New OPi Zero - Yet another high temperature issue...

bozden · May 19, 2017

My third Orange Pi Zero came last week, yesterday I plugged it and had heating issues... So I started reading, learning and testing....

Setup:

OPiZ-1: 256 MB, no SPI flash - Has extension board plugged - for comparison
OPiZ-2: 512 MB, with SPI flash
OPiZ-3: 256 MB, with SPI flash (new one with problem)

Test conditions:

All with (newish) Sandisk Ultra 16 GB Class 10
All new formatted (SD Formatter) and burned/validated (Etcher v1.0.0) with Armbian_5.25_Orangepizero_Debian_jessie_default_3.4.113.img
All on my desk - side by side, ambient temperature 25.0-25.1 C (20 cm from boards, no air flow)
No case, no passive/active cooling
Power supply 7 port 5V 36W total, 2.4A per port (wait, wait, I tried others, separated etc - I'm testing them with USB V/A/W tester - 5.2-5.3V / 0.260-0.350 A)
First install done, rebooted, apt-get update && apt-get upgrade'd, rebooted, tested

The results are given below...

I switched (i.e. reformatted/reinstalled) the SD Card of OPiZ-3, no change...

I plugged OPiZ-3 alone, no change...

I used 3 other power sources, no change...

Is the new OPiZ-3 a faultly board? What else can I test/check?

Thanks...

OPiZ-1:

h3consumption -p

Spoiler

Active settings:

cpu 1200 mhz allowed, 1200 mhz possible, 4 cores active

dram 408 mhz

hdmi/gpu off

usb ports active

eth0 100Mb/s/Full, Link: yes

wlan0 IEEE 802.11bgn ESSID:off/any
Mode:Managed Access Point: Not-Associated Tx-Power=20 dBm
Retry long limit:7 RTS thr:off Fragment thr:off
Encryption key:off
Power Management:on

armbianmonitor -m

Spoiler

Stop monitoring using [ctrl]-[c]
Time CPU load %cpu %sys %usr %nice %io %irq CPU
11:43:27: 1008MHz 0.01 1% 0% 0% 0% 0% 0% 49°C
11:43:32: 240MHz 0.01 1% 0% 0% 0% 0% 0% 48°C
11:43:37: 240MHz 0.01 1% 0% 0% 0% 0% 0% 47°C
11:43:43: 240MHz 0.01 1% 0% 0% 0% 0% 0% 48°C
11:43:48: 240MHz 0.01 1% 0% 0% 0% 0% 0% 48°C
11:43:53: 240MHz 0.01 1% 0% 0% 0% 0% 0% 48°C
11:43:58: 240MHz 0.01 1% 0% 0% 0% 0% 0% 49°C
11:44:04: 240MHz 0.01 1% 0% 0% 0% 0% 0% 48°C
11:44:09: 240MHz 0.01 1% 0% 0% 0% 0% 0% 48°C
11:44:14: 240MHz 0.08 1% 0% 0% 0% 0% 0% 49°C
11:44:19: 240MHz 0.07 1% 0% 0% 0% 0% 0% 48°C
11:44:25: 240MHz 0.07 1% 0% 0% 0% 0% 0% 48°C
11:44:30: 240MHz 0.06 1% 0% 0% 0% 0% 0% 49°C
11:44:35: 240MHz 0.06 1% 0% 0% 0% 0% 0% 48°C
11:44:41: 240MHz 0.05 1% 0% 0% 0% 0% 0% 48°C
11:44:46: 240MHz 0.05 1% 0% 0% 0% 0% 0% 48°C
11:44:51: 240MHz 0.04 1% 0% 0% 0% 0% 0% 49°C
11:44:56: 240MHz 0.04 1% 0% 0% 0% 0% 0% 48°C
11:45:02: 240MHz 0.04 1% 0% 0% 0% 0% 0% 47°C
11:45:07: 240MHz 0.03 1% 0% 0% 0% 0% 0% 47°C
11:45:12: 240MHz 0.03 1% 0% 0% 0% 0% 0% 49°C
11:45:17: 240MHz 0.03 1% 0% 0% 0% 0% 0% 48°C
11:45:23: 240MHz 0.03 1% 0% 0% 0% 0% 0% 50°C
11:45:28: 240MHz 0.02 1% 0% 0% 0% 0% 0% 47°C

