by changing value, and their values tend to affect the DC working point of the circuit, so the repair tech needs to be able to see their values at a glance.  Capacitors, not so much.

[wraper]

If you look how MLCC are produced, marking them individually would be pain in the ass not adding any additional value. Simply printing over bare ceramic once they are already produced won't work well.

[wraper]

On the other hand resistors are printed over protective layer which you need anyway and is done before separation from a panel.

https://youtu.be/wshRwO0MCSU

[Neilm]

Not so long ago there was a motion to remove markings from resistors. It would save money not having another step in the production.

[BravoV]

Imo, it comes down to a really basic fact and truth, that no human read it anymore.

Especially in these throw away buy new age.

PS: Human = Big customers / companies , not one off project for hobbyist.

[Zero999]

Imo, it comes down to a really basic fact and truth, that no human read it anymore.

Especially in these throw away buy new age.

PS: Human = Big customers / companies , not one off project for hobbyist.

If no human reads it, then why bother marking resistors?

I suppose the cost of marking resistors is so minimal it doesn't matter, but as mentioned above, it's more costly to mark ceramic SMT capacitors.

[T3sl4co1l]

Ah, but humans -- and far more importantly, machines as well -- can read it.  Labeled parts may be an important part of assembly inspection.  That would be a point in favor of marked caps as well, but there is a cost breakdown between avoiding or optimizing rework through inspection (which mind, can be at PnP time, or AOI), and part cost, and there are typically fewer caps than resistors on a board making the advantage even thinner.

Another plausible explanation, but probably still not a wholly motivating one.

Note that this would only apply at scale -- even in the thousands, the few percent of rework can be done by hand.  This would be in the 100k's where machine vision systems, as fully automated as possible, are justified.

Tim

[RES]

https://en.wikipedia.org/wiki/Ceramic_capacitor#Marking

[wraper]

If you look how MLCC are produced, marking them individually would be pain in the ass not adding any additional value. Simply printing over bare ceramic once they are already produced won't work well.

Vishay safety rated caps are printed. Surely it can be done at a mass production scale.

Then look at their size and price. Not to say marking safety caps makes sense.

[tooki]

If you look how MLCC are produced, marking them individually would be pain in the ass not adding any additional value. Simply printing over bare ceramic once they are already produced won't work well.

Why would it not work? You can inkjet print onto essentially any substrate.

[radiolistener]

Printing may leads to change it's value

[wraper]

Why would it not work? You can inkjet print onto essentially any substrate.

Printing on ceramic a way that it will survive reflow and washing? Unless it involves additional heat treatment stage, I don't think such marking will last.

[T3sl4co1l]

Epoxy or enamel ink would be the way to go.

I wonder also if it's complicated by manufacture variation: are they able to produce values accurately at low reject rate, or is there selection done?  If the latter, marking would have to be done as a final step, which sucks.

Handling also matters.  Caps may be handled loose, in the production or shipping process.  Which side is "up", isn't really obvious after that.  Non-square parts should be easy enough to sort through the handling process, but square parts, dunno.  If this is a key part of the process, it may simply not be worth keeping parts neatly taped and everything.

Whereas resistors are fabbed on ceramic plate, which certainly can't be handled or mounted sideways (at least, not nearly as easily), and they can only be laser trimmed from the same orientation.

Tim

[wraper]

Epoxy or enamel ink would be the way to go.

Which won't fly on already made individual components 30 seconds away from testing and packing. You cannot apply marking before final process stages, but by then it becomes pain in the ass just as I said before. Storing them for curing in particular orientation as they will go into a reel is not trivial.

[Ian.M]

They could be printed with a fireable ink (metal oxide pigment + frit) in between tumbling for corner rounding, and the final firing to sinter the terminations.   However that doesn't address Wraper's point about orientation, or T3sl4co1l's point about select on test for possible tolerance binning.  The only remaining practical option that addresses those issues would be  lazer marking immediately before application of the cover tape.  However the industry has had 30 years to get used to unmarked SMD ceramic caps, so bulk customers who would be willing to pay extra for marked caps are IMHO vanishingly rare.

[tooki]

Why would it not work? You can inkjet print onto essentially any substrate.

Printing on ceramic a way that it will survive reflow and washing? Unless it involves additional heat treatment stage, I don't think such marking will last.

They make solvent-resistant inks. Industrial inkjet is NOT the same as what we have at home! This is what is already used on some Bourns-style trimmers, for example. Remember, pretty much any reasonably thin liquid can be used as ink in a piezo-type inkjet head. This means there are a LOT of ink possibilities. (Like, you know how they make solvent-resistant permanent markers? This is no different!)

Another option is UV-cure lacquer inkjet. I would be surprised if this isn’t already used on electronic components, too. This could easily be done with the caps in their wells in the tape, just before the sealing film is applied. Literally just needs about a foot of space along the line. Bright UV light cures the lacquer literally instantly.

[Bud]

Not so long ago there was a motion to remove markings from resistors. It would save money not having another step in the production.

All of 0603 resistors I bought lately from Digikey came unmarked.

[T3sl4co1l]

Like I said, epoxy (heat or UV cure) or enamel (fired ceramic) would be entirely plausible, if all the other conditions for feasibility are met of course.

All of 0603 resistors I bought lately from Digikey came unmarked.

Panasonic?  I have a few Panasonic that are unmarked (just black enamel).

Digikey sells a dozen brands of chip resistors, that's not very specific!

Tim

[schmitt trigger]

Money.

MLCCs are produced and sold by the Gazillions. Even if you reduce a milliDollar or a milliEuro from its manufacturing costs, it will be serious money by the end of the day.

Also as other poster noted, big customers  will always machine-place them, and as long as the reel has a barcode or RFID tag that a machine interface can read, it is sufficient identification.
And because electronics nowadays are essentially intended to be non-repairable, not by the end users at least, it does not matter it can't be identified once that it is soldered to the board.

I sometimes visit a PWBA-assembly house. They have many tape tails, containing a few components at most, which they simply discard. They have allowed me to take some of those tails, and back home with the help of a capacitor tester equipped with tweezers, I have been able to sort and identify those tails. Sufficient to say that I haven't required to purchase any 0.01uF, 0.1uF or 1uF MLCCs in several years.