Do not use the macbook pro '11 under clamshell mode! Proof inside!

[snaky69]

Your keyboard gets warm due to air from the vent at the back blowing over the aluminium as the screen hinge diverts its flow. Its actually not the aluminium dissipating heat, its the aluminium absorbing heat that would otherwise be vented into the air.

That is totally wrong.

You don't seem to understand how heat dissipation works. It isn't directional, heat flows from hottest to coolest, and can propagate through radiation(infrared), plain old contact and convection in a fluid.

Heat is most definitely coming straight from the CPU, none of that nonsensical air from the fans theory.

Just take a look at this infrared picture if you're not convinced.

As you can see the air is diverted straight up and onto the screen, not over the keyboard.

 

[Eddyisgreat]

Just take a look at this infrared picture if you're not convinced.
Image

Oh wow we're getting thermal in this mofo

 

[snaky69]

Oh wow we're getting thermal in this mofo

I didn't take the picture, but I remembered seeing it on here and google image found it for me.

 

[GGJstudios]

Can someone also confirm there's actually no air *intake* at all coming in via the keyboard?

There is a solid sheet under the keyboard, preventing any meaningful air flow. All the venting (intake and exhaust) is done at the rear, near the hinge. Heat radiates in all directions, whether you have the lid open or closed. Aluminum acts like a heat sink, making it hotter than materials used in other notebooks. There is no problem at all running in clamshell mode. A degree or two of heat isn't going to make a meaningful difference. The title of this thread is misleading. It's perfectly fine to use any Mac notebook in clamshell mode.

 

[dsio]

That is totally wrong.

You don't seem to understand how heat dissipation works. It isn't directional, heat flows from hottest to coolest, and can propagate through radiation(infrared), plain old contact and convection in a fluid.

Heat is most definitely coming straight from the CPU, none of that nonsensical air from the fans theory.

Just take a look at this infrared picture if you're not convinced. Image

Actually, I think you are a little confused. On the 13" macbook pro, the hottest part of the chassis under load is the region directly above the heat sink exhaust vent. The CPU is offset around two inches to the left of the heatsink, as the heatpipe inside a Sandy Bridge 13" is an "L" shape.

There is plastic used, and even a strip of insulating foam mechanically separating the heatsink assembly from any part of the aluminium chassis, and the entire assembly is support by the logic board alone, which makes sense, especially given that the current Sandy Bridge chips will run above 90C under full load in normal circumstances, and having that as thermally isolated as possible means the aluminium casing will not get uncomfortable as quickly for the user.

That isn't to say there isn't any radiated heat being absorbed by shell, it just happens to be a tiny fraction of that being conducted to the heatsink and exhuasted.

As far as your comment heat isn't directional? If you have a reasonably efficient heatpipe connecting two devices, one of which is creating substantial heat, one of which is dissipating substantial heat, there very much is a direction. If that cold air is coming in, passing over the heatsink, being heated, and flowing out the back, it also has a direction. The fact is even the plastic cover of the screen hinge gets very warm when the machine is at full load, even though the plastic has no direct contact, and is a lousy thermal conductor at that. Its not magic, its the hot air constantly blowing over it, just like your desk also gets hot where that air is venting, and why your cat wants to sit behind it.

 

[Alexjones]

I run it in clamshell mode. The CPU purrs like a kitten and there is very little heat to boot. Try opening the lid about a half an inch just before the wake up point if the heat bothers you.

 

[wegster]

There is a solid sheet under the keyboard, preventing any meaningful air flow. All the venting (intake and exhaust) is done at the rear, near the hinge. Heat radiates in all directions, whether you have the lid open or closed. Aluminum acts like a heat sink, making it hotter than materials used in other notebooks. There is no problem at all running in clamshell mode. A degree or two of heat isn't going to make a meaningful difference. The title of this thread is misleading. It's perfectly fine to use any Mac notebook in clamshell mode.

Thanks (and dsio before) - I haven't needed to pull apart a macbook that far, and the chicklet key design certainly looked like it's possible that it was drawing some amount of airflow, although checking with a small piece of paper for airflow pointed towards very little or zero airflow. The 2011s run hot enough that I didn't want to consider a keyboard cover without confirming this one, thanks again. I still don't *like* the idea of a kb skin, but it's not an inexpensive machine, and I'm usually in desktop/external kb/mouse/display mode so I can probably tolerate it for meetings and travel.

