Rasberry Pi Zero W dies after a couple of hours

[Gal Gofrit]

Hey,

I have a Rapsberry Pi Zero W with a 3D printed case spying on my pet rabbit. 

Whenever I plug it in, it works for a couple of hours, after which it dies (web interface stops working and my NAS is no longer able to access the video feed). The Raspberry Pi is slightly hot to the touch even days later.

Does that mean it overheats? How can I debug this? 

Thanks

[Kevin Shumaker]

If you unplug it for an hour or so, does it start working again?

If you can see the console or ssh in run the command:

vcgencmd measure_temp

and report the number. If it's much above 50-55C it could be heat issues.

If you can check the logs after regaining access look in messages.log (under Expert Settings

Look for lines like:

Sep 24 11:50:49 meye-c0e188a9 user.notice throttlewatch: currently: ok, temperature: 43 C
Sep 24 11:51:49 meye-c0e188a9 user.notice throttlewatch: currently: ok, temperature: 43 C
Sep 24 11:52:49 meye-c0e188a9 user.notice throttlewatch: currently: ok, temperature: 46 C

--
You received this message because you are subscribed to the Google Groups "motioneye" group.
To unsubscribe from this group and stop receiving emails from it, send an email to motioneye+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/motioneye/3b7c9fe9-b16c-4744-ab96-8cc8298152fbn%40googlegroups.com.

--

Thanks

Kevin Shumaker

Personal Tech Support

N38° 19' 56.52"
W85° 45' 8.56"

Semper Gumby
“Don't tell people how to do things. Tell them what to do and let them surprise you with their results.” - G.S. Patton, Gen. USA
Ethics are what we do when no one else is looking.
Quis custodiet ipsos custodes?

“There is no end to the good you can do if you don’t care who gets the credit.” - C Powell

You know we're sitting on four million pounds of fuel, one nuclear weapon and a thing that has 270,000 moving parts built by the lowest bidder. Makes you feel good, doesn't it?

[Gal Gofrit]

Hey, thanks for your answer.

It works immediately when I unplug it and plug it back in, no need to wait. 

I just looked at messages.log and it seems that the highest temperature it ever got to was around 65~. It has been running since I posted this, so around 2:30 hours, and temperature seems about the same (went from 55 to 56.2), so I'm guessing this isn't a temperature issue...

Perhaps it's a memory leak? Or something crashed at some point?

Any ideas I can try?

[Kevin Shumaker]

Can you attach your motion, motioneye, and messages logs, and indicate when was the last time you observed it was good before you noticed it was not responding, please?

If you can ssh into it or have console access, can you also post the results to

df -h <enter>

command?

[Gal Gofrit]

Sure.

I don't recall any specific time, but it's a pattern that's visible in motioneye.log - server started, nothing for a couple of days, and "hello!" a couple of days later - that's me unplugging and plugging back in. 

So, for example, on 2020-09-15 23:34:33 and then on 2020-09-22 19:32:24.

Attaching all of the logs as well as the output of dh.

Thank you for your help :) 

[Kevin Shumaker]

I assume you did a power reset around

Sep 24 17:06:50 meye-a579285b user.notice throttlewatch: currently: ok, temperature: 44 C
                                                                                          Jan  1 02:00:06 meye-a579285b user.notice throttlewatch: currently: ok, temperature: 25 C
Jan  1 02:00:16 meye-a579285b user.info sntp[973]: sntp 4.2....@1.3847-o Sat Jun  6 11:41:06 UTC 2020 (1)
Jan  1 02:00:16 meye-a579285b user.info sntp[973]: 1970-01-01 02:00:16.982574 (-0200) +1600961200.125111 +/- 1067307466.753649 0.pool.ntp.org 162.159.200.1 s3 no-leap
Sep 24 18:26:57 meye-a579285b user.notice date: current system date/time set to Thu Sep 24 18:26:57 IDT 2020 via SNTP

When was the last recording prior to 17:06 on the 24th?

--
You received this message because you are subscribed to the Google Groups "motioneye" group.
To unsubscribe from this group and stop receiving emails from it, send an email to motioneye+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/motioneye/d2c53feb-6e57-4685-a8cf-bc2ab26d15ffn%40googlegroups.com.

[Gal Gofrit]

I actually unplugged it for like an hour today for something unrelated (needed the socket), so this specific one is unrelated

Going back, the only footage I have is from the 22nd, and it's only a couple of minutes - perhaps I can let it run now and update after it stops working with new logs if that'll help

[Kevin Shumaker]

Yep, that would be excellent, if you could do that, please, and note when the last good recording was, too, please.

All 3 logs, motion, motioneye, and messages...

Thank you for helping me to help you.

[Gal Gofrit]

Hey,

I kind of neglected it for two years, but now I'm trying to get it running again... Any chance you can still try to help me now?

I guess it's not a temperature problem, perhaps a power problem?

Is there any way to monitor the power that the Pi recieves?

Thank you...

On Thursday, September 24, 2020 at 9:59:39 PM UTC+3 StarbaseSSD wrote:

Yep, that would be excellent, if you could do that, please, and note when the last good recording was, too, please.

All 3 logs, motion, motioneye, and messages...

