Faulty Raspberries

[antoniomilano]

Two RPi purchased = two faulty boards.
Is there any QC at Raspberry Pi ?
Don't buy Raspberry Pi !
I'm gettin' back to my faithful BeagleBoard

[piglet]

Two faulty boards arriving for a single person would indicate PEBKAC to me.

If you will provide more information on symptoms seen I suspect that you will be told how to get your Pi boards working.

[jamesh]

It's certainly possible you got two faulty boards, but extremely unlikely. Yes, there is QC at the factories - they are electrically tested and all leave working. It would be foolish in the extreme not to test prior to dispatch.

So, what were the symptoms you were seeing? Because it's more likely you have a problem elsewhere rather than with both Raspberries.

If you have given up completely, and returned to the Beagleboard, and decided just to post to say don't buy Pi, well, that's not particularly helpful.

[antoniomilano]

You can find other people in this forum having issues with your boards.
My RPi's cease to work after 2-5 minutes without doing anything.
In my opinion they are not tested.
I think that neither Raspberry Pi nor RS are taking in consideration the problem.
I don't want to spend other money to send you back the boards without to be sure to receive tested and working ones.
sincerely
antonio

[RaTTuS]

you don't spend money with an RMA

what are your issues , exactly

there are not many posts about broken RPi's

I have 4 and no issues

have you looked at
http://elinux.org/R-Pi_Troubleshooting

what have you plugged in ?
what is your power as measured from TP1 TP2

[jamesh]

You can find other people in this forum having issues with your boards.
My RPi's cease to work after 2-5 minutes without doing anything.
In my opinion they are not tested.
I think that neither Raspberry Pi nor RS are taking in consideration the problem.
I don't want to spend other money to send you back the boards without to be sure to receive tested and working ones.
sincerely
antonio

Well, your opinion is wrong - they ARE tested. However, they are not SOAK tested - that would not be possible - this is a high volume production line, and there isn't time to soak test. Note that very few other commercial devices are soak tested either, it's not just the Raspi. I'm not sure of the exact tests that are done on the line, but I believe they are mostly testing electrical connections I believe, to ensure all the components are correctly seated and soldered.

As to other people commenting on the board about devices not working as expected - we are aware of that - there are a number of threads about various issues, including USB issues and SD card corruption issues. However, when you consider nearly a million boards have been sold the number of problems is fairly small in comparison. I should add that all faults are considered- we do NOT ignore problems (despite what some people think - just because there is nothing visible on the surface this does not mean things are not going on underneath the surface) and we are actively working on both the USB and SD card issues.

Without you being specific about what your problems are we are unable to say whether the problems you are seeing are related to a hardware fault, or a software fault. There is not point in RMAing a board if it's a software fault.

[FFAMax]

Hello, jamesh!

I also want to say that at this week we got broken PI
We got about 50pcs before and all working good.
But last 2 Pi was last revision with 512 MB, 1 of them booting ok, but second not booting.
Same SD card uses.
So, here HW problem or boot loader not working.

[drgeoff]

The posts in http://www.raspberrypi.org/phpBB3/viewt ... 28&t=20657 may indicate that some RPis are passing the factory test but would fail a soak test. However, without those boards being professionally examined it is not possible to come to a definite conclusion.

[FFAMax]

The posts in http://www.raspberrypi.org/phpBB3/viewt ... 28&t=20657 may indicate that some RPis are passing the factory test but would fail a soak test. However, without those boards being professionally examined it is not possible to come to a definite conclusion.

I can ship it to specialist which can make test. (Easy, but more time needed)
Or i can connect debugger (which?) to board and make needed operations. (May be fast but practice needed)

What we can do in my case?

I thinks that same cases important and must be investigated.

[FFAMax]

Nobody not want to resolve problem with broken Pi. It's so sadly.
People, just know, if you got broken Pi you can drop it to bin cuz no support here

[malakai]

First if you believe the board is defective contact the distributor. If you have 52 and only 1 isn't working requesting a replacement with the distributor it would not be difficult. You do not have to go to all the trouble to try and diagnose the solution to why it doesn't work. Attach power, test voltage, swap working cards, you have a known working power supply, keyboard, mouse, and video cable. I would assume this as you have more than 50 working pieces. Please post what you have tried. To claim there is no help there are probably 1,000 pages of troubleshooting steps on this board and the wiki. Just because others don't search for your answers does not mean there is no support.

