Ethernet chip gets raging hot


87 posts   Page 1 of 4   1, 2, 3, 4
by UnaClocker » Tue May 15, 2012 4:39 am
My ethernet/usb chip seems to be pretty hot at all times, but when downloading at 10mbit or so off the internet, it'll get almost too hot to touch, and about a minute into the download, the whole pi locks up and becomes unresponsive. My first thought is to glue a ram chip heatsink to it, and maybe install a fan in my case (already got a 30mm 5v fan on order). Is there a setting I can change to help this problem?
I've seen mention of the ethernet being picky about the 5v, and it needing to be precisely 5v. That doesn't make sense to me, is this chip a 5v chip? Not a 3.3v one? Seems like it should be running off the regulated 3.3v power.
Posts: 12
Joined: Sat May 12, 2012 5:48 pm
by Gert van Loo » Tue May 15, 2012 1:50 pm
It might be warm, it should never be hot. Can you check if it also gets hot with no USB or Ethernet plugged in? In general: if you have trouble try as many (safe!) different operating modes as possible. From the total picture it is easier to have an idea what is wrong. (Not like the email I once got from the Far-east which had only the text: "It don't work")
User avatar
Posts: 2389
Joined: Tue Aug 02, 2011 7:27 am
by psergiu » Tue May 15, 2012 9:01 pm
Mine gets very hot too. Over 50 deg. C - burn sensation on the tip of the finger.

Compiling some software on a USB HDD. Swapfile configured on the same USB HDD.

4.95V between the test pins.

No lockups yet using the kernel from here: viewtopic.php?p=75889#p75889
User avatar
Posts: 222
Joined: Mon Nov 07, 2011 8:36 am
Location: TX, U.S.A. (was: RO, E.U.)
by mahjongg » Tue May 15, 2012 9:12 pm
according to its dataheet, the chip should consume no more than 750mW @ 3V3, that is just 230mA.
It shouldn't get more than lukewarm from that, and therefore also the 3V3 regulator should stay cold.

Something may be really wrong when the chip gets this hot, are you using an external powered hub? does the chip cool down when not connecting the hub?

is the 3V3 regulator also hot?
User avatar
Forum Moderator
Forum Moderator
Posts: 9059
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands
by psergiu » Tue May 15, 2012 9:40 pm
Port 1 - Powered USB 1.1 HUB (that also powers the Pi) - suposedly 2.3 Amp PSU
Port 2 - USB HDD with Y cable - additional power from iPad (2 Amp) PSU.
Eth - 2m cable to 100BaseTX switch

4.94V between TP1 & TP2

RG1 - cold
RG2 - warm-ish
RG3 - to small to feel anything

LAN/USB chip still hot after compiling done.
I'l try tomorrow various configurations and report.
User avatar
Posts: 222
Joined: Mon Nov 07, 2011 8:36 am
Location: TX, U.S.A. (was: RO, E.U.)
by mmcv2008 » Sun May 20, 2012 11:18 am
My LAN/USB Chip gets burning hot. Only LAN connected - but in the switch/router boxes this type of chips get burning hot as well, so I placed a isolated metal block (only thing I had at hand) on top of the LAN/USB-chip to cool it a bit down - works somehow. Would be of interest if this chip gets hot generally on all Raspberry PI's ???
Posts: 7
Joined: Sun May 20, 2012 10:58 am
by mahjongg » Sun May 20, 2012 1:03 pm
Well, these kid of chips do have a metal pad on their underside that should be soldered to the PCB, if for some reason that isn't done the chip cannot dissipate its heat away effectively to the PCB's ground copper plane, it might explain why its getting hot.

It should normally dissipate just 3/4 of a single watt, but one often forgets how small it really is. If its not able to effectively use the PCB as its "heat sink" I can imagine its getting a bit warmer than expected.

The chip is 3V3 powered, so if it draws an abnormal amount of heat (due to an internal defect somehow) then the 3V3 regulator it is feeding off from (RG2) should also get blazingly hot. That not being the case I can only conclude that the chip isn't drawing abnormal amounts of current.

Can't really decide from the descriptions if there really is something wrong, I don't own a R-PI myself, so I cannot compare if this is abnormal. Just a "touch test" isn't the best way to decide that. Even 45 degrees can be too hot to touch for too long if the material touched is very heat conductive, it doesn't mean that the chip is in any way in danger. Remember for the body everything above 39 degrees can be dangerous, so if the body senses a part of its body getting hotter than than that its sensed as pain. That is because proteins coagulate at 40 degrees Celsius or so, that is why getting a high fever is so dangerous. But in fact a normal chip will work well up-to 85 degrees.
So "getting too hot to touch" isn't a real sign of a defect.
User avatar
Forum Moderator
Forum Moderator
Posts: 9059
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands
by johnbeetem » Sun May 20, 2012 2:31 pm
mahjongg wrote:Well, these kind of chips do have a metal pad on their underside that should be soldered to the PCB, if for some reason that isn't done the chip cannot dissipate its heat away effectively to the PCB's ground copper plane, it might explain why its getting hot.


