Page 1 of 1

Pi 3b+ lost Capacitor on PCB

Posted: Mon Jan 14, 2019 9:46 pm
by zimzia
On my pcb is missing these 2 capacitors, so be it?
Image

Re: Pi 3b+ lost Capacitor on PCB

Posted: Tue Jan 15, 2019 11:05 am
by Burngate
The resolution is to low for me to see what's missing.
What effect are you seeing?

Re: Pi 3b+ lost Capacitor on PCB

Posted: Tue Jan 15, 2019 2:44 pm
by zimzia
I corrected the photo, it seems to me that everything works, can it be from GPIO?

Re: Pi 3b+ lost Capacitor on PCB

Posted: Tue Jan 15, 2019 6:40 pm
by mahjongg
Only the designer knows.
There are no component references on the board, and there is no schematic, the PI is NOT "open source hardware".
an educated guess is that these are decoupling caps, 100nF 6V3 0402, but as I said I'm guessing.
Were the capacitors rubbed off, or never mounted?

Re: Pi 3b+ lost Capacitor on PCB

Posted: Tue Jan 15, 2019 7:01 pm
by zimzia
I looked closely, missing 6 elements, never mounted, 5 capacitor 1 rezistor. PCB model FCC ID 2ABCB-RPI3BP, IC 20953-RPI3P, 3.3 OC-138, 94V-0H.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Tue Jan 15, 2019 9:28 pm
by hippy
zimzia wrote:
Tue Jan 15, 2019 7:01 pm
I looked closely, missing 6 elements, never mounted, 5 capacitor 1 rezistor.
Contact who you bought it from and arrange to get a replacement.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Tue Jan 15, 2019 10:47 pm
by HawaiianPi
zimzia wrote:
Tue Jan 15, 2019 7:01 pm
I looked closely, missing 6 elements, never mounted, 5 capacitor 1 rezistor. PCB model FCC ID 2ABCB-RPI3BP, IC 20953-RPI3P, 3.3 OC-138, 94V-0H.
It's normal to have empty component locations on a complex PCB. There are various reasons for that, but almost every large PCB will have at least a few empty spots (and those may change over time or production runs).

Not sure about the 2 you asked about in the OP (mine has them). Only the RPi designers or engineers can say for certain.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 10:09 am
by Burngate
The right-hand end of each appears to be ground; that would tally with mahjongg's guess
Where the left-hand ends go I have no idea. Not, apparently, to the GPIOs.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 10:52 am
by 6by9
Neither are populated on at least one of my boards.

I can check with the hardware guys, but I wouldn't worry about it. There will be deliberate minor variations between builds as tweaks are made to reduce the BOM, none of which will affect the functionality.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 11:49 am
by hippy
6by9 wrote:
Wed Jan 16, 2019 10:52 am
Neither are populated on at least one of my boards.

I can check with the hardware guys, but I wouldn't worry about it. There will be deliberate minor variations between builds as tweaks are made to reduce the BOM, none of which will affect the functionality.
I guess this takes us back to the "undocumented change" issue.

With two boards, where one has things missing which are on the other, no one outside RPT knows if that's some sort of assembly error or intended, whether it's something to worry over or not.

To determine which it is means having to find someone who can find someone within RPT who can find the definitive answer.

There has to be some better, less resource hungry, way than that

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 12:25 pm
by 6by9
6by9 wrote:
Wed Jan 16, 2019 10:52 am
I can check with the hardware guys, but I wouldn't worry about it. There will be deliberate minor variations between builds as tweaks are made to reduce the BOM, none of which will affect the functionality.
Checked. Yes, they're just small decoupling caps that were found to be unnecessary and therefore removed from the BOM. There are also at least four around the PMIC also removed.
hippy wrote:
Wed Jan 16, 2019 11:49 am
I guess this takes us back to the "undocumented change" issue.

With two boards, where one has things missing which are on the other, no one outside RPT knows if that's some sort of assembly error or intended, whether it's something to worry over or not.

To determine which it is means having to find someone who can find someone within RPT who can find the definitive answer.

There has to be some better, less resource hungry, way than that
You've bought a board for which is provided a guarantee that it will work. It's not an open design.
Unless you actually have issues (in which case take it up through the supply chain), don't worry about it.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 12:28 pm
by bensimmo
Do RPT alter the design or is it Farnell etc, the licensed makers, that choose to reduce and redesign these small bits and bobs as needed ?

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 12:32 pm
by 6by9
bensimmo wrote:
Wed Jan 16, 2019 12:28 pm
Do RPT alter the design or is it Farnell etc, the licensed makers, that choose to reduce and redesign these small bits and bobs as needed ?
RPT.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 5:56 pm
by Burngate
6by9 wrote:
Wed Jan 16, 2019 12:25 pm
... they're just small decoupling caps that were found to be unnecessary and therefore removed from the BOM. There are also at least four around the PMIC also removed.
I wonder how it's determined that they're unnecessary.

Does someone take a representative sample of boards - say, a thousand - and remove odd components here and there, just to see what happens?
This seems unlikely.

Or does someone read the fine print in the datasheets and realise they got it wrong first time round?
Similarly unlikely.

Is my idle wondering about this sort of thing a waste of my time?
That's very much more likely.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 6:09 pm
by davidcoton
Burngate wrote:
Wed Jan 16, 2019 5:56 pm
6by9 wrote:
Wed Jan 16, 2019 12:25 pm
... they're just small decoupling caps that were found to be unnecessary and therefore removed from the BOM. There are also at least four around the PMIC also removed.
I wonder how it's determined that they're unnecessary.

Does someone take a representative sample of boards - say, a thousand - and remove odd components here and there, just to see what happens?
This seems unlikely.

Or does someone read the fine print in the datasheets and realise they got it wrong first time round?
Similarly unlikely.
My guess is that the original design allows for different manufacturers or versions of one or more ICs. The capacitors are not fitted if the IC variant used doesn't need it, and obviously there is benefit (to RPT) in sourcing those versions (price otherwise being equal). It's quite possible that some early batches have to be built with the ICs that do require extra decoupling, until the supply chain becomes established and stable.
The design allows fallback to the "other" ICs without redesign costs -- just some reprogramming of the assembly robots.
Is my idle wondering about this sort of thing a waste of my time?
That's very much more likely.
The answer to that question is independent from the first two Q&As. :o :shock: :? :lol: :roll:

Re: Pi 3b+ lost Capacitor on PCB

Posted: Wed Jan 16, 2019 11:17 pm
by jdb
Burngate wrote:
Wed Jan 16, 2019 5:56 pm
6by9 wrote:
Wed Jan 16, 2019 12:25 pm
... they're just small decoupling caps that were found to be unnecessary and therefore removed from the BOM. There are also at least four around the PMIC also removed.
I wonder how it's determined that they're unnecessary.

Does someone take a representative sample of boards - say, a thousand - and remove odd components here and there, just to see what happens?
This seems unlikely.

Or does someone read the fine print in the datasheets and realise they got it wrong first time round?
Similarly unlikely.

Is my idle wondering about this sort of thing a waste of my time?
That's very much more likely.
The answer is: It's Complicated (tm).

Decoupling capacitors are always over-specified on reference designs - you sprinkle them around to excess, and forget about it. For volumes that the Pi is produced in, every penny counts, so we do a more in-depth design review (typically conducted with multiple board trials and PCB analysis tools). If we can omit a few 100nF capacitors from noncritical decoupling sites on the board, then we save money.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 17, 2019 10:21 am
by Burngate
Thankyou, jdb - I wasn't really expecting an answer, but that makes sense.

So the answers to my questions turn out to be:
Yes, sort of, in that it's done in a far more rational manner
Only partly, in that the "wrongness" was already built in to the datasheets, and the designer was already aware of it
No, because my idle wondering has led me to learn something.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 24, 2019 9:08 am
by Yukon Cornelius
I've had a few questions myself about components and layout .

But as another member pointed out , Pi is not " open-source hardware " ,
and although it can be a little annoying for an old timey electronics tech like myself ,
I think that it's a smart decision .

If nothing else , it makes it a bit harder for the would-be counterfeiters .
A vast amount of resources went into creating and developing the Pi
and AFAIK it's a non-profit organization so it's vital to protect the intellectual property .
It would be a bad scenario if folk ended up buying fakes unwittingly !

There are some block diagrams on the wiki page , and some good board photos for the various models
but that's all I could find that might help .

Here's an example
https://en.wikipedia.org/wiki/Raspberry ... Bottom.jpg
:lol:
I would have called that the " track side " rather than " Bare Bottom " ,
but now that we have smd components , and layered boards with plated-through holes....
.... I guess the term has lost meaning

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 24, 2019 3:34 pm
by mahjongg
Cost optimizing a design by removing unnecessary components (after a thorough Assessment process) is standard in any industry where products are made in such large numbers that the cost of assessment is worth the cost reduction.

as for board side naming, standard is component (top) side versus solder (bottom) side, which side is the "component side" is determined on which side the primary components, like connectors, are placed. But yes, its just a rule of thumb naming convention, it stems from the time all components were through hole, and boards were (only) wave soldered.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 24, 2019 4:20 pm
by jamesh
Simples. If your board works, simply don't worry about it.

Any PCB in any product almost ever will have areas on the PCB that appear to have parts missing, but in fact are not needed. It's just that because they are in cases, you never see them. Because we sell a bare PCB, any component that appears to be missing is much more obvious, but there is still not reason to worry, because the devices are tested as they come off the production line to make sure they work, and in fact it is incredibly rare for a pick and place machine + the test systems to miss anything.

Of course it does happen sometimes, but mostly it means the Pi won't work when you get it, and you just need to return to the supplier for a replacement. But that's not necessary unless it doesn't work. And like I said - really REALLY rare.

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 24, 2019 6:12 pm
by Will5455
Which means that at least most of the people there are doing there job right!

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 24, 2019 6:34 pm
by florca
Reminds me of the story my Dad used to tell about how Marconi (in their later incarnations) designed TVs - Start off with a really good TV and then take out components one by one until it stops working. Replace the last component. Job done...

Re: Pi 3b+ lost Capacitor on PCB

Posted: Thu Jan 24, 2019 8:14 pm
by drgeoff
florca wrote:
Thu Jan 24, 2019 6:34 pm
Reminds me of the story my Dad used to tell about how Marconi (in their later incarnations) designed TVs - Start off with a really good TV and then take out components one by one until it stops working. Replace the last component. Job done...
Not quite that simple. The order in which you take components out is rather important. The CRT was the most expensive component. Start with that one? :)

Re: Pi 3b+ lost Capacitor on PCB

Posted: Sun Jan 27, 2019 7:17 pm
by mahjongg
florca wrote:
Thu Jan 24, 2019 6:34 pm
Reminds me of the story my Dad used to tell about how Marconi (in their later incarnations) designed TVs - Start off with a really good TV and then take out components one by one until it stops working. Replace the last component. Job done...
I assume you are confused with "Earl "Madman" Muntz" famed for selling eight-track systems for cars, and for the process of "Muntzing", which means taking a side cutter to an electronic device, (like a black and white TV with tubes) and starting removing components (and soldering them back) to see if the device could do without.

https://en.wikipedia.org/wiki/Muntzing