MacBookPro 15" 1.1: Strange GPU temp sensor behaviour

Hello to all,
Since I let my trusty MBP alone for a week two months ago some real strange behaviour occured.
I left my MBP in standby as usual. When I came back from my trip, the battery was completely discharged and I had to connect the power adapter. MBP awoke from suspend to disk (grey screen, progress bar) and a few seconds after beeing ready for work, fans started to rev up. 5 min later the fans were running at full speed.
Puzzled, I did a restart. Same behaviour, high RPM right after boot logo, full 6000rpm after a few minutes.
When I took a look at iStatMenus I saw that the GPU Heatsink Sensor Temperature was around 35 degrees Celsius (around 70 Fahrenheit) higher than the GPU Temp sensor itself. It was the highest readout by all sensors, which makes no sense at all.
I tried: PRAM reset, SMC reset, PMU reset. Apple Hardware Diagnostic CD by me and a German ASP, and by a Genius Bar expert. No errors shown.
But over the two months the difference between the values sank to 18-20deg Celsius, and I found a tool to let only the CPU sensor control the fan, so I was happy again.
Yesterday I came back from another trip, guess what: battery empty, woke up from standby..... the difference is now around 36 deg Celsius, and the MBP shuts down automatically even on the smallest YouTube clip when GPU Heatsink Temp exceeds around 86deg Celcius.
I am clueless, everybody around me too. So maybe one of you has an idea?
Regards, Matthias

Not sure about sensors adapting to reality by themselves. This sounds like very strange behavior. I wonder if there could be something like a loose connection somewhere.
Did the computer misbehave in front of the genius? If it didn't, it might be worth while seeing if you can determine the exact circumstances under which it will act up so you can make it act up in front of the genius.
For example, my iBook used to freeze, but only when warmed up. I went to the genius bar early so I could give it time to warm up so that it would be sure to malfunction in front of them. This is especially important to do if you are still under warranty. It's important to establish that a problem first occurred under warranty, especially if your warranty is close to running out.
Good luck!

Similar Messages

GPU 0 Die Temp Sensor ERROR on a MacBook Pro. Can anybody help?

Hello.
My wife has a 15-inch, 2.4GHz MacBook Pro model 3.1 laptop.
All of sudden the screen started to flicker, and after that, it could no longer be booted. Applelogo visible but the timewheel stops and nothing happens.
I managed to get hold of a copy of ASD (Apple Service Diagnostic) for her computer, and I found out what was wrong:
There is a test called "Temp Sensor - GPU 0 Die Temp" that fails. I get the following message:
"Temp sensor GPU 0 Die Temp test 1 - Sensor Reads Within Operating Range
- Check to ensure that sensor reads within operating range.
ERROR - Unable to read sensor
- TEST FAILED -
Temp sensor GPU 0 Die Temp test 8 - Sensor Noise Test
- Checks that sensor Noise is within spec.
ERROR - Unable to read sensor
- TEST FAILED -"
If I look in the "Hardware Profile" under "sensor" I find the following:
"Sensor
type:            Temperature
location:            TG0D
low-limit:            15
high-limit:        85
reading:            1.0
description:        GPU 0 Heatsink Temp
noise-tolerance:    1.0"
The thing is that the problem seems to be intermittent. It comes and goes. All of a sudden the test passes and I can boot the computer. Some days the computer can work for a whole day, so it doesn´t seem to be related to temperature.
What do you guys and girls think? Is this fixable by myself? Does anybody know where this sensor is located? Maybe it´s just a bad connection, since it´s intermittent?
I feel I have come a little bit on the way to be able to solve this issue, and it would feel like a failure to be forced to leave the computer at an Apple Service Center, if it´s even worth it to do that. There is a risk that the computer is not worth a repair.
Could someone please chime in here for a clever advice?
Thanks.
Bo Reimer.

Welcome Bo!
It may be related to a known problem with MBPs 3,1 and 4,1. Some had a bad Nvidea Geforce 8600M GT video chipset that could fail under high heat cpnditions. There is an Apple Repair Extension Program (REP) that will fix the defect at no charge for up to four years from date of purchase:
http://support.apple.com/kb/TS2377
You'll need to have an Apple retail store or an authorized service provider check the computer with a special test gadget for positive ID of the problem to qualify. If it does, you get a new logic board. Reported turnaround for the fix is from same day to three days.
If their test does not turn up the Nvideo curse, you still have an affordable repair option if you live in North America (It would he helpful if you fill in the "Geographic Information" field in your profile--at this point I'm having to assume you're in N. America)).
The option is a depot repair. Apple sends your computer to their central refurbishment facility and bascially fixes anything outside of cosmetic damage for a flat fee of about US$315--much cheaper that a new computer. Taxes and shipping may apply.
People who have used this option are reporting excellent results and a turnaround of about one business week.