According to the LAN9512 data sheet, the ground pad is the only ground connection, so if it's not connected I can't see how the LAN9512 would work at all. OTOH, there might be enough solder to make decent electrical contact but not enough for good thermal contact. According to the data sheet the LAN9512 dissipates as much as 763 mW which is a lot for such a tiny package. I tried to find the thermal resistance of the QFN package at smsc.com but couldn't.
User avatar
Posts: 944
Joined: Mon Oct 17, 2011 11:18 pm
Location: The Coast
by mmcv2008 » Sun May 20, 2012 3:01 pm
Just fo avoid sudden death of the chip from getting overheated, I leave this metal block on it and as for the USB/LAN oscillator topic (http://www.raspberrypi.org/phpBB3/viewtopic.php?t=5766), I hope this is fixed on future devices.
Posts: 7
Joined: Sun May 20, 2012 10:58 am
by abishur » Sun May 20, 2012 4:49 pm
In a thread I made over here, I listed a heat sink and thermal tape combo that I used on the SoC and ethernet chips. It cost me around 8 bucks I think, but it gives me peace of mind for those hot little chips.
Dear forum: Play nice ;-)
User avatar
Posts: 4478
Joined: Thu Jul 28, 2011 4:10 am
Location: USA
by mahjongg » Sun May 20, 2012 11:38 pm
johnbeetem wrote: According to the LAN9512 data sheet, the ground pad is the only ground connection, so if it's not connected I can't see how the LAN9512 would work at all.


Ah yes indeed, good catch I had not realised that, but you are absolutely right in your assessment and conclusions.
User avatar
Forum Moderator
Forum Moderator
Posts: 9059
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands
by jbeale » Mon May 21, 2012 12:14 am
In case this is useful, from another thread. Personally, I'd say a 40 C plastic surface is very warm, but not quite too hot to touch. BTW, IR thermometers are as cheap as $25 now.

viewtopic.php?f=28&t=5900&start=25#p78973
Re: Temp Sensitive Boot Problem
Postby Jongoleur » 16 May 2012 15:09
In the interest of getting some basic temperature figures from a Pi that appears to be working in a sensible manner, I've waved an (admittedly fairly cheap and inaccurate) infra-red thermometer over the various components of a Pi that has been running for several hours. It has been rebooted several times during this period and restarted flawlessly.

CPU/RAM 36C/96.8F
LAN/Ethernet 40.9C/105.6F
Big voltage reg by large capacitor 38.3C/100.9F
SDHC card 40.8C/105.3F at connector end, 25.6C/77.3F at other end :-)
User avatar
Posts: 3226
Joined: Tue Nov 22, 2011 11:51 pm
by johnbeetem » Mon May 21, 2012 6:11 pm
jbeale wrote:In case this is useful, from another thread. Personally, I'd say a 40 C plastic surface is very warm, but not quite too hot to touch. BTW, IR thermometers are as cheap as $25 now.


I measure chip surface temperature in seconds -- specifically, how many seconds I can leave my finger tip on a component before I need to pull it away :-) I need to calibrate this some day.
User avatar
Posts: 944
Joined: Mon Oct 17, 2011 11:18 pm
Location: The Coast
by spurious » Mon May 21, 2012 7:05 pm
I've just measured my R-Pi temp, which has been on for 5 days
    CPU = 36.5C
    Eth Chip = 31.5C
    HDMI Port = 26.2C
    Eth Port = 23.4C
    USB Port = 25.6C
    SD Card = 34.3C
    PSU Port = 26.9C
    Ambient = 21.4C
Posts: 343
Joined: Mon Nov 21, 2011 9:29 pm
by johnbeetem » Mon May 21, 2012 7:30 pm
spurious wrote:I've just measured my R-Pi temp, which has been on for 5 days
    CPU = 36.5C
    Eth Chip = 31.5C
    HDMI Port = 26.2C
    Eth Port = 23.4C
    USB Port = 25.6C
    SD Card = 34.3C
    PSU Port = 26.9C
    Ambient = 21.4C

Can you try running a bunch of Ethernet traffic and see what happens to the USB/Ethernet chip? It may use more current when there's actual traffic, depending on software's use of shutdown modes.
User avatar
Posts: 944
Joined: Mon Oct 17, 2011 11:18 pm
Location: The Coast
by spurious » Mon May 21, 2012 8:22 pm
johnbeetem wrote:Can you try running a bunch of Ethernet traffic and see what happens to the USB/Ethernet chip? It may use more current when there's actual traffic, depending on software's use of shutdown modes.