Thank you for helping me to help you.

[StarbaseSSD]

It is warm (anything over 60c is problematic). The problem with the temperature in messages.log is that it's the instantaneous temperature, and may miss the 'higher' temperatures because it is throttling, and doesn't run until it cools back down a bit...

If you can SSH in, you van check voltage and temperature issues with

vcgencmd get_throttled

if it reports anything but 0x0, post it here.

[Gal Gofrit]

I don't see it throttling (leaving a tab with `watch vcgencmd get_throttled` in the background, and I don't ever see it change),  but based on vcgencmd measure_temp, temperature seems to be steady around 64c...

If it's problematic, then there's something wrong here...

Room temperature is about 24c, and the Pi is in a case with an open back.

Maybe it has something to do with Pi cam that I'm using?

[StarbaseSSD]

I can write a script for you to check the temp and voltage and throttle status and log them for you.

That way, when the Pi fails you'll get the log with the info. Give me a bit, unless you can write it yourself. I'd use /data/log folder for the file.

[StarbaseSSD]

Here is a quick script to log info:

mkdir /data/usr

nano /data/usr/status.sh

copy and paste the following (minus the '- - - -')

- - - -

#!/bin/bash
temp=$(vcgencmd measure_temp)
volt=$(vcgencmd measure_volts)
throttled=$(vcgencmd get_throttled)
echo $(/bin/date) $volt $temp $throttled >> /data/log/info.log

- - - -

[ctrl]x y [enter] to save

chmod 777 /data/usr/status.sh

crontab -e (select 1 for editor)

add the following:

* * * * * /data/usr/status.sh

[ctrl]x y [enter] to save

It should now save the info once a minute to /data/log/info.log with a time/date stamp.

When it crashes the next time, post the log here.

[Gal Gofrit]

Thanks for the script!

The Pi crashed pretty quickly, but with nothing unusual - and I was thinking cron's 1 minute interval is not granular enough... I changed it to run every ten seconds, you can see that half way through the log.

Temperatures seem to constantly rise, but I don't see throttling...

I think all the time gaps are crashes, but not even sure at this point (if it is throttling everything, bash might get lower priority as well, and I guess the script may be delayed? but I don't really know...)

[StarbaseSSD]

I hate to say it, but...

Disk full?

df -h

Other options...

1) Backup first, then New SDCard, new image and copy the scripts over and restore.

2) The PiZero itself could be dying. I don't see anything in the logs. Depressing...

For grins and giggles, install Raspberry Pi OS Buster / Legacy Desktop, and fire up one of the camera apps, like Cheese and see if the Pi crashes...

I'm at a loss...

[dew...@gmail.com]

I know everyone focuses on amperage, but is that 1.35 voltage right for any RPi ?

 

Dewey

--

You received this message because you are subscribed to the Google Groups "motioneye" group.
To unsubscribe from this group and stop receiving emails from it, send an email to motioneye+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/motioneye/72c58556-fea4-4c6f-a8e9-bfbb61492c5dn%40googlegroups.com.

[StarbaseSSD]

It's the internal CPU voltage, and yes, it's 'close enough'.

https://www.raspberrypi.com/documentation/computers/os.html#measure_volts-block

[StarbaseSSD]

From my PiZeroW:

[root@meye-4cec78b8 ~]# vcgencmd measure_volts
volt=1.3500V
[root@meye-4cec78b8 ~]#
[root@meye-4cec78b8 ~]# vcgencmd measure_volts core
volt=1.2000V
[root@meye-4cec78b8 ~]# vcgencmd measure_volts sdram_c
volt=1.2000V
[root@meye-4cec78b8 ~]# vcgencmd measure_volts sdram_i
volt=1.2000V
[root@meye-4cec78b8 ~]# vcgencmd measure_volts sdram_p
volt=1.2250V
[root@meye-4cec78b8 ~]#

[dew...@gmail.com]

Thanks… WOW, I knew about the 3.3 and the 5, but I never knew the RPi stepped down into the 1’s.

 

Thanks again,

Dewey

To view this discussion on the web visit https://groups.google.com/d/msgid/motioneye/09d17087-c470-4588-a3d3-ee3aaea0e3f4n%40googlegroups.com.

[Gal Gofrit]

Mine are almost identical (core is 1.3500V in my case, the rest are identical) - I guess it's either the SD card or the Pi then, I'll try swapping them around... there seems to be plenty of disk space left.

Do you think my Picam could somehow be the cause? Maybe something on the module is shorting or messed up?

It's from AliExpress, supposedly official, but I can't really know for sure

[StarbaseSSD]

No different than PC hardware. Check out the BIOS settings for CPU and RAM voltages in the overclocking arena....

[StarbaseSSD]

I doubt it, but it is possible the camera itself overheats, especially if it sits right over the PiZeroW CPU and Wifi chips...

[Gal Gofrit]

It's pretty cramped in the case, and it does sit right over the Pi, around the center (I'm not home so can't take a photo right now)

I don't suppose there's a way to check the Pi cam module's temperature? :/

In any case, when I get home, I'll try pulling it out of its case and seeing if it fixes the issue. And I guess the next step would be to try a different SD card.