MSI 6600GT Temp sensor install "how to" ?

Got all my new Thermaltake fans in today. I now have a ton of unused temp sensor probes. Was wondering if there's a "How To" out there somewhere on how to properly install one of these things on my Vid card.The fan controller I got handles 4 fans and will monitor 4 different temps.

Quote from: Trailryder on 17-February-05, 01:44:04
I've also read that the temp shown by the card bios is way off and probably why nvidia didn't enable it.
It is MSI that did not enable it not nvidia.
Because there are a lot of other 6600GT's that have it enabled by default. MSI left it disabled.
http://www.firingsquad.com/hardware/nvidia_geforce_6600_gt_agp_roundup/page19.asp
The 6600GT GPU is the same on all the cards, so I think it is accurate. For some strange reason MSI decided not to enable it, maybe they thought no one would care.
If you could do the flash with no risk I say go for it. There is a risk though. You can pretty much minimize that risk as long as you have a cheap PCI video card to start the computer with if something goes wrong, and you have your bios set to start PCI VGA first. I couldn't save my video card for two reasons. I don't have a cheap PCI graphic card, and my bios was set to start from AGP first. I never expected my computer to not start from the bad flash. I thought it would start, and I could at least get to the floppy drive to restore a backup which it wouldn't do.
Now that I have the probe attached to mine, I see no reason to flash the bios. It gives me real time temperature, and I don't have to start any software to see what the temperature is on the card.
If you already have a probe, and you don't have to buy anything then I would use the probe. It has to be a very thin probe to fit between the heatsink and on the side of the GPU. It can't be any higher then the GPU. You should also use some nail clippers and clip as close as possible all aorund the tip. You also want to put a small drop of themal compound on the tip that touches the side of the GPU. I use double sided tape to stick mine to the board and to the area around the GPU so it doesn't move.

IMac iSight GPU Temp?

Hi,
I just got my iMac iSight a couple days ago, and i was monitoring the temp of the iMac with Temperature Monitor when i noticed that sometimes my GPU Temp will go up to around 96°C/205°F ... is that normal? i dont want to burn up my graphic card... thanks in advance for any help on this..

Valhalla:
Are you referring to you CPU temperature or the GPU because there are two GPU temperature sensors. I assume you mean the CPU. I don't care what you were told. To me 205 is too hot. Noone else has reported a temperature that high. In fact, I believe the earlier versions shut down around 180+. Why should this one be able to go up to over 200?
I would call help desk again find out what is the accepted range, tell them you think it is running to hot and a new computer or any computer should run that hot. Better yet, find out at what temperature the computer shuts down. Get a case number and make sure you tell the individual that you fear it will burn up your graphic card.
Or tell Apple you are returning the computer. I would be very tempted to return it and don't fret about how long it would take to get another.

HD7950 temp sensor always zero + auto fan always 100%

After installing new card and latest CCC software and drivers as soon as windows starts, the card fan goes to 100% even at idle. In Performance section of CCC GPU temp reads 0c at all times even under load.
Tried uninstall CCC, remove card, re-install card and then CCC, no dice... same problem. I can manually adjust the fan speed in CCC but it's a PITA and shouldn't be necessary.
Do I have a faulty temp sensor perhaps?
Thanks....

Quote from: flobelix on 26-November-12, 13:46:47
Obviously either driver isn't correctly installed or it's indeed a sensor misreading caused by a defect.
Remove current driver in safe mode using>>Driver Fusion<<. Reboot and install latest MSI driver:
Thanks for the suggestion. Gave it a whirl but still the same deal. Temp 0c - Power 0% - Activity 0% - Fan 100% from the moment Win7 starts no matter what the card's doing.
Only thing I didn't follow to a T from syour suggestion is that the free version of Driver Fusion said it required upgrading to Pro to remove all driver items even in Safe Mode.