Just ran a load test for 5 minutes on nginx / MySQL / PHP, from My PC
new figures (bear in mind the sun is no longer shining through the window onto the R-Pi):
    CPU = 32.3C
    Eth Chip = 28.1C
    Ambient = 21.0C
Posts: 343
Joined: Mon Nov 21, 2011 9:29 pm
by mmcv2008 » Sat May 26, 2012 11:23 am
I did buy now some heatsinks (http://www.zerotherm.net/eng/product/ZH100.asp on amazon for 6.90€ incl. shipping), normally for VGA boards, but two of them fit exactlly onto the CPU and the LAN chip and looking nice also.
Posts: 7
Joined: Sun May 20, 2012 10:58 am
by mahjongg » Sat May 26, 2012 11:56 am
With an ethernet chip that normally should be at something around 30 degrees Celsius you shouldn't have to put on a heatsink.
If your chip gets much hotter than that putting on a heatsink is like treating the symptoms of a disease, instead of curing the patient.
User avatar
Forum Moderator
Forum Moderator
Posts: 9059
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands
by mmcv2008 » Sat May 26, 2012 3:37 pm
It is at peak a bit more than 30 degree, but the heatsink looks cool now - I will post a picture later :mrgreen:
Posts: 7
Joined: Sun May 20, 2012 10:58 am
by jbeale » Sat May 26, 2012 3:49 pm
CPU = 32.3C
Eth Chip = 28.1C
Ambient = 21.0C


Something seems very odd here. If that's the normal temperature of the chips when they are working hard, there is no way the Eth chip should be "burning hot" unless the input voltage is way out of spec, or there is an internal or external short-circuit somewhere. You aren't connecting to a PoE port by any chance (power over ethernet)? -and even that, in theory wouldn't be a problem unless the "missing magnetics ethernet jack" problem has returned.

I'd really like to hear from one of the R-Pi hardware engineers on this, because I think something is very wrong.
User avatar
Posts: 3226
Joined: Tue Nov 22, 2011 11:51 pm
by mmcv2008 » Sat May 26, 2012 4:30 pm
Here the picture
Attachments
IMAG0124.jpg
IMAG0124.jpg (47.62 KiB) Viewed 4994 times
Posts: 7
Joined: Sun May 20, 2012 10:58 am
by Gert van Loo » Sat May 26, 2012 10:36 pm
I can imagine that you do not want to give up your Pi now.
But wen a few more are available I would really send it back so somebody
can find out why it gets so hot.
User avatar
Posts: 2389
Joined: Tue Aug 02, 2011 7:27 am
by simonbr » Sat May 26, 2012 11:02 pm
johnbeetem wrote: I tried to find the thermal resistance of the QFN package at smsc.com but couldn't.


I found a datasheet of another component with the QFN64 package that says 20 degC/W typical without airflow. This assumes the use of 49 (7*7) thermal vias of 0.3mm diameter. At 3/4W the die temperature would rise 15 degC above ambient temperature.

But looking at the board there are at most some 25 thermal vias under the chip and they look rather thin. So on the Pi the actual thermal resistance, and the temperature rise, may be considerably higher.

Best regards
Simon
Posts: 23
Joined: Sun May 13, 2012 5:02 pm
by dics » Sun May 27, 2012 12:29 am
My Pi stayed on for three days with apache/php/mysql/vsftpd running.
No problems whatsoever until I tried to transfer a 2GB file using FTP. After one hour it spit out a kernel panic; ethernet chip was quite hot to touch
I only have one board and it just arrived a few days ago but since I am not the only one complaining about heat I am wondering if it's not actually a problem with the last batch of RPis.
Posts: 14
Joined: Mon Feb 06, 2012 3:54 pm
by mahjongg » Sun May 27, 2012 2:06 am
I have to stress that the talk about "an ethernet chip" is wrong!
There is no "ethernet chip" on the RPI,

the chip mentioned isn't an ethernet chip, its a three port HUB chip, which on one of its ports it has a built in USB ethernet adapter.

So when one says that the "ethernet chip is hot", what is actually meant is that the "USB hub, annex USB ethernet adapter" chip is hot, so it could be equally hot because its USB hub part is hot as it could be hot because its ethernet part is hot.

The chip could get hot because of something to do with USB communication, instead of ethernet communication.
User avatar
Forum Moderator
Forum Moderator
Posts: 9059
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands