Experimental Thermal Characteristics of C1

Post Reply
User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Experimental Thermal Characteristics of C1

Post by odroid » Tue Feb 03, 2015 12:36 pm

Experimental Thermal Characteristics of C1

Goshener raised a serious issue of Over-heating.
http://forum.odroid.com/viewtopic.php?f=115&t=7772
So we performed some evaluations with a few C1 boards as below.

1. ODROID-C1 as it is
We ran the Kodi with 1080p/H.264 BBB test video file for 1 hour and kept checking the CPU temperature. We didn’t apply the passive cooler nor plastic case. The ambient room temperature was around 22C. The CPU temperature could go up to 68C.

2. ODROID-C1 with original case
We put the board into the plastic case and ran the same test. It could reach to 82C after 30minutes test. It means our plastic case might cause a serious problem of ventilation.

3. ODROID-C1 with original case in the chamber (35C)
We installed the board in the chamber and it could reach to 90C after 1 hour test. Yes, I could reproduce the similar symptom.

4. ODROID-C1 with original case and the heatsink in the chamber
And we installed our official heatsink on the board with a grey color thermal pad(TIM).
The maximum temperature is around 83C.

5. ODROID-C1 with original case and the heatsink in our office
The temperature is less than 70C at 20C of ambient temperature on my desk.


Our plastic case has not enough hole/slot to make a good air-flow and it is one of the main root causes of over-heating.
We recommend you to use a heatsink when you put the board in the case.
Don't forget to add the thermal pad(TIM) between board and heatsink otherwise there will be no cooling effect.

Sorry about that we didn't perform the full thermal test with the official case.

Note that the CPU intensive loads don't make any serious heat at 1.5Ghz.
The GPU/VPU intensive multimedia playback makes more heat.


Special thanks to Goshener who raised this issue seriously.

mlinuxguy
Posts: 842
Joined: Thu Feb 28, 2013 10:28 am
languages_spoken: english
ODROIDs: X, X2, XU, XU3, XU4, C1, C1+, C2, N1, USB-IO
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by mlinuxguy » Tue Feb 03, 2015 12:44 pm

In my overclocking tests I used my arctic silver thermal paste compound (what I used to use when overclocking my PC's)
I ordered some of this stuff off Ebay from an Asian supplier:

Code: Select all

5 pcs Adhesive Thermal Grease Paste Compound Silicone for LED CPU IC Heat sink
thermal conductivity 0.671W / m.k
Note: This item is Adhesive Solder,while the Adhesive solder vulcanizing, the heat sink and part can not be separated!
After checking data sheets that would not be a good option over thermal grease or my arctic silver
Arctic silver: independent tests at 0.94 W/(m·K)
My preference would be to use an electrically insulator one like: Céramique

beatmag
Posts: 28
Joined: Sat Jan 03, 2015 9:36 pm
languages_spoken: english
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by beatmag » Tue Feb 03, 2015 8:38 pm

What is the maximum temperature for C1??

When overheating will the CPU clock down...........or will the board/cpu start to burn?

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Tue Feb 03, 2015 8:49 pm

Max clock is changed to 1.2Ghz from 1.5Ghz when the temperature is higher than 85C.
But it seems not to be enough in the edge condition.
We are considering 700~900Mhz now. But we need to do further test.

dnom
Posts: 18
Joined: Sat Jan 31, 2015 1:00 am
languages_spoken: english
ODROIDs: C1, C2, C3, C4 ...
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by dnom » Tue Feb 03, 2015 9:21 pm

odroid wrote: clock is changed to 1.2Ghz from 1.5Ghz when the temperature is higher than 85C.
But it seems not to be enough in the edge condition.
We are considering 700~900Mhz now. But we need to do further test.
But this won't do much if the heat cause is the GPU/VPU. I've never reached temp over 65 C when I did CPU benchmarks for hours without connecting it to a display but I easily reached 90C just by playing a video on kodi and nothing else.

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Tue Feb 03, 2015 10:30 pm

Yes.. good pointing. :oops:

mdrjr
Site Admin
Posts: 11753
Joined: Fri Feb 22, 2013 11:34 pm
languages_spoken: english, portuguese
ODROIDs: -
Location: Brazil
Has thanked: 1 time
Been thanked: 25 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by mdrjr » Wed Feb 04, 2015 2:29 pm

I tried to burn a C1 for the sake of tests..
Even putting my soldering iron on top of the SoC and leaving it there.. the best reaction I got was..
When the temperature reached 110C it shutdowns gracefully. No issue.

Did this test several times and the board is rock solid.

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Wed Feb 04, 2015 4:26 pm

Yes! Over-temperature protection is working well.
Thank you for the confirmation with the 300C of soldering iron. :)

Apokalypz
Posts: 178
Joined: Thu Dec 26, 2013 12:58 pm
languages_spoken: english
ODROIDs: • Odroid -XU
• Odroid-C1
Location: USA
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by Apokalypz » Wed Feb 04, 2015 9:40 pm

mdrjr wrote:I tried to burn a C1 for the sake of tests..
Even putting my soldering iron on top of the SoC and leaving it there.. the best reaction I got was..
When the temperature reached 110C it shutdowns gracefully. No issue.

Did this test several times and the board is rock solid.
Have you tried microwaving it? I'm sure that'd burn it...I need to be sure this thing can hold up against a nuclear attack or the sun exploding. J/K

mdrjr
Site Admin
Posts: 11753
Joined: Fri Feb 22, 2013 11:34 pm
languages_spoken: english, portuguese
ODROIDs: -
Location: Brazil
Has thanked: 1 time
Been thanked: 25 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by mdrjr » Wed Feb 04, 2015 9:43 pm

Sec.. calling my friend to get some Plutonium ....

mAineAc
Posts: 28
Joined: Tue Feb 03, 2015 6:46 am
languages_spoken: english
ODROIDs: C1. XU3
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by mAineAc » Fri Feb 06, 2015 12:12 pm

OK, I just ordered a case for mine. Maybe this was a mistake? Is there a heat sink with a fan option for the c1? This doesn't sound good.

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Fri Feb 06, 2015 12:15 pm

According to our test result, the heat-sink should be enough if the TIM PAD is properly installed.
http://www.hardkernel.com/main/products ... 1637555371

gdstew
Posts: 4
Joined: Tue Feb 17, 2015 2:40 pm
languages_spoken: english
ODROIDs: C1
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by gdstew » Tue Feb 17, 2015 3:21 pm

I have a question about the heat sink and the thermal pad(s). Specifically about the installation instructions that says that there are two thermal pads being used.
One that is already on the heat sink, and the second is the one applied to the CPU/memories before installing the heat sink. Is this correct? Using two thermal
pads increases the thermal resistance from the CPU/memories to the heat sink. Am I missing something ?

I wanted to buy a case but now I'm not so sure. From what I'm reading in this forum it looks like active cooling will be required to go 1.5 GHz in a case.

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Tue Feb 17, 2015 3:43 pm

Yes, the double layer of thermal materials can increase the thermal resistance.
But it is not too bad to trigger the thermal throttling because the thermal throttling threshold level is 85~86C.

rflownn
Posts: 28
Joined: Sun Jan 04, 2015 12:47 pm
languages_spoken: english
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by rflownn » Tue Feb 17, 2015 5:05 pm

Any chance we can see thermal gradient pictures from thermal IR cameras? I run some of my odroid boards really hot for several hours and would like to know what to expect. I'd do those but I don't have a thermal IR camera.

edit: Thermal movie would be best :) Then we can do some basic measurements for heat transfers.

gdstew
Posts: 4
Joined: Tue Feb 17, 2015 2:40 pm
languages_spoken: english
ODROIDs: C1
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by gdstew » Tue Feb 17, 2015 11:33 pm

OK, let me put it this way. Is there a reason for the two layers? It's not can increase thermal resistance it is will increase thermal resistance.
There is no advantage to using two thermal pads and unless there is some kind of physical reason for it (needed to provide good contact pressure?) I don't
understand why the second pad is being used.

I'm seeing 80+ degrees C is several of these posts and that is much hotter than my overclocked 3.8 GHz quad-core AMD Phenom II and it is at the very
edge of industrial grade ICs (85 deg. C) and well above the rating for commercial grade ICs (70 deg. C). Is the Amlogic S805 SoC rated for industrial use?

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Wed Feb 18, 2015 12:12 am

The thin white layer should be enough if the PCB surface was even.
But height of some passive components is slightly taller than CPU and DRAM.
It breaks the conjunction between the CPU and the heatsink.
So we needed to add another layer.
I think you can remove the white layer, if you can add a non-conductive material to protect the short circuit by metallic heatsink.

User avatar
rooted
Posts: 7741
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 718 times
Been thanked: 206 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by rooted » Wed Feb 18, 2015 12:43 am

Would 3 small heatsinks attached provide more cooling? I doubt it as the large heatsink gets rather hot.

The hottest I have seen is 72C and thats inside the case. With kodi running my idle temperature is 55C but it's only 17C in the room.

gdstew
Posts: 4
Joined: Tue Feb 17, 2015 2:40 pm
languages_spoken: english
ODROIDs: C1
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by gdstew » Wed Feb 18, 2015 1:15 am

Thanks for the info. I figured there was a good reason, it looks like X3 and U17 are the main culprits. I just wanted to make sure I knew what is was so I could come up
with an alternative if needed. It would be interesting to see what difference a thin 30mm or 40mm fan mounted on the heat sink would make. I was waiting to find out about
the two pads question before purchasing a heat sink so I may try the fan solution after I purchase one. Using three heat sinks as suggested by rooted is another possibility
although it looks like two would also work, one for the AM S805 and one for the memories.