I understand about heat pipes, have built hundreds of systems and managed thousands, but the MBP isn't quite the same as a workstation nor rackmount servers or blade chassis with removable fan and power supply modules, etc. What I haven't seen is a diagram of the air intake and airflow inside a MB/Pro - it seems more than slightly 'odd' if the air intake as well as exhaust is all in the back, without any intake (and thus meaningful air flow) coming across the front of the system. Maybe I missed something obvious when i last had it apart or in teardown pics, but I'm also not seeing any obvious channeling to help direct air except across the heat pipe radiators? Is airflow solely drawn in from the rear, and maybe the speaker grille alone, and the rest is mostly just dissipation via the chassis? In other words, what provides actual airflow over the RAM, HD, battery, as the fans really just blow across the ends of the heat pipes - is it just minor 'side-effect' airflow versus a more intentional path that would move air across more components?

 

[Eddyisgreat]

What I haven't seen is a diagram of the air intake and airflow inside a MB/Pro - it seems more than slightly 'odd' if the air intake as well as exhaust is all in the back, without any intake (and thus meaningful air flow) coming across the front of the system. Maybe I missed something obvious when i last had it apart or in teardown pics, but I'm also not seeing any obvious channeling to help direct air except across the heat pipe radiators? Is airflow solely drawn in from the rear, and maybe the speaker grille alone, and the rest is mostly just dissipation via the chassis? In other words, what provides actual airflow over the RAM, HD, battery, as the fans really just blow across the ends of the heat pipes - is it just minor 'side-effect' airflow versus a more intentional path that would move air across more components?

Ok do this unscientific test on the 13" Macbook Pro (it's different for the 15 and 17" because they have two fans vs 1).

Go to terminal
Type in "yes > /dev/null" < - this will generate CPU activity. Let it sit for a while and hear your fans start to spin up.
Grab a sticky note/post-it
With the lid closed (clamshell mode) and the front facing you, put the nonadhesive edge of the sticky note on the back of the machine, anywhere right of the apple. Notice how it falls.
Now do the same thing on the left side, about a few inches away from the magsafe. Notice how it stays up. Right next to it will be the warm exhaust air.

The battery hdd and ram don't really need their own cooling modules. Stock kit isn't over clocked or high RPM. Battery shouldn't warm up that much either.

 

[dsio]

Thanks (and dsio before) - I haven't needed to pull apart a macbook that far, and the chicklet key design certainly looked like it's possible that it was drawing some amount of airflow, although checking with a small piece of paper for airflow pointed towards very little or zero airflow. The 2011s run hot enough that I didn't want to consider a keyboard cover without confirming this one, thanks again. I still don't *like* the idea of a kb skin, but it's not an inexpensive machine, and I'm usually in desktop/external kb/mouse/display mode so I can probably tolerate it for meetings and travel.

I understand about heat pipes, have built hundreds of systems and managed thousands, but the MBP isn't quite the same as a workstation nor rackmount servers or blade chassis with removable fan and power supply modules, etc. What I haven't seen is a diagram of the air intake and airflow inside a MB/Pro - it seems more than slightly 'odd' if the air intake as well as exhaust is all in the back, without any intake (and thus meaningful air flow) coming across the front of the system. Maybe I missed something obvious when i last had it apart or in teardown pics, but I'm also not seeing any obvious channeling to help direct air except across the heat pipe radiators? Is airflow solely drawn in from the rear, and maybe the speaker grille alone, and the rest is mostly just dissipation via the chassis? In other words, what provides actual airflow over the RAM, HD, battery, as the fans really just blow across the ends of the heat pipes - is it just minor 'side-effect' airflow versus a more intentional path that would move air across more components?

I can only speak to the 13", but in that, the intake air comes from beside the exhaust (that is, directly to the left of the heatsink behind the hinge as the laptop sits open in front of you). There's a trail of dust on the inner surface of the baseplate that runs in an arc from the intake vent around to the fan that shows its path, the DDR3 in laptops doesn't generate enough heat to need any active cooling and even a 7200RPM Momentus XT only runs 3W total so its not really necessary.

I would say the most likely intake point for the 15 is between the two fans at the back.

 

[Ronnoco]

I run it in clamshell mode. The CPU purrs like a kitten and there is very little heat to boot. Try opening the lid about a half an inch just before the wake up point if the heat bothers you.

This is great...thanks...it works in Lion and at the very least gives me some piece of mind that I'm allowing some trapped heat to escape (even if it is just a small amount) as I've already had the logic board replaced and the MBP is only 13 months old...I just want to do everything I can to keep this thing cool when I'm stressing the CPU with things like video processing...

