Red LED on Faint Green LED

[Darkness]

I got a reply from farnell indicating that there might indeed be a problem with a specific series in the batch.

[zbuh]

I got a reply from farnell indicating that there might indeed be a problem with a specific series in the batch.

whats the email that you contacted?
I'm not getting any answer...

[crt404]

I too have been hit by the Red light- Green no light issue. I've contacted Farnell for an RMA and am awaiting their reply.

For info's sake, my Pi has been delivered from the latest batch - I'm just so glad it's not just me having problems - I thought I'd done something wrong with my Pi!

[Darkness]

I got a reply from farnell indicating that there might indeed be a problem with a specific series in the batch.

whats the email that you contacted?
I'm not getting any answer...

I replied to the mail that I got with the tracking number. In my case BelgiumSales (at) farnell.com but it got forwarded internally to a few other people. Eventually someone in Holland replied. If you're from another country I'm not sure if they will help.

[zbuh]

I got a reply from farnell indicating that there might indeed be a problem with a specific series in the batch.

whats the email that you contacted?
I'm not getting any answer...

I replied to the mail that I got with the tracking number. In my case BelgiumSales (at) farnell.com but it got forwarded internally to a few other people. Eventually someone in Holland replied. If you're from another country I'm not sure if they will help.

I already got a phone call from Farnell.
I got an RMA

Going to send it back....

I Hope they don't take much time to deliver a new one.

[andrewuk]

Can't offer any help really but to say I have exactly the same issue.

Steady red power and faint pin-prick green 'OK' LED with a board I received yesterday from RS. Have tried several power supplies and different keyboards mice etc plugged in. Have also tried several SD cards, including the recommended one that came with the board from RS, with the Debian image and had zero success...

Very disappointing - I realise this is a development board but I didn't think for one moment end users would have to fix it before starting out.

Had the kids in school very excited when it turned up yesterday - now it's a bit flat...

[Darkness]

I also tried a different power supply, rated for 1000mA. No change.
Then I tried another SD card and with that it boots fine so far.

The SD card that did not work was a Kingston SDHC class 4, 4GB.
The SD card that works is an EMTEC SDHC Class 10, 8GB.

So at this point, it seems to me that some of the recent board are a little more picky on what SD card you feed it...

[zbuh]

I'm felling really dumb...


I miss read... and jumped this step:

Using the device name of the partition work out the raw device name for the entire disk, by missing out the final "s1" and replacing "disk" with "rdisk":
e.g. /dev/disk1s1 => /dev/rdisk1

http://elinux.org/RPi_Easy_SD_Card_Setup


Could you all that had this problem, please check this?

[Andrew Hogg]

Got my board from Farnell in the last couple of days, had suitable power supply, 4Gb SanDisk class 4 card, put the image on no problem, fired it up - nothing. Blank screen, same Red light and feint green light. Switched power supplies, changed OS, rewrote OS couple of times, no progress.

Switched to a different SanDisk 2Gb card - worked and booted first time.

Really does seem to be very picky on exacrtly which cards it likes to read - I don't see any reason why the 4Gb card failed but happy that at least got it booted up with 2Gb.

[greeneyedtx]

Same problem here! However, i called Farnell and the lady was happy to ship me another one without asking the 21 questions. It seem a bit odd, all thing considering, but i expected a fight and was very much surpised. Almost as if this was expected. All she asked was that i wrote a note about the issue and include it with the return. Hell i think that its going to cost more to ship it back then its worth for the both of them. ohh well, i've waited longer for less.

[ferik]

Anyone having the same issues with a Neward/Element 14 board?

[hackery21]

Honestly, the best solution now for you guys is to get a refund from RS Components LTD or element14. They have spare boards for these exchanges and should ship out pretty fast.

[Darkness]

something else you could try is downloading these files from the boot folder:
https://github.com/raspberrypi/firmware ... aster/boot

and extract them to the FAT32 partition on the SD card (which is in fact /boot for the linux image).
That way, you have a newer kernel which includes some fixes for SD cards. It might help with some boot problems.

[majenko]

Well, I must say, this is rather a disappointment.

Yes, I have the RED on, Green pinpoint problem.

Mine is from RS, and arrived (after months of waiting) this morning.

It's not the power supply on mine, of that I can be certain. I started out with my 1A HTC charger adapter, and moved up to my 17A bench power supply (switched mode - no shortage of current there!)

I have tried both Debian and Arch, both on my 4GB SanDisk Class 4 card, which I know works well, as I use it with RetroBSD on a PIC32 chip.

I suppose I am going to have to try and find a local supplier for a different SD card and try that, but I don't live in the middle of a bustling city, so I may be out of luck.

So far, user experience (and I'm *NOT* a beginner by any means - I build embedded systems for a living): very very very bad.

[majenko]

Ditto... ditto... ditto...

Tried 1A phone charger.

Tried 17A SMPS power supply.

Tried old class 4 SanDisk 4GB.

Been out and bought Class 10 4GB SanDisk Extreme.

All to no avail.

Supplied by RS.

Time to call them.

[eben]

There's been some speculation in this thread that there may be a problem of some sort with this week's batch of units, and I received an email from majenko asking for an official comment. I've just spoken with our distributors to ask whether there as been a spike in RMA rates, and it would appear that there has not; whoever Darkness spoke to at Farnell was unfortunately misinformed.

