I was wondering, since multilayer ceramic caps have layers like this:
Aren't they supposed to (for better performance) be soldered on the PCB as the above picture shows aka parallel to the PCB layer, for example like this:
and NOT vertically like this:
(Which I think will also make the layers radiate around the PCB when their power is noisy and has ripple.)
If "Yes," is there a way to tell which side I will be soldering when I solder with soldering iron/oven?
The original question was about the MLCC orientation on the PCB and whether better performance could be obtained by mounting the capacitor in the vertical orientation.
This characteristc has been known of since the early days of microwave engineering of the 1970s. Mounting the capacitor in the vertical manner would reduce the occurrence of the parasitic self resonances low down in the rf spectrum. The resonances then only become apparent at much higher frequencies. So the capacitors behave as capacitors and do NOT become inductive till the microwave regions.
The relevant question then is - does this help or is it relevant to the design in question? The answer depends on your application. If you are dealing with very high speed switching designs with very fast rise and fall times and need to exploit the non inductive effects of the capacitor (i.e. minimum L in parasitics) then this is something that would be of good use.
The research has been done by American Technical Ceramics, Johanson, Dilectric Labs, KVX, Waycom and a few others.
Here are a few application notes/papers.
https://www.johansontechnology.com/understanding-chip-capacitors
http://www.electrical-integrity.com/Quietpower_files/QuietPower-30.pdf
Parallel vs Vertical Electrode Mount
https://www.mouser.com/datasheet/2/218/DLI_catalogue_10021_17-1510962.pdf
https://exxelia.com/uploads/PDF/ceramic-capacitors-catalog-new-62949dc1491f1.pdf
In terms of circuit simulation, Modelithics allows the user to select and exercise this analysis feature.
You would need to work out if this type of capacitor mounting would be of benefit to your design.
This defect would be called "tombstone" , when a part mounted on its side or it stands up. Generally it is an indicator of quality of manufacture, but on certain applications the orientation may matter electrically or have poor contact.
MLCC generally mounted automatically from a tape and reel in the orientation that is most solderable. A suction nozzle grabs the part and puts it down on the board, the tape and reel technology is designed to have the part in the correct orientation for the automatic equipment.
Occasionally the component is rotated within the tape and reel , a packaging error, this may result in a poorly mounted part and is caught during inspection
For hand soldering, it is generally pretty easy to tell when side is "down" on the capacitor pancake , but with small parts like 0402 or 0603 it is common to tombstone the part when hand soldering for novices. Usually this is only a mechanical concern, but if you believe that there is an electrical need it only reinforces the interpretation of this as a defect.
Very occasionally the part is "square " (usually high density small sized caps), generally the manufacturer does not indicate the preferred axis and it would be considered over constraint for you to prefer an axis, likely for this kind of sensitive application there are specialty caps with asymmetry to have a fixed orientation. (Like a three/four lead feedthrough capacitor)
I found something, just the thing I was suspicious about.
The video doesn't mention anything else but Id like to see such datasheet with a plot about orientation of placement vs L change.
Of course, this is about inductors, the question was about capacitors.