 

[unclewebb]

The multipliers look wrong at every state in that Throttlestop program,

They look wrong compared to what? Compared to other software like CPU-Z that doesn't bother to follow the Intel recommended monitoring methods because he has chosen consistent validations over accuracy.

Here is the Intel White Paper that explains the recommended monitoring method for Core i processors.

http://files.shareholder.com/downlo...C8-A433-E28F64CB8EF2/TurboBoostWhitePaper.pdf

ThrottleStop follows this method while most other monitoring software does not. When I asked the programmer of CPU-Z to explain this, here was his reply.

Of course I admit that CPU-Z is not accurate anymore at idle on latest Intel generations, that is why TMonitor was developped.

You may not understand what ThrottleStop is all about but it has turned the laptop PC industry on its ear. Ask Dell how accurate the data coming from ThrottleStop is. Then ask Asus, Acer, HP or Lenovo. The public backlash Dell received when users discovered that one Dell laptop model after another was throttling cost them a fortune in lost sales, on site service calls and forced them to completely redesign their entire laptop line up including a world wide recall of power adapters that were proven to be totally inadequate by ThrottleStop.

If you want to see the significant performance improvement that ThrottleStop is capable of then check out this video.

http://www.youtube.com/watch?v=0JdRUHHn0UM

That is a performance gaming laptop that is throttling so bad that is is unusable. ThrottleStop is the only software in the world that can solve problems like this in a wide variety of laptops and best of all it is absolutely free.

World records like scook9 running his QX9300 at 4333 MHz were not possible without ThrottleStop.

http://ro.hwbot.org/submission/1014630_scook9_cpu_z_core_2_qx9300_2.53ghz_4333_mhz

While you're doing some Google home work, check out what all of the 920XM and 940XM world record holding laptops are using. Maybe ask svl7 if it is possible to run his 920XM fully loaded at 4.0 GHz without ThrottleStop.

http://forum.techinferno.com/genera...els-increasing-cpu-core-voltage.html#post9712

The first post in this thread clearly shows that when fully loaded, the cooling system in this laptop is completely inadequate and the CPU is throttling like crazy. Maybe Apple could learn a lesson or two from Dell and others.

The second picture shows an average multiplier of 19.32 on each thread. The default multiplier for a Core i7-2720QM is 22. A 2720QM should never go below this value when it is fully loaded. When it does, that's throttling. The lower average multiplier is a clear sign of that. The lower VID voltage that the CPU is requesting as well as the lower CPU power consumption number also confirms that this CPU is throttling more because of the increased heat. Run this exact same test on an M18x with the exact same CPU and you will understand the difference that proper cooling can make to the performance of these new Sandy Bridge CPUs.

... looks like something slapped together in Visual Basic, so I wouldn't put a great amount of faith in it.

I am a little biased but I would recommend that users put more faith in a well proven app than a pretty looking app. Maybe some Apple engineers could learn a thing or two about laptop design if they understood what this app was telling them.

 

[Macsavvytech]

Actually, no it doesn't its blocked off from the motherboard.

I think you misunderstood what I was saying, in terms of fans it is completely disconnected, heat can dissipate through the aluminium but the effect would be fairly minimal.

Doesn't this this tell you that there is no air *intake* through the keyboard?

Can someone also confirm there's actually no air *intake* at all coming in via the keyboard?

The posted numbers aren't surprising, the CPU looks like it's temperature throttling itself, no? Even if the keyboard doesn't increase airflow, the upper body still likely has small amount of heat dissipation via airflow over it, versus closed in clamshell mode. Without taking samples of temps over time or ensuring enough cooldown to start at the same exact CPU temp in open vs clamshell mode, across multiple tests, it's not exactly conclusive, but seeing a difference of roughly 1*C or so isn't a real surprise.

Maybe I'm misunderstanding the data in the screenshots - look at the variance in the frequency multiplier in the second shot's open mode - with max temps of 65*C the multiplier is all over the place - doesn't seem to be exactly under load so much as possible sporadic in and out of boost from background processes, versus doing anything maxxing the CPU out.. ??

Thanks Eddy for stating it once more.

There is absolutely no air intake at all coming in via the keyboard.

I could go into a long winded discussion about heat pipes and heat sinks and airflow and fan design but I try to avoid doing so.