One thing worth bearing in mind is that in the last ten days something like ten times as many Raspberry Pis have shipped as in the entire prior history of the product. While our partners are of course aiming for a defect rate of zero, this will inevitably lead to an increase in the total number of defective units in the wild. Right now the RMA rate is roughly 1 in 500, and only around half the returned units that I have seen are actually defective; most defective units have had USB or RJ45 issues rather than boot problems.

Boards which are dead on arrival, failing to boot correctly when used with a good power supply, and a correctly flashed SD card from the approved list, are extremely rare, as you would expect given the last thing the factory does before shipping them out is to boot them in order to program the MAC address and check the display, audio, USB and RJ45 functionality. If you do experience this behaviour, please do ask for help (and try Debian, an approved card and a good PSU).

Needless to say, if your board turns out to be genuinely defective, you shouldn't experience any issues getting an RMA number from either of our partners.

[majenko]

Should the GPIO pin that sinks the current for the OK LED be measuring 1.56V?

I'd expect that to be 3.3v (or near) for LED off, and as good as 0v for the LED on. 1.56V seems a very odd value from a digital IO pin.

It's a static DC voltage as well - not a square wave or anything.

[carl_retrotext]

So it turns out the postmen are de soldering components in transit or its the end users fault.

I was lucky (!) to get a 1 in 500 board and Farnell were brilliant in handling my issue and changed the board over without fuss.

I just hope that there is a little more empathy of the difficulties there are, some people are getting put off with the 'you are not good enough to own a PI' attitude that I detect from some posts around the forum.

[eben]

Obviously if you get the 1 board in 500, you personally experience a 100% failure rate, which is less than ideal.

So far we've been very impressed both with our users' perseverance (a false RMA rate of 1 in 1000 is remarkable) and with both companies' RMA procedures. As you note, people with faulty boards seem to be getting replacements very quickly.

Liz and the mods have been stamping on any "not smart enough to own a Pi" posts they come across (this sort of attitude is a quick way to get banned), and will continue to do so.

[27feet]

Mine was shipped by RS this week and appears dead. It has a damaged pin on the sd connector - but it appears to still be okay to make contact so was hopeful it would be okay anyway.

I've tried 8 SD cards (5 off the "okay" list), 5 different supply units (1 was below recommended current, rest were in range + gave within 1/4v of th 5v required when measured). I've even tried different distributions. I haven't tried all combinations but I'm going to have to RMA it.

I had a similar issue with it arriving and telling colleagues + kids at work (a school) only to disappoint. I resorted to my arduino and fez - neat but not a raspberry pi and not driving a screen.

Hoping I can get a speedy replacement - I'll RMA after the bank holiday in the event I have any success trying before then - anyone know if there's any form of readable output on a GPIO port or simiar from power on that may help? Sure I could set the fez or arduino to record output (seen somewhere it's 3.3v rather than "standard" pc serial)

[carl_retrotext]

Obviously if you get the 1 board in 500, you personally experience a 100% failure rate, which is less than ideal.

So far we've been very impressed both with our users' perseverance (a false RMA rate of 1 in 1000 is remarkable) and with both companies' RMA procedures. As you note, people with faulty boards seem to be getting replacements very quickly.

Liz and the mods have been stamping on any "not smart enough to own a Pi" posts they come across (this sort of attitude is a quick way to get banned), and will continue to do so.

thanks Eben, thank you for replying.

I am aware that members are energised and enthused about the PI,this is a good thing, I have had help from some great members ( Andrews S for that Boot code ) and I am sure the Mods will carry on doing a good job here too.

When you feel that you are in a hopeless position, this is the place you start.

[Tharg]

I am pretty sure 1st my board was faulty, Farnell were great, I had the replacement the next day. It worked perfectly (well I need a new USB keyboard I think but I knew that would be a possibility).

Personally I have no complaints. We all know there is a lot that can go wrong. But I have been very impressed by the reaction from I got Farnell when it did go wrong.

[majenko]

Personally I have no complaints. We all know there is a lot that can go wrong. But I have been very impressed by the reaction from I got Farnell when it did go wrong.

You're one of the lucky ones then. I am cursed with the hell that is RS's customer service. I am still waiting for a reply. You can't talk to a human there. I tried phoning, but all I got was a "You have to email to this specific address" thrown at me. An address which doesn't seem to have anyone reading it!

[majenko]

Right, I'm getting somewhere.

I get a very brief flash of data on the SD card data lines, and the CS line flickers high briefly, at the moment of power-on.

The SD card's clock line is permanently running. As this is a dedicated clock coming from GPIO48 I'm not sure if that is really correct.

So there is certainly life in the old girl yet.

I will give the most recent boot code a try as suggested by Darkness - once the git repo has cloned...

I am still concerned by the 1.56V level on the low-side of the OK LED. Should GPIO16 really be at that level? It doesn't make much sense - unless it has a rather high internal resistance...

[majenko]

I will give the most recent boot code a try as suggested by Darkness - once the git repo has cloned...

Nope - no different. I await my new card from Farnell in the morning (I hope) - if that doesn't work, then I am at a loss.

It would be really nice to get some feedback from the Pi.

What would have been really really nice would be if the MicroUSB socket were more than just a power connector. It should really be connected to something akin to an FT-232 chip to provide serial console access to the board. Just think of the wealth of debugging information and such that could be available then. A simple message on the serial saying "SD Card Read Fail" would solve all these headaches.

Or even just a pattern of flashes on the status LED that could be cross-referenced with a table to find the error condition.

Can we get the source code to the boot system, so we can add our own debugging and such to it? Or is this open hardware closed?