For the other posts yes you can find those that may have boards that were shipped with a defect but I bet if you counted them all up and then realized that around one million boards are out there you would find approximately 1% find a manufactured product with less failures than that and choose them as your favorite manufacturer. Because that is a very decent rating in the industry.

[RaTTuS]

Hello, jamesh!

I also want to say that at this week we got broken PI
We got about 50pcs before and all working good.
But last 2 Pi was last revision with 512 MB, 1 of them booting ok, but second not booting.
Same SD card uses.
So, here HW problem or boot loader not working.

you got 50 one is suspect - make sure by trying against the working ones then contact the distributer if it still fails and they will RMA it.

[jamesh]

Nobody not want to resolve problem with broken Pi. It's so sadly.
People, just know, if you got broken Pi you can drop it to bin cuz no support here

Sorry, but what? Have you talked to the people who supplied your board? They are the people to talk to if there is a hardware fault. Meanwhile, you can ask specific questions here and will be many many answers, which may or may not help. Just saying "my board doesn't work" is not good enough. Details are need to diagnose problems.

But to suggest there is no support is plainly wrong.

[FFAMax]

Sorry, but what? Have you talked to the people who supplied your board? They are the people to talk to if there is a hardware fault. Meanwhile, you can ask specific questions here and will be many many answers, which may or may not help. Just saying "my board doesn't work" is not good enough. Details are need to diagnose problems.

Yes, Farnell already informed. But what? RMA? Farnell have RMA procedure?
Does not matter.

Correct position - find bug and fix.
If you tell that boards testing at manufacture then my board must been detected at manufacture.

I repeat: RMA - it is not solution - this is workaround.
I ask "who can investigate my case" - no answers.
I understand this as "nobody do not want to analyse it".

There are 2 positions:
1. You got broken board? - ok, we replace it.
2. You got broken board? - how it possible? We have Quality Control at factory and this is not possible! We want to fix our fail and find bad worker (or etc.) and prevent it in the future.

Now i see position #1.

And find main thing - I am searching man who can diagnoze my fail board and (this man must be from Rasp. team). This is must be professional who can detect problem in manufacture process.

Note: this is not my first board and i know quality but think about:
new member order Rasp. pi and got broken board and think: "damn, it was my last 35$"

And second: China != UK or you want to broke this statement?

I not will be to discuss following treads and post about fails.
I wrote enough about this problem and my position. You may want to do good product or not.

Good Luck!

[Burngate]

I believe English may not be your first language, but you use English better than I could use your first language, whichever language that is
However I do have difficulty understanding what you write

...I repeat: RMA - it is not solution - this is workaround.
I ask "who can investigate my case" - no answers.
I understand this as "nobody do not want to analyse it".

There are 2 positions:
1. You got broken board? - ok, we replace it.
2. You got broken board? - how it possible? We have Quality Control at factory and this is not possible! We want to fix our fail and find bad worker (or etc.) and prevent it in the future.

Half a million boards out there, less than 1% failure rate, so that could be 5000 bad boards - should they sack 5000 workers?
If you RMA your faulty board, they may be able to find out what's wrong, and prevent it happening again.
If you don't RMA it they can't, so other people will suffer.

And find main thing - I am searching man who can diagnoze my fail board and (this man must be from Rasp. team). This is must be professional who can detect problem in manufacture process.

Surely man who can diagnose work for manufacturer - send it back and he can diagnose

[jamesh]

You seem to under the impression that manufacturing is a perfect science. It isn't. There is a acceptable failure rate for any production line. Many faults are detected before products leave a factory, but some are not - they may be faults that only become evident after transport (perhaps a weak solder joint) , or after the device has been turned on for a certain amount of time. There are many reasons. Manufacturers make a judgement on what an acceptable failure rate is, and design their production line testing around that.

You could make a production line extremely robust, by having product on shakers, or soak testing them, but then the product takes so long to make and test it becomes too expensive to make and sell! The failure rate is a happy medium.

Short answer, you need to RMA you device if you think it is faulty. That's why companies have RMA systems, because they KNOW there will be some failures. If there were NO failures, there would be no need for returns, no need for repair shops, no need for warranties.

[jamesh]

I believe English may not be your first language, but you use English better than I could use your first language, whichever language that is
However I do have difficulty understanding what you write

...I repeat: RMA - it is not solution - this is workaround.
I ask "who can investigate my case" - no answers.
I understand this as "nobody do not want to analyse it".