Msi Z68 GD80 (G3) strange Cpu Temp.reading

Hello all,this is my first post here.
I just purchased an Msi Z68 GD80(G3) and I think I have strange cpu temp. reading as seen in the Hw monitor.My i5 2500k idle/load CPU temp. is equal or higher than core temps which is strange for an intel cpu,as an intel user for years I believe it should be 10-12C lower than core temps.(my previous mobo was giga p67 ud5 and never had this issue).Double checked my cpu cooler-spire thermax 2-no problems.Updated to last bios-no change-Any recommendations/solutions welcomed here.
Rest of my system as follows:
Asus gtx560
4*2 gb gskill ddr3 ram
1tb samsung hd
high power 650w power supply

Have you checked what the CPU temp shows in BIOS just after start up ? How does that compare with the HW temp.
HW Monitor's temp interpretation may just not be calibrated for the specific Fintek chip used on your MB. Maybe you want to check with the HW monitor home site as well.
How do the temps compare with whats read/displayed in Speedfan, Real Temp or Core Temp ?
If they are all similar, then it could just be a badly calibrated sensor on the board as well.
Do some comparative testing and see if there are any differences at all.
Also check with MSI technical to see if they are aware of a general tendency with regards to this, or whether a BIOS revision may have addressed such an issue. >>How to contact MSI.<<
Just recheck your heatsink seating again and maybe even reseat the CPU and do a full socket pin inspection at the same time.

Getting the GPU temp on a Radeon Card, and is a Radeon X850 PRO 256MB PC card flashable for use in Mac G5?

I decided to post this as a new topic since it really didn't belong on the tail end of my X800 XT thread of woe...
Is there any program that can get the GPU temp of a Radeon video card on a Mac? I have done quite a bit of looking online and have come up with nothing. Strange, since this is usually a concern for all computer owners.. I don't even care if it's unstable, beta, hack city.
The second part of this question is: I have a PC AGP video card "Radeon X850 PRO 256MB", and I wanted to know if it is flashable to use in a Powermac G5.
The card looks a lot like the X800 XT, except for the chip on the underside of the card that says "ATI Rage Theater" on it.
Here is a link to a pic of the underside of the card that I found on PhotoBucket:
http://media.photobucket.com/image/recent/dualamdmp/DSCF6296.jpg
Here is the top side:
http://s1238.photobucket.com/albums/ff489/dualamdmp/?action=view&current=DSCF629 2.jpg
In case my efforts completely strike out getting a video card with an ADC port (basically the X800 XT or X850 XT), I am starting to look for a reasonably priced DVI to ADC converter.
Thanks!

Is there any program that can get the GPU temp of a Radeon video card on a Mac?
No.
Radeon X850 PRO 256MB...I wanted to know if it is flashable to use in a Powermac G5.
Possibly. BUT, it won't output digital, which is what you need to use a DVI to ADC adapter.
All flashed Radeon X800/X850 will only output analog.
A flashed FireGL X3 , using a Mac Radeon X800XT ROM will output both ports digital, and one of those is also dual link DVI (in case one wants a 30" Cinema Display).
That's the only flashed X800 that outputs digital.
If you're going the adapter route, consider an OEM or a flashed Geforce 6800 GT or Ultra.
The Ultra actually out performs the X800XT in some areas, with the GT not far behind.
A flashed Geforce 6800 GT can be a real performance value, often times a bargain.

Difficulty with CJC using on board temp sensor - SCB-68

I am trying to do Cold Junction Compensation using the built in temperature sensor on my SCB-68 connector block. I am using the 6224 M Series PCI DAQ card. I am having two problems:
1. I found a procedure on this forum for reading the temperature sensor which, in my case, is at channel 16. So, I have my connector block switches set so the Temp sensor is enabled, and I am in differential mode. I created a global channel called "TempSensor" and set it to channel 16. I applied a scaling of m = 180 and b = 32 to get farenheit. When I do this, and I test the channel, it reads a temperature of 75 degrees. This seems a little hot for my office, which is probably more like 68 degrees. Any idea what might be wrong? I don't have to connect anything to the board to read the temp sensor, do I?
2. I've connected a type T thermocouple to channels 20 and 28 (differential). For CJC, I can set it manually by estimating the temperature in my office as 68 degrees and puting that value in the CJC Source Constant field. When I do this, the TC seems to work great. It reads around 68 degrees in my office, and more like 85 if I squeeze it between my hands. So, now I want to use the "TempSensor" global channel to do my cold junction compensation automatically. When I select my CJC Source as "Channel" and enter "TempSensor," I run the test and get temperatures of 176 degees F. What am I doing wrong?
Thanks for any help!
Kevin

Dear Kevin,
There is a really easy way to test your thermocouple using the Built-In CJC of the SCB-68. The only time I set up a global channel for the CJC is to make sure that it is working correctly. The CJC temp does not indicate the temperature of your office, it indicates the temperature of the SCB-68 board. My CJC value is 2 degrees warmer than the room temperature. 75 degrees is a reasonable temperature to get.
After you check the CJC to make sure that it is working, you can set up the task in the Measurement and Automation Explorer (MAX). Expand Devices and Interfaces>>NI-DAQmx Devices. Right click on your device and select Properties. Click on the Accessory Tab and add the SCB-68 to the second connector and click Configure. Check the box for Reference Temperature Enabled. Click the two OKs to save your changes. Then create your thermocouple task normally and select your CJC source as Built-In.
Now, it will detect temperature changes that you add. Please let me know if you have any questions. Have a great day!
Sincerely,
Marni S.

Where to place external temp sensor?

I have a MSI 845 Ultra (6398) MB with a PIV 2Ghz processor and I have just changed case to a ThermalTake XaserII 6000. The case came with an external temp sensor that "can be placed anywhere". The reason I changed case was that my temp readings was very high (CPU Idle 50C, stressed 72C, MB 40C idle, 46C stressed, HD0 62C, HD1 48C).
The XaserII 6000 comes with 5 fans and I now have the following readings: CPU 45-50C idle, 58-62C stressed, MB 31C idle, 35C stressed, HD0 35C and HD1 28C.
I managed to get the temps down substantially with the new board but I still think that the reading for the CPU is a bit high and since I understand that there are issues with MSI temp readings I thought that I would use the external temp sensor with the CPU and see what happens. The questions is where to I put the sensor (under the processor, between the processor and the heatsink (bad idea?), attached to the heatsink) to get the most accurate/true reading?
The XaserII lets me control the CPU speed (high, medium, low) and the question is shall I do this manually or shall I leave the CPU speed control to the board or to a program like SpeedFan?
Any input is higly appreciated.
Hans

GAWD!! I purchased Yamaha MOTIF samples in EXS format and I have spent TWO hours trying to figure out how, where, what, etc, etc... to install these to use with LOGIC Express. I've searched the net, search apple, etc... still, I've yet to hear a BLEEP outta LOGIC. I get asked for permission to drag folders, type permission, but nothing will go where I'm thinking it should go....
God only knows how much I miss good ole OS9.......

Where are the K7N2 Delta temps sensors located?

Where are the K7N2 Delta temps sensors located?
Okay I know one reads from the diode... altough the reading still seems to be low compared to other diode readers (like the 8K3A+ which gives you readins in the 50's).
Where is the other one located?

MSI Tech... how about it?

Were is the temp sensor on a k7t pro MS-6330

i have a k7t pro MS-6330 and would like to know were to "system" or "case" temp sensor is as i wana make sure it is not reading falesly (hiden behind a ribon cable etc (may be the msi H/G coul help)
cos i needed to know the anserw
( i've looked for it but dont even know what to look for ? )
please help someone!

An Athlon XP could run in your board but it would only run at about 75% of it's clock speed due to the fact your board doesn't support 266MHz FSB.
Officially you should only use a Duron1300 or old Athlon 1400/200 if you could still find one.
A 250w PSU could limit you depending on the rest of your hardware and the exact model/ratings of this PSU.

KT6V Chassis Temp Sensor?

Anyone know where the KT6V LSR chassis temp sensor is? I am sitting here running Prime 95 on a new watercooled rig. Ambient temp si 22 C, motherboard monitor is reporting the chassis temp at 35C and the CPU at 38C. I love the fact that the CPU is this low under 100% stress, but cant imagine why the chassis is reporting 13 degrees over ambient. There is a fan blowing air on the mosfet and caps next to the cpu to keep them cool.

Quote
Originally posted by Mr_Woodchuck
also because of electronics' ability to increase temp by 200% in under 10 seconds without any form of cooling, unless you've got a hurricane blowing freezing cold air over the cpu I'd say that having the cpu at ambient temperature is physically impossible
Did not expect to get to ambient. That was not my question. If you reread the original post, I am trying to determine the location of the sensor on the motherboard that reports the chassis temp. I am more than happy with my cpu temp of 38 C. I am trying to figure out why the chassis is reporting so high with the side off the case, and I have one case fan blowing on the area between the cpu and io plate to cool mosfets and caps since there is no wash now coming off of any cpu air cooler.

Temp sensor

Is the temp sensor or corecentre weird on this board? Everytime I start pc, the temp says about 52, But I know thats wrong because If I start and go staright to bios it will say about 35-38. and also the fan stays normal when i get the bad reading.
To make core centre read correct I have to start pc, then restart straight away, otherwise i get the dodgy temp reading.
Think I better get speedfan or mbm

The whole core temperature reading scheme on the K8N NEO2 Platinum is peculiar. I'd swear that the bios is conditioning the actual reading from the Winbond Super I/O chip, for whatever reason. Perhaps an attempt to keep the uninformed blissful. No way I can keep the Vcore the same, reduce the CPU speed 20% and the Prime95 Torture temperature stays the same. No way!
I'm not sure, but if I can find an easy access point to the SMBUS, like the header you find on high quality motherboards, I may implement an external reader IC.
Hoot

GPU temp increasing rapidly after 10.10.3when I open photo app

Just updated to 10.10.3. After the update I see that the photo app is very unstable and just opening the app increases the GPU temps and it does not stop going up until I close the app. I don't even have to use it, just leaving the app open cause temp skips. I laptop GPU when up-to 83°C . Anyone else facing this issue? Do we have a solution?
I am using MacBook Pro (Retina, Mid 2012).
2.3 GHz Intel Core i7
16 GB 1600 MHz DDR3
Dual GPU card(Intel HD Graphics 4000 1024 MB,NVIDIA GeForce GT 650M)
Thank you.

Reinstalling OS X Without Erasing the Drive
Boot to the Recovery HD: Restart the computer and after the chime press and hold down the COMMAND and R keys until the menu screen appears. Alternatively, restart the computer and after the chime press and hold down the OPTION key until the boot manager screen appears. Select the Recovery HD and click on the downward pointing arrow button.
Reinstalling OS X Without Erasing the Drive
Repair the Hard Drive and Permissions: Upon startup select Disk Utility from the main menu. Repair the Hard Drive and Permissions as follows.
When the recovery menu appears select Disk Utility and press the Continue button. After Disk Utility loads select the Macintosh HD entry from the the left side list. Click on the First Aid tab, then click on the Repair Disk button. If Disk Utility reports any errors that have been fixed, then re-run Repair Disk until no errors are reported. If no errors are reported click on the Repair Permissions button. Wait until the operation completes, then quit Disk Utility and return to the main menu.
Reinstall OS X: Select Reinstall OS X and click on the Continue button.
Note: You will need an active Internet connection. I suggest using Ethernet if possible because it is three times faster than wireless.
Alternatively, see:
Reinstall OS X Without Erasing the Drive
Choose the version you have installed now:
OS X Yosemite- Reinstall OS X
OS X Mavericks- Reinstall OS X
OS X Mountain Lion- Reinstall OS X
OS X Lion- Reinstall Mac OS X
Note: You will need an active Internet connection. I suggest using Ethernet
if possible because it is three times faster than wireless.

HDD temp sensor

Can anyone tell me how to tell if my HDD temp sensor
is connected, or if it's connected correctly, or anything
would Help.
I've connected a new SSD and HDD in my late 2009 iMac 27,
everything seems to be working OK, I just want to make susure this HDD
temp sensor is hooked up right before I put the LCD. and glass back on.
I'm using an external monitor to test everything as I go.
I've installed iStat pro and it says the SSD is 0 degrees,and the
HDD is 25 degrees. This doesn't make sense to me.
The HDD has what looks like a .1 connector with a double row
of 4 pins. And my temp sensor cable has a similar female coconnector with
only 3 pins per row, with the 2 sensor wires connected in the center.
So I can swap this around 4 different ways. But I don't know how to
to tell when it's right.
Anyhelp thanks.
BTW, the fans are 3900 for the optical, and 1000 or so for ththe other two.
I'm using OS X 10.8

I should also note that I have reset the PRAM and SMC to no avail. Also the system seems to have totally crashed and now if I try to boot from the SSD, I get the prohibited sign, but it will boot from the recovery partition. Running Disk Utility on the SSD from the recovery partition gives no errors.

MacBookPro 15" 1.1: Strange GPU temp sensor behaviour

Similar Messages

Maybe you are looking for