OPiZ-2:

h3consumption -p

Spoiler

Active settings:

cpu 1200 mhz allowed, 1200 mhz possible, 4 cores active

dram 408 mhz

hdmi/gpu off

usb ports active

eth0 100Mb/s/Full, Link: yes

wlan0 IEEE 802.11bgn ESSID:off/any
Mode:Managed Access Point: Not-Associated Tx-Power=20 dBm
Retry long limit:7 RTS thr:off Fragment thr:off
Encryption key:off
Power Management:on

armbianmonitor -m

Spoiler

Stop monitoring using [ctrl]-[c]
Time CPU load %cpu %sys %usr %nice %io %irq CPU
11:33:57: 1008MHz 0.06 4% 1% 1% 0% 1% 0% 48°C
11:34:02: 240MHz 0.05 4% 1% 1% 0% 1% 0% 48°C
11:34:07: 240MHz 0.05 4% 1% 1% 0% 0% 0% 48°C
11:34:12: 240MHz 0.04 4% 1% 1% 0% 0% 0% 49°C
11:34:18: 240MHz 0.04 4% 1% 1% 0% 0% 0% 47°C
11:34:23: 240MHz 0.04 3% 1% 1% 0% 0% 0% 48°C
11:34:28: 240MHz 0.03 3% 1% 1% 0% 0% 0% 48°C
11:34:33: 240MHz 0.03 3% 1% 1% 0% 0% 0% 49°C
11:34:39: 240MHz 0.03 3% 1% 1% 0% 0% 0% 48°C
11:34:44: 240MHz 0.03 3% 1% 1% 0% 0% 0% 48°C
11:34:49: 240MHz 0.02 3% 1% 1% 0% 0% 0% 47°C
11:34:55: 240MHz 0.02 3% 1% 1% 0% 0% 0% 47°C
11:35:00: 240MHz 0.02 3% 1% 1% 0% 0% 0% 48°C
11:35:05: 240MHz 0.02 3% 1% 1% 0% 0% 0% 46°C
11:35:10: 240MHz 0.02 3% 1% 1% 0% 0% 0% 49°C
11:35:16: 240MHz 0.01 3% 1% 1% 0% 0% 0% 49°C
11:35:21: 240MHz 0.01 3% 1% 1% 0% 0% 0% 47°C
11:35:26: 240MHz 0.01 3% 1% 1% 0% 0% 0% 48°C
11:35:31: 240MHz 0.01 3% 1% 1% 0% 0% 0% 47°C
11:35:37: 240MHz 0.01 3% 1% 1% 0% 0% 0% 47°C
11:35:42: 240MHz 0.01 3% 1% 1% 0% 0% 0% 49°C
11:35:47: 240MHz 0.01 3% 1% 1% 0% 0% 0% 46°C
11:35:53: 240MHz 0.01 3% 1% 1% 0% 0% 0% 48°C
11:35:58: 240MHz 0.01 3% 1% 1% 0% 0% 0% 48°C
11:36:03: 240MHz 0.01 3% 1% 1% 0% 0% 0% 48°C
11:36:08: 240MHz 0.01 3% 1% 1% 0% 0% 0% 47°C
11:36:14: 240MHz 0.01 3% 1% 0% 0% 0% 0% 47°C
11:36:19: 240MHz 0.01 3% 1% 0% 0% 0% 0% 47°C
11:36:24: 240MHz 0.01 3% 1% 0% 0% 0% 0% 46°C
11:36:29: 240MHz 0.00 3% 1% 0% 0% 0% 0% 46°C

OPiZ-3:

h3consumption -p

Spoiler

Active settings:

cpu 1200 mhz allowed, 1200 mhz possible, 4 cores active

dram 408 mhz

hdmi/gpu off

usb ports active

eth0 100Mb/s/Full, Link: yes

wlan0 IEEE 802.11bgn ESSID:off/any
Mode:Managed Access Point: Not-Associated Tx-Power=20 dBm
Retry long limit:7 RTS thr:off Fragment thr:off
Encryption key:off
Power Management:on

armbianmonitor -m

Spoiler

Stop monitoring using [ctrl]-[c]
Time CPU load %cpu %sys %usr %nice %io %irq CPU
11:36:35: 768MHz 0.06 2% 1% 1% 0% 0% 0% 67°C
11:36:40: 240MHz 0.20 2% 1% 1% 0% 0% 0% 67°C
11:36:45: 240MHz 0.18 2% 1% 1% 0% 0% 0% 66°C
11:36:50: 240MHz 0.17 2% 1% 1% 0% 0% 0% 66°C
11:36:56: 240MHz 0.15 2% 1% 1% 0% 0% 0% 66°C
11:37:01: 240MHz 0.14 2% 1% 1% 0% 0% 0% 67°C
11:37:06: 240MHz 0.13 2% 1% 1% 0% 0% 0% 67°C
11:37:12: 240MHz 0.12 2% 1% 1% 0% 0% 0% 67°C
11:37:17: 240MHz 0.11 2% 1% 1% 0% 0% 0% 67°C
11:37:22: 240MHz 0.10 2% 1% 1% 0% 0% 0% 67°C
11:37:27: 240MHz 0.09 2% 1% 1% 0% 0% 0% 67°C
11:37:33: 240MHz 0.09 2% 1% 1% 0% 0% 0% 67°C
11:37:38: 240MHz 0.08 2% 1% 1% 0% 0% 0% 67°C
11:37:43: 240MHz 0.07 2% 1% 1% 0% 0% 0% 66°C
11:37:48: 240MHz 0.07 2% 1% 1% 0% 0% 0% 67°C
11:37:54: 240MHz 0.06 2% 1% 0% 0% 0% 0% 67°C
11:37:59: 240MHz 0.06 2% 1% 0% 0% 0% 0% 66°C
11:38:04: 240MHz 0.05 2% 1% 0% 0% 0% 0% 67°C
11:38:10: 240MHz 0.05 2% 1% 0% 0% 0% 0% 67°C
11:38:15: 240MHz 0.04 2% 1% 0% 0% 0% 0% 67°C
11:38:20: 240MHz 0.12 2% 1% 0% 0% 0% 0% 67°C
11:38:25: 240MHz 0.25 2% 1% 0% 0% 0% 0% 66°C
11:38:31: 240MHz 0.23 2% 1% 0% 0% 0% 0% 67°C

prisma · September 24, 2017

I would recommend no one to buy a Zero currently.

I ordered 5 (got V1.4), 2 were DOA and 3 didn't last longer than 1 Week (idleing).

Something is shorting the zero, so voltage falls and current goes to max. Then the power supply drops out and restarts. So I get something like a boot loop whit blinking Ethernet port. Power supply I used delivers 5.20V@2A, and it works perfectly on a V1.1. Does someone have an Idea howto quickly check the voltage regulator, replacing this is something I might be bothered.

As I need a cheap board with ethernet, is the Nanopi Neo a better choice, or does anyone know similar alternatives?

Igor · September 25, 2017

7 hours ago, prisma said:

As I need a cheap board with ethernet, is the Nanopi Neo a better choice, or does anyone know similar alternatives?

All small boards face some common issues: overheating and lack of proper power regulator. Opi Zero (1.4) is a clear winner in this - who heats most - competition.

Give 10$ more for OrangePi PC+ or 2E+ if for some reason needs a gigabit ethernet.

pfeerick · September 25, 2017

What Igor said. From what I've seen mentioned, it sounds like in this case "newer != better" :-( I have two V1.1 Orange Pi Zeros that I purchased 2-3 months ago via their aliexpress page, and (after testing that both worked!) I have had one running 24x7 since I got it, open air without a case, plugged into ethernet running node-red and some other bits. I turned everything down power-wise and down-clocked it using the h3consumption script, and haven't had any issues with it, and it runs nice and cool. It does have a heatsink on it, but it certainly doesn't *need* it, but it helps keep it cool and let it throttle up for longer when needed.

Jan Dovrtěl · October 7, 2017

This is a big shame at Orange Pi Zero H2 + v1.4 it's a great device at an unbeatable price ($ 7 256MB + $ 4 post: o /)
70 ° C in a quiet state is really a lot. If it were a solution to get the temperature down to 30 ° C (undercoating, cooler), I'd like to get acquainted with them.

ullbeking · November 12, 2017

I am an Orange Pi newbie and was recommended the OPi Zero by a trusted source due to its simplicity. I ordered a few as spares (I may have to do some soldering, and also experimentation in the way I supply power to the board). I haven't actually used it yet. How can I tell if I got one of the "bad" revisions?

Oh, I also got a few OPi R1 boards at the same time, as they seemed similarly simple to the Zero. Are there any known problems with the R1?

chwe · November 12, 2017

7 minutes ago, ullbeking said:

How can I tell if I got one of the "bad" revisions?

By reading this whole thread or using the search engines...

10 minutes ago, ullbeking said:

Are there any known problems with the R1?

By reading another thread....

mpmc · November 12, 2017

Just thought I'd pop in and report that on my zero running the nightly stretch, the temp seems to have gone down a hell of a lot (when idle).

It used to get quite toasty (58c+) idle but now it's much better & this is with the xradio enabled (I had it off before)!

Well done guys!

Spoiler

mark@192.168.1.100's password:
___ ____ _ _____
/ _ \ _ __ __ _ _ __ __ _ ___ | _ \(_) |__ /___ _ __ ___
| | | | '__/ _` | '_ \ / _` |/ _ \ | |_) | | / // _ \ '__/ _ \
| |_| | | | (_| | | | | (_| | __/ | __/| | / /| __/ | | (_) |
\___/|_| \__,_|_| |_|\__, |\___| |_| |_| /____\___|_| \___/
|___/

Welcome to ARMBIAN 5.34.171113 nightly Debian GNU/Linux 9 (stretch) 4.13.12-sunxi
System load: 0.00 0.00 0.00     Up time: 10:01 hours
Memory usage: 9 % of 493MB     IP: 192.168.1.100
CPU temp: 44°C
Usage of /: 8% of 15G

ullbeking · November 14, 2017

On 12/11/2017 at 11:51 PM, mpmc said:

Just thought I'd pop in and report that on my zero running the nightly stretch, the temp seems to have gone down a hell of a lot (when idle).

It used to get quite toasty (58c+) idle but now it's much better & this is with the xradio enabled (I had it off before)!

@mpmc Is this with OPi Zero rev. 1.4?

And are you saying that recent kernel versions include changes so that it doesn't run so hot? So that the effects of this hardware change can be mitigated by this software change?

I looked at my boards (I ordered a few) and they all have the rev. 1.4 changes pictured near the start of the thread. Unless I can get them to idle cooler, and also perform cooler under load, they will be practically useless.

mpmc · November 14, 2017

3 minutes ago, ullbeking said:

@mpmc Is this with OPi Zero rev. 1.4?

And are you saying that recent kernel versions include changes so that it doesn't run so hot? So that the effects of this hardware change can be mitigated by this software change?

I looked at my boards (I ordered a few) and they all have the rev. 1.4 changes pictured near the start of the thread. Unless I can get them to idle cooler, and also perform cooler under load, they will be practically useless.

I believe it's the older version.

Michael Dehn Klit · November 14, 2017

6 hours ago, ullbeking said:

Unless I can get them to idle cooler, and also perform cooler under load, they will be practically useless.

I got 4 of the 1.4 versions, and with sysbench on the next kernel, they run around 70C, with passive cooling, and around 44-48C "idle", they run Node-Red and mosquitto. My Opi plus2 runs around the same temperature with active cooling with a minecraft server, and no problems so far.

lzb · December 1, 2017

So there is some improvment over recent kernels or 1.4 is just "randomly broken"?

konsgn · December 12, 2017

Just a heads up, I think the bigger issue between the revision 1.1 and 1.4 is that they removed the "U5" buck AVCC/RTC 3.3V converter, Instead in its place they put a "R9" 0 ohm resistor from the GPIO 3.3v regulator "u55". As such it makes sense that they removed the "Q11" gpio voltage enable switch, since now that switch must always be on.

In testing the voltages, the Rev 1.1 GPIO VCC is 3.37v, whereas the AVCC/RTC Vcc is 3.27v.

For Rev1.4 though, the GPIO/AVCC/RTC Vcc is actually 3.4v.

I am currently removing "R9" from rev 1.4 and installing a diode to drop some voltage. It may help to keep the device from overheating. If that doesn't help, perhaps adding a secondary 3.3v buck regulator can fix the issue.

Update: With the diode to drop voltage, the resulting voltage on AVCC/RTC is 2.91V.
This results in the following temp readings stable/finger tested:

19:02:53: 1200MHz 0.71 19% 11%   6%   0%   1%   0%   -2°C
19:02:58: 1008MHz 0.73 19% 11%   6%   0%   1%   0%   -4°C
19:03:03: 240MHz 0.67 19% 11%   6%   0%   1%   0%   -9°C
19:03:08: 240MHz 0.62 19% 11%   6%   0%   1%   0% -10°C
19:03:13: 240MHz 0.57 18% 10%   5%   0%   1%   0% -11°C
19:03:18: 240MHz 0.52 18% 10%   5%   0%   1%   0% -17°C
19:03:24: 240MHz 0.48 17% 10%   5%   0%   1%   0% -20°C
19:03:29: 240MHz 0.44 17% 10%   5%   0%   1%   0% -21°C

This leads me to believe that the temperature may not be significantly if at all different between v1.1 and v1.4. It would make sense for the slight voltage difference of voltage on the AVCC pins would change the internal temperature readings. They probably don't have any sort of voltage reference internally, and that would lead to any sort of internal reading based on analog voltages to be affected by voltage changes of the AVCC power.

I will test this with a power supply on the seperated AVCC/RTC to see if it does indeed result in different internal readouts.

Alright, here's what I found:

All of these results are running the armbianmonitor right after start up and each show a "finger test" draw represents draw on the Avcc/rtc from the power supply, and is quite stable.

Orangepi v1.4 test: r9 removed, AVCC&RTC=3.27 about 50mA draw during test:

Quote

Time        CPU    load %cpu %sys %usr %nice %io %irq   CPU
19:00:55: 1200MHz 1.28 37% 18% 12%   0%   6%   0%   35°C
19:01:00: 240MHz 1.26 35% 17% 11%   0%   5%   0%   31°C
19:01:05: 240MHz 1.21 32% 16% 10%   0%   5%   0%   30°C
19:01:10: 240MHz 1.11 30% 15% 10%   0%   4%   0%   30°C
19:01:15: 240MHz 1.02 28% 14%   9%   0%   4%   0%   30°C
19:01:20: 240MHz 0.94 26% 13%   8%   0%   4%   0%   30°C
19:01:26: 240MHz 0.87 24% 12%   8%   0%   3%   0%   30°C
19:01:31: 240MHz 0.80 23% 11%   7%   0%   3%   0%   30°C
19:01:36: 240MHz 0.73 22% 11%   7%   0%   3%   0%   25°C
19:01:41: 240MHz 0.67 21% 10%   7%   0%   3%   0%   25°C
19:01:47: 240MHz 0.62 20% 10%   6%   0%   3%   0%   25°C
19:01:52: 240MHz 0.57 19%   9%   6%   0%   3%   0%   24°C
19:01:57: 240MHz 0.52 18%   9%   6%   0%   2%   0%   25°C
19:02:02: 1200MHz 0.48 18%   9%   6%   0%   2%   0%   31°C
Time        CPU    load %cpu %sys %usr %nice %io %irq   CPU
19:02:07: 1200MHz 0.52 18%   9%   6%   0%   2%   0%   34°C
19:02:12: 1200MHz 0.56 18%   9%   6%   0%   2%   0%   36°C
19:02:17: 1200MHz 0.60 19% 10%   6%   0%   2%   0%   37°C
19:02:23: 1200MHz 0.63 19% 10%   6%   0%   2%   0%   38°C
19:02:28: 1200MHz 0.66 19% 10%   6%   0%   2%   0%   39°C
19:02:33: 1200MHz 0.69 19% 10%   6%   0%   2%   0%   40°C
19:02:38: 1200MHz 0.71 19% 11%   6%   0%   2%   0%   41°C
19:02:43: 1200MHz 0.73 20% 11%   6%   0%   2%   0%   41°C

Orangepi v1.4 test: r9 removed, AVCC&RTC=3.4V about 50mA draw during test:

Quote

Time        CPU    load %cpu %sys %usr %nice %io %irq   CPU
19:00:55: 1200MHz 1.08 35% 18% 12%   0%   4%   0%   58°C
19:01:00: 240MHz 1.00 32% 17% 11%   0%   3%   0%   52°C
19:01:06: 240MHz 0.92 30% 16% 10%   0%   3%   0%   51°C
19:01:11: 240MHz 0.84 27% 14%   9%   0%   3%   0%   51°C
19:01:16: 240MHz 0.78 26% 13%   9%   0%   2%   0%   51°C
19:01:21: 240MHz 0.71 24% 13%   8%   0%   2%   0%   51°C
19:01:27: 240MHz 0.66 23% 12%   8%   0%   2%   0%   51°C
19:01:32: 240MHz 0.60 21% 11%   7%   0%   2%   0%   51°C
19:01:37: 240MHz 0.56 20% 11%   7%   0%   2%   0%   51°C
19:01:42: 240MHz 0.51 19% 10%   6%   0%   2%   0%   50°C
19:01:48: 240MHz 0.47 18% 10%   6%   0%   2%   0%   45°C
19:01:53: 240MHz 0.43 18%   9%   6%   0%   1%   0%   44°C
19:01:58: 240MHz 0.40 17%   9%   6%   0%   1%   0%   43°C
19:02:03: 1200MHz 0.37 17%   9%   5%   0%   1%   0%   48°C
Time        CPU    load %cpu %sys %usr %nice %io %irq   CPU
19:02:08: 1200MHz 0.34 17%   9%   6%   0%   1%   0%   49°C
19:02:13: 1200MHz 0.39 17%   9%   6%   0%   1%   0%   50°C
19:02:18: 1200MHz 0.44 18% 10%   6%   0%   1%   0%   50°C
19:02:23: 1200MHz 0.48 18% 10%   6%   0%   1%   0%   54°C
19:02:28: 1200MHz 0.53 18% 10%   6%   0%   1%   0%   56°C
19:02:34: 1200MHz 0.56 18% 10%   6%   0%   1%   0%   57°C
19:02:39: 1200MHz 0.60 19% 11%   6%   0%   1%   0%   58°C
19:02:44: 1200MHz 0.63 19% 11%   6%   0%   1%   0%   59°C
19:02:49: 1200MHz 0.66 19% 11%   6%   0%   1%   0%   60°C
19:02:54: 1200MHz 0.69 19% 11%   6%   0%   1%   0%   60°C

Orangepi v1.4 test: r9 removed, AVCC&RTC=2.90V about 40-50mA draw during test

Quote

Time        CPU    load %cpu %sys %usr %nice %io %irq   CPU
19:00:52: 1200MHz 1.56 39% 19% 12%   0%   7%   0% -10°C
19:00:57: 240MHz 1.51 38% 18% 12%   0%   7%   0% -11°C
19:01:02: 240MHz 1.39 35% 17% 11%   0%   6%   0% -17°C
19:01:07: 240MHz 1.28 32% 16% 10%   0%   6%   0% -18°C
19:01:12: 240MHz 1.18 30% 14%   9%   0%   5%   0% -18°C
19:01:18: 240MHz 1.08 28% 13%   9%   0%   5%   0% -18°C
19:01:23: 240MHz 1.00 26% 13%   8%   0%   4%   0% -18°C
19:01:28: 240MHz 0.92 25% 12%   8%   0%   4%   0% -18°C
19:01:33: 240MHz 0.84 24% 11%   7%   0%   4%   0% -18°C
19:01:38: 240MHz 0.78 22% 11%   7%   0%   4%   0% -18°C
19:01:44: 240MHz 0.71 21% 10%   7%   0%   3%   0% -18°C
19:01:49: 240MHz 0.74 20% 10%   6%   0%   3%   0% -24°C
19:01:54: 240MHz 0.76 19%   9%   6%   0%   3%   0% -24°C
19:01:59: 240MHz 0.70 19%   9%   6%   0%   3%   0% -24°C
Time        CPU    load %cpu %sys %usr %nice %io %irq   CPU
19:02:05: 1200MHz 0.64 19%   9%   6%   0%   3%   0% -19°C
19:02:10: 1200MHz 0.67 19%   9%   6%   0%   3%   0% -17°C
19:02:15: 1200MHz 0.70 19% 10%   6%   0%   3%   0% -17°C
19:02:20: 1200MHz 0.72 19% 10%   6%   0%   2%   0% -16°C
19:02:25: 1200MHz 0.82 20% 10%   6%   0%   2%   0% -13°C
19:02:30: 1200MHz 0.84 20% 10%   6%   0%   2%   0% -11°C
19:02:35: 1200MHz 0.86 20% 11%   6%   0%   2%   0% -10°C
19:02:40: 1200MHz 0.87 20% 11%   6%   0%   2%   0%   -9°C
19:02:45: 1200MHz 0.88 20% 11%   6%   0%   2%   0%   -7°C

Results: The internal temperature sensing cannot be trusted, especially since there is no voltage reference.

Tomorrow I will try to place a 150mA 3.3V LDO regulator instead of "r9" and see if I can find a way to actually test temperature.

PS: I would add images if I could figure out how to do so outside of hosting them somewhere else.

Sidenote:

"R9" is glued down, so to minimize chance of lifting pads when desoldering try this:

1- apply generous amounts of flux from flux pen.

2) use solder wick to remove as much solder as possible

c Once all the solder is gone, rotate the part 90 degrees with flat end needle nose pliers while slightly pushing into the board.

If you managed to get enough solder off, the part should break free without lifting the pads. And if they do lift, you can always solder onto the test points on the board.

Edited December 12, 2017 by konsgn

TonyMac32 · December 12, 2017

25 minutes ago, konsgn said:

I think the bigger issue between the revision 1.1 and 1.4 is that they removed the "U5" buck AVCC/RTC 3.3V converter, Instead in its place they put a "R9" 0 ohm resistor from the GPIO 3.3v regulator "u55".

Great observation.

chwe · December 12, 2017

30 minutes ago, konsgn said:

Results: The internal temperature sensing cannot be trusted, especially since there is no voltage reference.

Tomorrow I will try to place a 150mA 3.3V LDO regulator instead of "r9" and see if I can find a way to actually test temperature.

Great work! This might have a big impact on all H2+/H3 boards.

bozden · December 13, 2017

On 12/12/2017 at 4:11 AM, konsgn said:

Results: The internal temperature sensing cannot be trusted, especially since there is no voltage reference.

Great work... Good to have someone with the knowledge, skills and equipment.

Is there any possibility that you have calibrated equipment to measure the case temperature in each step? I'm pretty sure untempered v1.4 boards run quite a lot hotter (ref. finger test).

konsgn · December 14, 2017

22 hours ago, bozden said:

Great work... Good to have someone with the knowledge, skills and equipment.

Is there any possibility that you have calibrated equipment to measure the case temperature in each step? I'm pretty sure untempered v1.4 boards run quite a lot hotter (ref. finger test).

Unfortunately not, the closest I am to having a test setup is 10meters of un-joined type k thermocouple, and a thermocouple module. It's not a real set up. Waiting to get a hold of a Ir thermometer, or some other sort of temp sensor I can bond to the case / point at it with.

I now do agree that it feels hotter when the AVCC is run at 3.4V, and right now I'm debating wether a better fix for the rev1.4 would be to drop the general voltage of 3.4 down to 3.3 by tweaking/replacing the relevant resistors, or by breaking R9 and putting a 3v3 regulator to AVCC.

Janusz Kowalski · December 22, 2017

Is this high temperature issue also valid for the H3 boards or just H2?

raschid · December 22, 2017

This issue seems to specific to one specific revision of the Orange Pi Zero (Rev1.4).

konsgn · December 28, 2017

I managed to test the 1.4 version with the set up in the image attached.

I admit the ti sensor tag thermopile sensor is not at all a good way to test it.

The test I ran:

I let the device cool down completely between tests, then set the power supply to a selected voltage (for powering AVCC/RTC) and logged in. Then I waited for about 3 minutes while looking at the output of the armbianmonitor -m for the temperature to stabilize. Then I ran the following command:

sysbench --test=cpu --cpu-max-prime=20000 run --num-threads=$(grep -c '^processor' /proc/cpuinfo)

while watching the remote sensing temperature output from the sensor tag.

here is what I got:

Quote

@3.301 AVCC steady state:
19:26:57: 240MHz 0.02   2%   1%   0%   0%   0%   0%   48°C
41.8c external temp seen.

Maximum prime number checked in CPU test: 20000

Test execution summary:
    total time:                          202.0714s
    total number of events:              10000
    total time taken by event execution: 808.1024
    per-request statistics:
         min:                                 61.07ms
         avg:                                 80.81ms
         max:                                119.91ms
         approx. 95 percentile:              95.59ms

Threads fairness:
    events (avg/stddev):           2500.0000/1.87
    execution time (avg/stddev):   202.0256/0.04

maxed out at 54.2 c external temp

and :

Quote

@3.396v AVCC
steady state:
19:48:48: 240MHz 0.04   7%   4%   2%   0%   0%   0%   56°C
40.0 ext temp
...

Maximum prime number checked in CPU test: 20000

Test execution summary:
    total time:                          241.4225s
    total number of events:              10000
    total time taken by event execution: 965.4710
    per-request statistics:
         min:                                 61.07ms
         avg:                                 96.55ms
         max:                                148.29ms
         approx. 95 percentile:             113.54ms

Threads fairness:
    events (avg/stddev):           2500.0000/3.08
    execution time (avg/stddev):   241.3677/0.03

max temp 50.9 Celsius external temp.

....According to the results, it runs faster and hotter at the expected temp of 3.3V.

I only tested a few times, but I am now of the opinion that these chips will run hot one way or another.

For my application, I will be giving it a 3.3V regulator and a nice heatsink.

Further testing would probably benefit from a more directly attached temperature sensor/ more accurate thermopile & logging.

Rollmeister · January 7, 2018

Maybe try my blower fan, mosfet, 10kohm resistor and FanD software solution?

Cost me about 8 euro to do. Cheaper if you use a regular fan and have holes in the top of the case.

noxlin · January 11, 2018

I got opi zero H2+ 512Mb , v1.4

and install armbian 5.35 debian.

The board display -20 degree temperature in idle state.

i think maybe 20degree really.(adding 40)

i run a test using ffmpeg encoding about 2 hour。

the temperature about 30degree(really 70degree)

but the system is stable。

piknew · June 22, 2018

Hi, case is related to Opi Zero H2+:

Quote

root@PKOTHER:/var/log# uname -a
Linux PKOTHER 4.14.50-sunxi #1 SMP Mon Jun 18 23:35:22 +07 2018 armv7l GNU/Linux

I have done cpuburn test and some time after start it thrown following message:

Quote

root@PKOTHER:/var/log# cat kernel_2018-06-21.log
Jun 21 16:55:43 pkother kernel: thermal thermal_zone0: critical temperature reached (105 C), shutting down

Is there any throttling here?

enjoyneering · December 20, 2018

Hi,

Any plans to fix hi temperature issue after shutdown on ARMBIAN 5.65 stable Debian GNU/Linux 9 (stretch) 4.14.84-sunxi?

p.s. It's nice and cool on Debian GNU/Linux 8 (jessie)

WarHawk_AVG · December 21, 2018

Biggest solution for my OPiZero was limit max CPU to 816mhz, and minimum to 240mhz, then set conservative governor...

So far doing that..my OPiZero runs at 45°C, if processor kicks up w/ all cores heavily it gets to around 50°C and remains stable

A CPU heatsink and active cooling would definitely help reduce the CPU heat...mainly because even if a heatsink is placed on the processor...it ends up heat soaking it and the heat remains higher longer.

In fact my OPiPC rate limited has been holding solid for a long time as well...for some reason when processor usage spikes the processors just run hot as heck!

ridho13 · December 27, 2018

On 12/21/2018 at 2:06 PM, WarHawk_AVG said:

Biggest solution for my OPiZero was limit max CPU to 816mhz, and minimum to 240mhz, then set conservative governor...

So far doing that..my OPiZero runs at 45°C, if processor kicks up w/ all cores heavily it gets to around 50°C and remains stable

A CPU heatsink and active cooling would definitely help reduce the CPU heat...mainly because even if a heatsink is placed on the processor...it ends up heat soaking it and the heat remains higher longer.

In fact my OPiPC rate limited has been holding solid for a long time as well...for some reason when processor usage spikes the processors just run hot as heck!

could you tell me how to do that? i have looking how to do it but no success. thank you

guidol · December 27, 2018

7 hours ago, ridho13 said:

could you tell me how to do that? i have looking how to do it but no success. thank you

edit the file /etc/default/cpufrequtils

the save and reboot:

# WARNING: this file will be replaced on board support package (linux-root-...) upgrade
ENABLE=true
MIN_SPEED=240000
MAX_SPEED=816000
# GOVERNOR=ondemand
GOVERNOR=conservative

I got my OPi Zero in the original black cube case and the Zero is arounf 54-55 degree:

12:52:57:  240MHz  0.03   1%   1%   0%   0%   0%   0% 54.7°C  0/6
12:53:02:  240MHz  0.10   2%   1%   0%   0%   0%   0% 54.6°C  0/6
12:53:08:  240MHz  0.09   1%   1%   0%   0%   0%   0% 54.1°C  0/6
12:53:13:  240MHz  0.09   1%   0%   0%   0%   0%   0% 54.5°C  0/6
12:53:18:  240MHz  0.08   1%   1%   0%   0%   0%   0% 54.7°C  0/6
12:53:24:  240MHz  0.07   1%   1%   0%   0%   0%   0% 55.4°C  0/6
12:53:29:  240MHz  0.07   1%   0%   0%   0%   0%   0% 54.5°C  0/6

jstefanop · January 16, 2019

Is there anyway to disable GPU and lower DRAM frequency with mainline? Looks like overlays are the best way to do this? Haven't been able to find anything online though.

tbsky · September 12, 2019

how about the temperature of the new orange pi zero LTS?

orange pi zero ver 1.0 is cool. ver 1.4 and ver 1.5 are hot.

Lutz Fechner · October 14, 2019

My Zero ver. 1.5 in the black case with the expansion board attached reports 50°C right after booting and settles at around 46°C shortly after.

There is no heatsink on the SOC yet and load on the system is pretty low (only using it as OpenVPN Gateway).

I can live with that temperatures as they seem to be stable over time so far.

dolphs · November 29, 2019

On 9/24/2017 at 11:17 PM, prisma said:

I would recommend no one to buy a Zero currently.

How are we doing at this moment, referring to the newer OPi Zero LTS

It states " low running temperature and low power consumption. ", but ... ... ?

Anyway my goal is to run domoticz, which will be connected to my smart meter in the meter cupboard

Sign In

New OPi Zero - Yet another high temperature issue...

Question

Link to comment

Share on other sites

105 answers to this question

Recommended Posts

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Join the conversation

Important Information