Capacitors from/to ground (in 1207 package ...)

[oliv3r]

Hey all,

I ran into something confusing I can't quite figure out.

After finding this section in a schematic where a single capacitor R224 and R225 is only connected to ground, with both its legs. These parts surround U200, the magnetic sensor, as can be seen in the foto , where we see a crystal oscillator top left, then top-mid the sensor (U200) and top right R225 and R224. With the missing capacitor (why is it even labeled R?) one can also very clearly see it's the ground traces being connected.

This photo is from iFixit, where R225 seems to have been broken off. On my board, R224 is broken off. The picture does show it was broken. Not sure if this was during repair at some point, or they do this in the factory.

But first, I need to understand what the purpose is/could be. What's even more weird, the part is specced as 1207. Measuring it seems to be somewhat around 1.3 mm * 0.72 mm. The width seems to be spot on, where the length can be explained by the tin on the sides? Anyway, I measured R225, as I lack R224. I ordered a cap of the right value as a 1608, so slightly bigger, but does it matter?

[TheMG]

They could be dummy components for board version/revision identification, and in this case they probably used capacitors as they likely already had them used elsewhere on the BOM so that's what they put in there.

Quite common to see this done so at the factory they can visually identify the board's firmware or hardware version amongst otherwise visually similar or identical boards, by looking at which of the dummy components are populated or not.

I don't see it serving any other functional purpose, the pads are obviously shorted together on the PCB.

[oliv3r]

Are we certain this is not somehow used with the magnetic sensor?

I can agree with the resistor trick, but they are using ton of 0603's even zero ohm resistors, are those nut much cheaper then those oddly sized caps?

See here for the hardware identification pins used on this board for example. So why not tie 'blanks' in there as well if it is for identification purposes?

[NorthernWing]

That's... weird, and it wouldn't surprise me if it's a mistake made by a drafter that laid this board out who didn't really get the purpose of those capacitors either.  I've seen:

• Dummy parts used for revision ID like everyone else has noted
• Some old Analog Devices ADCs where the reference design calls for a guard poured around the analog half coupled to the analog GND with a cap under one specific pin

- neither one of which really makes sense here because of the context of the schematic and the position of those caps.  They wouldn't have tried placing them like they're decoupling caps if they were just for board ID, I'd think.

Maybe they had a purpose at some point, but I'd say the EE just handed this schematic to an inexperienced drafter, didn't explain what those caps were actually intended for, and that's where they ended up.  That's just my guess.

EDIT: The more I look at this thing, the more I hate it lol.  Why does it look like none of those stitching vias are aligned to a grid? Why are they using an 'R' refdes with a capacitance value populated with a cap?  Are those the decoupling caps for what looks like an oscillator hanging on like 1/2" of trace?  Are the 'positive' terminals on the rightmost two of those capacitors even connected to anything on the top copper? What is an AXICS?

[gamalot]

Those 1000nF/10000nF capacitors drive me crazy! 

[oliv3r]

Yeah, I found this schematic also very poor; but this level of engineering is what drives most of our day to day cheap electronics. It's _very_ sad indeed.

Haven't figured it out still though, and the sensor 'works' as it is supposed too ... so *shrug*