There are 2 positions:
1. You got broken board? - ok, we replace it.
2. You got broken board? - how it possible? We have Quality Control at factory and this is not possible! We want to fix our fail and find bad worker (or etc.) and prevent it in the future.

Half a million boards out there, less than 1% failure rate, so that could be 5000 bad boards - should they sack 5000 workers?
If you RMA your faulty board, they may be able to find out what's wrong, and prevent it happening again.
If you don't RMA it they can't, so other people will suffer.

And find main thing - I am searching man who can diagnoze my fail board and (this man must be from Rasp. team). This is must be professional who can detect problem in manufacture process.

Surely man who can diagnose work for manufacturer - send it back and he can diagnose

Eventually, boards do get back to the Foundation for testing (if manufacturer cannot find the fault), but only if they are RMA to the manufacturer first!

[Lob0426]

Most electronic devices will fail, if they are going to fail, within the first 30 days of use. A large number of them will fail in the first 24 of use. The testing of the devices is usually less than a minute.

I could see your point if this were a $10,000 device, it should have been thoroughly tested and guaranteed working. But at $35, if it quits, send it back and get another one or a refund.

I had one that the SD card corrupted every write. I called them and they gave me an RMA E-mail. I had the replacement in less than a week.

Take the time to tell us what it is or is not doing, the version and maybe we can help you either find a fix or decide to send it back. Your "shoot the tester" talk is not really getting you anywhere. If we shot the testers there would be no one left at Toyota or Ford or Chevrolet or Dell or HP, well hopefully you get the point by now!

[FFAMax]

Eventually, boards do get back to the Foundation for testing (if manufacturer cannot find the fault), but only if they are RMA to the manufacturer first!

How i can send item to the Foundation for testing?

No way to RMA with Farnell for 1 pcs.

I can send directly - it possible.
To Farnell - not possible (very very very hurdly).

[jamesh]

Eventually, boards do get back to the Foundation for testing (if manufacturer cannot find the fault), but only if they are RMA to the manufacturer first!

How i can send item to the Foundation for testing?

No way to RMA with Farnell for 1 pcs.

I can send directly - it possible.
To Farnell - not possible (very very very hurdly).

You really need to RMA to Farnell (you should be able to do that fairly easily, but I do not know the exact mechanism), that way you will get a replacement - sending to the Foundation you won't because they don't keep any stock. Farnell will then test, and if they cannot find the fault, they will usually send it on to the Foundation. Once the Foundational gets it, it'll be looked at, and if necessary the problem will be escalated to the SoC supplier. That's pretty unlikely.

[redhawk]

What level of checks do factories perform on the PI is it simply a case of plug in and check the TV works or do they actually test the LAN and USB ports too??

I remember a while ago someone had a problem with his PI it was rebooting pretty much as soon as the power was on.
It turned out that 2 of the GPIO pins had been bridged together thanks to a blob of solder.

Now if every PI was tested before shipping out how the hell did no one noticed a problem with this board??

To be honest I'm not totally convinced every PI is tested it's probably more likely to be 1 in every 100 batch.

Richard S.

[hop]

I haven't had a single issue with mine.

[SN]

As a multi Pi owner of ooooo six months now and frequent visitor to this board I have to say that the incidence of "broken Pi's" seems to have jumped significantly in the last month or so. Now this is either because

a) there's an underlying problem with a recent batch or batches
b) there are tons more Pi's out there and thus chances of failure are higher
c) people are just more vocal about issues now

Just my view. Would be interested to know the Foundations view...

[RaTTuS]

RMA via farnell is very easy
http://uk.farnell.com/jsp/bespoke/bespo ... tid=415779
^ the UK version - but it will have a local version for you.

if there is a problem in production - this can help to find it -
if that passes then it will go back to the foundation for Extra lookings

[jamesh]

What level of checks do factories perform on the PI is it simply a case of plug in and check the TV works or do they actually test the LAN and USB ports too??

I remember a while ago someone had a problem with his PI it was rebooting pretty much as soon as the power was on.
It turned out that 2 of the GPIO pins had been bridged together thanks to a blob of solder.

Now if every PI was tested before shipping out how the hell did no one noticed a problem with this board??

To be honest I'm not totally convinced every PI is tested it's probably more likely to be 1 in every 100 batch.

Richard S.

I don't know the specific checks, but I can try and find out. It may also depend on which factory is building the boards (I believe there are three). I think every board is tested in some way.