Raspberry Pi doesn't stay powered up for long.

[Bruce Crowthorne]

.... and another!
Same number - like everyone else waited for weeks for RS to ship.
Mine lasts about a minute or so - then hangs / dies.

BLAST

Thanks for posting the return form - but it's a bit ominous that the form exists! I guess they have lots of them failing.

[onepoint21]

Just received mine from Newark this evening. Same issue with the LAN and USB dropping off randomly. Really disappointed. My first gen Pi has been awesome.

[stargeezer]

Add me to the list. - e234156sl-mru94v-01238 and FNXM121115338. Bought from newark/element14. This seems to be a major mess up. Can anyone say quality control? Will see how RMA works.

[DemonJim]

I received two new 512MB RasPis from RS recently and one of them had the exact same problems described in this thread.

One just kept losing USB and Ethernet after 30 seconds or so, and the HDMI even cut out sometimes. Usually booted up OK but never lasted over a couple of minutes before I had to reconnect the keyboard or reboot.

One was perfectly solid when using the exact same image, kb+mouse and PSU, so I knew it wasn't a voltage or software issue. Am SO glad I ordered two, and also glad that at least one worked fine! Spent ages trying different cards, images, power options etc.

Sent it back to RS (costing £2.70) with that form filled out a couple of days ago so will report back if I get a replacement any time soon (and if they send a 256MB one).

[nickfilary]

I had the same issue, tried everything - no luck

So I re-flowed the board with my heat gun - CAREFULLY -
and WA-LA its is working great 4 days now no crashes!!!!!!

if you have nothing to loose give re-flow a try - don't use the oven unless you have to - more precision with a heat gun.

-Nick

[onepoint21]

I had the same issue, tried everything - no luck

So I re-flowed the board with my heat gun - CAREFULLY -
and WA-LA its is working great 4 days now no crashes!!!!!!

if you have nothing to loose give re-flow a try - don't use the oven unless you have to - more precision with a heat gun.

-Nick

What particular part of the board did you focus on? I'd be willing to try.

[DemonJim]

I had the same issue, tried everything - no luck

So I re-flowed the board with my heat gun - CAREFULLY -
and WA-LA its is working great 4 days now no crashes!!!!!!

if you have nothing to loose give re-flow a try - don't use the oven unless you have to - more precision with a heat gun.

-Nick

Great work reflowing and solving it! I will be trying that if the replacement is faulty. Maybe all the faulty ones that get returned could be given to schools for electronics lessons, and the ones that get fixed in "Reflowing Solder 101" mean more working Pis for the school

[onepoint21]

I can confirm now that the poly fuse is blowing. The pi is booting to the desktop and I was able to open a couple of apps before the mouse failed. Measuring the voltage drop across the fuse I can safely say its the cause. Any thoughts on a fix?

[toxibunny]

bridge the polyfuse?

[onepoint21]

Well I took the risk of using a heat gun to heat the poly fuse. Let it sit for half an hour and I'm posting this from the Pi itself. I'm going to let it run for the day to see how stable it is. So far so good. It's been stable for the past 15 minutes.

[georgekok97]

Received two E4612RS2V13B1.0 pis

no problem at all ....
CPU stable (stress tested)
NOT turning off
NOT getting hot (both PCB and CPU)
eth0 turns off if power hungry device be plugged in and coming back when device is unplugged

[onepoint21]

Well I took the risk of using a heat gun to heat the poly fuse. Let it sit for half an hour and I'm posting this from the Pi itself. I'm going to let it run for the day to see how stable it is. So far so good. It's been stable for the past 15 minutes.

Been running the Pi for about 3 1/2 hours now. It's working perfectly. Seems like heating and cooling the fuse did the trick.

[georgekok97]

Well I took the risk of using a heat gun to heat the poly fuse. Let it sit for half an hour and I'm posting this from the Pi itself. I'm going to let it run for the day to see how stable it is. So far so good. It's been stable for the past 15 minutes.

Been running the Pi for about 3 1/2 hours now. It's working perfectly. Seems like heating and cooling the fuse did the trick.

which exactly polyfuse ? Could you post a picture ?? thanks

[onepoint21]

Well I took the risk of using a heat gun to heat the poly fuse. Let it sit for half an hour and I'm posting this from the Pi itself. I'm going to let it run for the day to see how stable it is. So far so good. It's been stable for the past 15 minutes.

Been running the Pi for about 3 1/2 hours now. It's working perfectly. Seems like heating and cooling the fuse did the trick.

which exactly polyfuse ? Could you post a picture ?? thanks

Sure. It's F3 on the opposite side from the USB power. I basically I heated the area with a heat gun moving back and forth for about 30 seconds or so. Then I let it cool for a half hour and tried again. The board is still up and running with no freezing or network drops.

[georgekok97]

Sure. It's F3 on the opposite side from the USB power. I basically I heated the area with a heat gun moving back and forth for about 30 seconds or so. Then I let it cool for a half hour and tried again. The board is still up and running with no freezing or network drops.

it will stop restarting and stopping ethernet when i plug a power hungry device ???

[onepoint21]

it will stop restarting and stopping ethernet when i plug a power hungry device ???

You still need to have a good power supply. I wanted to see if heating the fuse up would essentially allow it to reset. For me it did work. I'm now using it with a mouse, keyboard, Ethernet and HDMI plugged in.

[verm71]

My Raspberry Pi came in today and I've been reading around the forums to try to gather solutions for my problem. It also powers off after a while, randomly. Usually when it's doing something like sudo apt-get upgrade, or startx.

I ran cat /proc/cpuinfo and apparently my revision # is 2, which, from what I understand, means I've got a 256MB unit, despite having ordered a 512MB. I'm running off a 4GB card I got from Newark (Element 14). I'm going to try a newer flash of the SD card, in case the older OS misreports the revision # (which I doubt)..

My board says "Made in China", which I understand some may have had issues.

My power supply is the one that came with the Raspberry Pi from Newark, a 1000mA. I've also tried the power supply from my BlackBerry Playbook, 1.8A.

My keyboard is connected via a powered USB hub.

As far as I can tell, I've done everything as right as I could.

When I measure TP1 and TP2, i get a pretty regular 4.89-4.92V, even when, say, startx starts up but crashes. It then goes to 5.01V, and all LEDs go off, except PWR.

It sounds like I should RMA that board, doesn't it?

[verm71]

I should have also mentioned that I don't have a heatgun, nor am I experienced enough in that area to feel comfortable doing that.

But if I did decide to go out and get or borrow a heatgun, is that all you did, run it over the general area for 30sec? Any particular temperature setting, or nozzle?

Update: With the new image, it seems it's now Rev 000f, and free command reports the full 512MB. Now to check for stability. I'm slightly more optimistic this time around

[onepoint21]

I should have also mentioned that I don't have a heatgun, nor am I experienced enough in that area to feel comfortable doing that.

But if I did decide to go out and get or borrow a heatgun, is that all you did, run it over the general area for 30sec? Any particular temperature setting, or nozzle?

Update: With the new image, it seems it's now Rev 000f, and free command reports the full 512MB. Now to check for stability. I'm slightly more optimistic this time around

I am in no way suggesting you should use a heat gun. It might be more worth your while getting it RMA'd. However I did use the high setting on my heat gun while from several inches away sweeping back and for to reduce the chances of melting compontents.

[verm71]

I am in no way suggesting you should use a heat gun. It might be more worth your while getting it RMA'd.

Good idea. I'm getting the ball rolling with Newark to get an RMA. In the meantime, I was able to get more testing done with the new SD image. At one point I had enough time to get about one quarter of the way through a game of Reversi. I wanted to see if CPU load could trigger the problem. While I played, the CPU seemed to be going in the 80%-90% utilization. And it was fine. Then it just spontaneously shutdown again.

My serial #, FWIW, is: FNXM121127451 (board made in China)

[dani190]

also having issues

Sticker on board shows FNXM121127435

Using power supply from newark, 5v 1000ma

[nickfilary]

I had the same issue, tried everything - no luck

So I re-flowed the board with my heat gun - CAREFULLY -
and WA-LA its is working great 4 days now no crashes!!!!!!

if you have nothing to loose give re-flow a try - don't use the oven unless you have to - more precision with a heat gun.

-Nick

What particular part of the board did you focus on? I'd be willing to try.

I aimed my heat gun at the two chips, the underside of the samsung lsi chip and the topside of the lan controller. I didn't focus on the fuse. When I did this I held my heat gun on a piece of solder to gauge how hot my heat gun was, and was careful not to hold it on the chip for to long. Work great for me, also I let it cool for about 30 minutes or so before powering it back on. Good Luck!!

[medyk]

Again same issue here. After half a year waiting I got dying Pi.

Sticker: E4612RS2V13B1.0
Board number: 1238
Made in China

[dfreedman]

Just received two Rev-B boards from Newark/Farnell.

Sticker: FNXM121127704.
Tried multiple usb supplies/hubs/keyboards/mice. Downloaded and updated latest software on the preflashed SD card from Newark. TP1-TP2 is > 5.0 V.

The USB and Ethernet just reset, it seems under stress. Once I upgraded the Debian software and stopped the GUI from starting automatically, both devices seemed to last a little longer, but not much.

I was never able to get IP address.

It sounds like there's some bad batches and that they really need a better quality control process. Any suggestion on how I should proceed?

[neomat]

Same problem here !

Impossible to get my Pi (E4612RS2V13B1.0) running for more than few minutes.
I tried several power supplies (2 smartphones, RS, ...), Archlinux and Raspbian, and 2 SD cards.
I will certainly try to send it back to RS...
Any advice ?