 

[dsio]

They look wrong compared to what? Compared to other software like CPU-Z that doesn't bother to follow the Intel recommended monitoring methods because he has chosen consistent validations over accuracy.

I've not used CPU-Z since back when it was wcpuid and never really cared for windows anyway tbh

Here is the Intel White Paper that explains the recommended monitoring method for Core i processors.

http://files.shareholder.com/downlo...C8-A433-E28F64CB8EF2/TurboBoostWhitePaper.pdf

ThrottleStop follows this method while most other monitoring software does not. When I asked the programmer of CPU-Z to explain this, here was his reply.

My problem isn't with your method of extracting values, its with your attempt to represent them, showing an average multiplier is an inaccurate way of representing that data, I can see why you did it, as its probably the easiest way to do it, it just isn't a particularly effective way.

You may not understand what ThrottleStop is all about but it has turned the laptop PC industry on its ear. Ask Dell how accurate the data coming from ThrottleStop is. Then ask Asus, Acer, HP or Lenovo. The public backlash Dell received when users discovered that one Dell laptop model after another was throttling cost them a fortune in lost sales, on site service calls and forced them to completely redesign their entire laptop line up including a world wide recall of power adapters that were proven to be totally inadequate by ThrottleStop.

Starting to sound a bit like self promotion... if you genuinely think you're the first person to discover that the quality of laptops has been in a dramatic decline over the last decade, you may need to get out more. Dell have been stepping on their own Johnson since forever, hell they even had to withdraw the T7700 M1330 from sale because in their effort to get to the top of the review sites they exceeded the thermal limits of their cooling by a mile. Makers of gaming laptops will continue trying to win battles on paper with no regard for their owners, thats just the way it is.

If you want to see the significant performance improvement that ThrottleStop is capable of then check out this video.

http://www.youtube.com/watch?v=0JdRUHHn0UM

That is a performance gaming laptop that is throttling so bad that is is unusable. ThrottleStop is the only software in the world that can solve problems like this in a wide variety of laptops and best of all it is absolutely free.

All that proves is that the owner should have taken his laptop back where he bought it and asked for a cash refund.

World records like scook9 running his QX9300 at 4333 MHz were not possible without ThrottleStop.

http://ro.hwbot.org/submission/1014630_scook9_cpu_z_core_2_qx9300_2.53ghz_4333_mhz

While you're doing some Google home work, check out what all of the 920XM and 940XM world record holding laptops are using. Maybe ask svl7 if it is possible to run his 920XM fully loaded at 4.0 GHz without ThrottleStop.

http://forum.techinferno.com/genera...els-increasing-cpu-core-voltage.html#post9712

this isn't actually worth googling

The first post in this thread clearly shows that when fully loaded, the cooling system in this laptop is completely inadequate and the CPU is throttling like crazy. Maybe Apple could learn a lesson or two from Dell and others.

Just shows he should return his laptop. If you have a look at a recently thread here some guy had a MBP that was so caked with melted thermal pad that it had leaked down through the heatsink and out the other side, that laptop is a "dud" and needs to go back. If this guy's laptop doesn't work under normal usage, it should go back as well.

The second picture shows an average multiplier of 19.32 on each thread. The default multiplier for a Core i7-2720QM is 22. A 2720QM should never go below this value when it is fully loaded. When it does, that's throttling. The lower average multiplier is a clear sign of that. The lower VID voltage that the CPU is requesting as well as the lower CPU power consumption number also confirms that this CPU is throttling more because of the increased heat. Run this exact same test on an M18x with the exact same CPU and you will understand the difference that proper cooling can make to the performance of these new Sandy Bridge CPUs.

Nobody's disagreeing with that.

I am a little biased but I would recommend that users put more faith in a well proven app than a pretty looking app. Maybe some Apple engineers could learn a thing or two about laptop design if they understood what this app was telling them.

Coming from linux and embedded systems gear, I've got nothing against ugly apps, I just have little to no respect for VB or .net devs in general, which is possibly an unfair generalization, but has proven pretty accurate in the past.

 

[nnoob]

They look wrong compared to what? Compared to other software like CPU-Z that doesn't bother to follow the Intel recommended monitoring methods because he has chosen consistent validations over accuracy.

Here is the Intel White Paper that explains the recommended monitoring method for Core i processors.

http://files.shareholder.com/downlo...C8-A433-E28F64CB8EF2/TurboBoostWhitePaper.pdf

ThrottleStop follows this method while most other monitoring software does not. When I asked the programmer of CPU-Z to explain this, here was his reply.



You may not understand what ThrottleStop is all about but it has turned the laptop PC industry on its ear. Ask Dell how accurate the data coming from ThrottleStop is. Then ask Asus, Acer, HP or Lenovo. The public backlash Dell received when users discovered that one Dell laptop model after another was throttling cost them a fortune in lost sales, on site service calls and forced them to completely redesign their entire laptop line up including a world wide recall of power adapters that were proven to be totally inadequate by ThrottleStop.

If you want to see the significant performance improvement that ThrottleStop is capable of then check out this video.

http://www.youtube.com/watch?v=0JdRUHHn0UM

That is a performance gaming laptop that is throttling so bad that is is unusable. ThrottleStop is the only software in the world that can solve problems like this in a wide variety of laptops and best of all it is absolutely free.

World records like scook9 running his QX9300 at 4333 MHz were not possible without ThrottleStop.

http://ro.hwbot.org/submission/1014630_scook9_cpu_z_core_2_qx9300_2.53ghz_4333_mhz

While you're doing some Google home work, check out what all of the 920XM and 940XM world record holding laptops are using. Maybe ask svl7 if it is possible to run his 920XM fully loaded at 4.0 GHz without ThrottleStop.

http://forum.techinferno.com/genera...els-increasing-cpu-core-voltage.html#post9712

The first post in this thread clearly shows that when fully loaded, the cooling system in this laptop is completely inadequate and the CPU is throttling like crazy. Maybe Apple could learn a lesson or two from Dell and others.

The second picture shows an average multiplier of 19.32 on each thread. The default multiplier for a Core i7-2720QM is 22. A 2720QM should never go below this value when it is fully loaded. When it does, that's throttling. The lower average multiplier is a clear sign of that. The lower VID voltage that the CPU is requesting as well as the lower CPU power consumption number also confirms that this CPU is throttling more because of the increased heat. Run this exact same test on an M18x with the exact same CPU and you will understand the difference that proper cooling can make to the performance of these new Sandy Bridge CPUs.



I am a little biased but I would recommend that users put more faith in a well proven app than a pretty looking app. Maybe some Apple engineers could learn a thing or two about laptop design if they understood what this app was telling them.

This is what I have been telling all along and BTW this is the author of Throttlstop

Apple throttles the macbook pro CPU for GPU intensive task

https://forums.macrumors.com/threads/1203927/

This is why I use throttlestop..

 

[dsio]

This is what I have been telling all along and BTW this is the author of Throttlstop

Apple throttles the macbook pro CPU for GPU intensive task

https://forums.macrumors.com/threads/1203927/

This is why I use throttlestop..

If your Macbook is throttling under normal use, the common sense approach is to return it under warranty...

 

[nnoob]

All that proves is that the owner should have taken his laptop back where he bought it and asked for a cash refund.

Maybe you should return your macbook because it does the same thing and apple is no better then the rest of the pc worlds....
See this for proof
https://forums.macrumors.com/threads/1203927/

 

[dsio]

Maybe you should return your macbook because it does the same thing and apple is no better then the rest of the pc worlds....
See this for proof
https://forums.macrumors.com/threads/1203927/

Sorry what exactly is that meant to prove... you posted the exact same thing in another thread... and then kept agreeing with yourself...

 

[nnoob]

If your Macbook is throttling under normal use, the common sense approach is to return it under warranty...

It a power requirement issue, not because of heat issues. When you do CPU- GPU task. it throttles down to 800 to starve off some power from the 85 watt power adapter.. with Throttlestop, It even barrows charges from the battery because the power requirement is too high on the macbook power brick even when the internal LCD is off...
Apple need bigger power adapter to fix the problem

 

[dsio]

It a power requirement issue, not because of heat issues. When you do CPU- GPU task. it throttles down to 800 to starve off some power from the 85 watt power adapter.. It even barrows charges from the battery because the power requirement is too high on the macbook power brick even when the internal LCD is off...

Then what in the hell was the point of the title of the thread you just created: Do not use the macbook pro '11 under clamshell mode! Proof inside!


Do you think clamshell mode increases total power consumption?

Please... when you figure out what you are on about, let us know...

 

[nnoob]

Then what in the hell was the point of the title of the thread you just created: Do not use the macbook pro '11 under clamshell mode! Proof inside!


Do you think clamshell mode increases total power consumption?

Please... when you figure out what you are on about, let us know...

It called off topic

 

[dsio]

It called off topic

OK... Well then what was the point of the original topic?