Vort
Posts: 153
Joined: Fri Nov 22, 2013 10:04 pm
languages_spoken: english
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by Vort » Wed Feb 18, 2015 11:44 pm

gdstew wrote:..., it looks like X3 and U17 are the main culprits.
What if you drill holes in the heatsink at this places to avoid contact with them?

gdstew
Posts: 4
Joined: Tue Feb 17, 2015 2:40 pm
languages_spoken: english
ODROIDs: C1
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by gdstew » Thu Feb 19, 2015 12:32 am

There are several more parts on the periphery of the heat sink that might also be a problem but yes that would work. Another thing
I'm looking at is I have a left over graphics card memory heat sink that I could cut it up and use one part for the S805 and the other
for both of the memories using some Arctic Silver thermally conductive epoxy I have to mount them.

User avatar
dtruesdale
Posts: 83
Joined: Sat Jan 10, 2015 3:55 pm
languages_spoken: english
ODROIDs: C1 x's 5
C1 Tinkering Kit
XU3-Lite x's 3
Oshow2
WeatherBoard
C1 3.2 inch Touchscreen Display Shield
(Thinking of purchasing a VU)
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by dtruesdale » Sun Apr 05, 2015 1:59 am

In my testing of the odroid with standard heatsink and case I have found that we need a heatsink with a fan on it. With a fan in the case I can keep the C1 at about 48C-53C with 2 usb cameras on it and it running zoneminder. Without it the fan and one usb camera using motion to send video to the other C1 using zoneminder it runs at about 59C-65C.
May not seem like a huge difference but it is when you consider the one with the fan has 2 usb cameras and is the zoneminder server.

A simple heatsink like we have for the XU3 made for the C1 solution made simple. If it could be controlled it would be even better but just the heatsink with the fan I would be greatful since in my testing these boxes could do even more.

Metropolis
Posts: 35
Joined: Thu Jan 08, 2015 12:17 am
languages_spoken: english
ODROIDs: C1
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by Metropolis » Thu Apr 16, 2015 2:46 pm

Did your C1 ever reach 85C during your tests without a fan?

User avatar
odroid
Site Admin
Posts: 34542
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean, Japanese
ODROIDs: ODROID
Has thanked: 807 times
Been thanked: 696 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by odroid » Thu Apr 16, 2015 4:15 pm

When I put the C1 in the plastic case while running some heavy video files on Kodi, the CPU temperature could be near 85C.
If I added a passive coolder(heat sink), it was less than 70C.
I belive you don't need any fan (active cooler).

/tmp
Posts: 21
Joined: Fri Mar 13, 2015 9:42 am
languages_spoken: English, Deutsch,
ODROIDs: C1
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by /tmp » Mon Apr 27, 2015 4:01 pm

I would suggest a copper heatsink to further decrease temps as copper
has shorter heat dissipation as long as more even heat convexion.

tuxedo_herbert
Posts: 10
Joined: Tue Apr 28, 2015 3:38 am
languages_spoken: german
ODROIDs: U3, (RasPi2)
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by tuxedo_herbert » Wed Apr 29, 2015 12:55 am

Hi, any supplier in germany that you can suggest for getting passiv-coolers? not an onlinestore, i mean a manufacturer. we're interested in higher amount.

thanks a lot.

patoberli
Posts: 403
Joined: Mon Feb 23, 2015 6:32 pm
languages_spoken: english, german
ODROIDs: C1 with RTC, Bluetooth 4.0, Wireless stick 4.0
Location: Switzerland
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by patoberli » Wed Apr 29, 2015 5:58 am

Try https://www.wlw.de/treffer/aluminiumkuehler.html maybe one of those can produce something in the right size. Although it's probably MUCH cheaper to buy it in China, www.alibaba.com is fairly useful for that, but beware of fraud there.
-
pato

Shonk
Posts: 29
Joined: Tue May 05, 2015 12:04 am
languages_spoken: english
Has thanked: 0
Been thanked: 0
Contact:

Re: Experimental Thermal Characteristics of C1

Post by Shonk » Wed May 06, 2015 6:59 pm

Mine has been playing a blu-ray mkv for the last 2 hours max temp 55

C1 in case with heatsink and vent area cut out of case running at 1.73 ghz
dropped to 52 about 2 mins after it finished and its going down slowly

its also in a tight space only 2cm above it is glass
so i dont think heat is an issue if you remove the vent area plastic

http://s23.postimg.org/3p0yzz4ln/image.jpg

crazyquark
Posts: 336
Joined: Thu Jan 15, 2015 4:22 pm
languages_spoken: english, french, romanian
ODROIDs: C1,C1+,C2,HC1,XU4
Has thanked: 3 times
Been thanked: 16 times
Contact:

Re: Experimental Thermal Characteristics of C1

Post by crazyquark » Sun May 10, 2015 2:14 pm

I recently also realized that the C1 is overheating and throttling the CPU speed down(while watching movies with Kodi). So I applied a lamer fix, I used one of those small and quiet USB fans to blow air directly onto it.
The C1 feels much better and is always cool. I am talking about this:
http://www.arctic.ac/eu_en/arctic-breeze-mobile.html

Post Reply

Return to “Hardware and peripherals”

Who is online

Users browsing this forum: No registered users and 3 guests