Extremely HOT! 85W MagSafe Power Adapter

[grapii]

Hi All,

Just got my MBP on wednesday, and am getting a bit worried on how hot the power adapter is getting. The laptop itself is fine, but the pa doesn't seem normal.

I know these kings are meant to get warm, I've had plenty of dell laptops historically, but this mac one is really warm, as far as being hot.

The only way I can describe the heat is that if I was washing my hands under the tap, and this was the temperature of the water, I'd have to turn it down and make it cooler.

Is this normal, or should I call AppleCare?

 

[ouimetnick]

If it burns your flesh, call Apple. they will ask you a lot of things regarding safty, and replace the adapter. I've noticed the Apple iBricks get warmer than PC bricks, but the Mag safe never burned me.

 

[grapii]

If it burns your flesh, call Apple. they will ask you a lot of things regarding safty, and replace the adapter. I've noticed the Apple iBricks get warmer than PC bricks, but the Mag safe never burned me.

It's not burning me as such, and I can hold it in my hand indefinitely if i wanted, but it would be uncomfortable.

it's as hot as a freshly made hot water bottle. (boil kettle > pour into hot water bottle and seal > now touch the rubber bottle....that's how hot my adapter is.

 

[Timur]

It's "normal". This happens because the PSU are dimensioned too small and thus often run at their limit. The Mac even has to draw concurrent power from battery sometimes because it needs more than 85W.

 

[rmbrown09]

**reaches down and touches 85w**
cool as can be, and sitting on carpet

 

[SchneiderMan]

Yeah mine has been running all day and it's at room temp. Maybe because it's not charging? I have it plugged in all the time so.

 

[Timur]

Running normal operations on OS X isn't much of a problem (13-45 W depending on what you are doing), problems arise when you are stressing the Mac a bit more.

Using the dedicated graphic needs more power. Loading the battery does, too (keep it plugged in whenever you can, also helps battery-life). Using bootcamped Windows draws more power, as does using virtualization like Parallels. Look out for certain Firefox plugins that may keep your CPU and HD working harder (check via Activity Monitor).

 

[DoFoT9]

**reaches down and touches 85w**
cool as can be, and sitting on carpet

Yeah mine has been running all day and it's at room temp. Maybe because it's not charging? I have it plugged in all the time so.

if your machine is fully charged then the charger no longer needs to charge the battery as well as provide power to the computer - it just provides power to the components (once charged, the battery receives no additional power).

Timur - running a VM doesnt add that much more power requierements.. a few extra load cycles here and there...

 

[Timur]

My 2009 MBP 17" draw around 25 W running OS X Snow Leopard idle with non-dedicated graphic-card with display brightness at 50%.

Average idle power consumption increases by around 5 W when using Parallels 5 (running a Windows 7 64-bit VM of my Bootcamp partition).

The dedicated 9600M GT draw around 7 W more power compared to the inbuild 9400M on my 2009 MBP 17".

The 17" display draws quite more power than the 15" one while both use the very same 85W PSU. My former 2008 15" MBP had a minimum consumption of only 13 W running Leopard (the only other difference is the use of a 7200 rpm HD vs a 5400 one, everything else same specs).

Using an external display increases power consumption by around 1-2 W.

Loading the battery draws around 20 W more power additionally, also in parts because less power if draw from the battery for operation (which usually happens even when plugged in). That means 57 W for an idle OS X desktop with dedicated GPU which leaves less than 30 W headroom for pushing the CPU and GPU to their limits (this is why you need to keep the battery plugged in to get full CPU power out of a Macbook).

You should try Istat Menu, Activity Monitor and the terminal command "top" to find out what's drawing CPU cycles and power if the PSU keeps getting too hot.

 

[Timur]

Two things to mention:

1. The power draw from loading the battery changes depending on the current load status of the battery (and maybe temperature, too). The closer the battery gets to full load the less wattage is used.

The battery is reloaded (red led on PSU) whenever it's load drops below 95%. Load between 95-100% will not trigger the loading mechanism (green LED).

2. The average increase of around 5 W for using Parallels corresponds somewhat with the average higher power draw in Bootcamp (around 7 W). So it's not Parallels that draws the additional power, but Windows itself (apart from some CPU overhead especially for HD access and intense graphics).

 

[DoFoT9]

Average idle power consumption increases by around 5 W when using Parallels 5 (running a Windows 7 64-bit VM of my Bootcamp partition).

which translates into how much of a change in how long the battery lasts? 30mins?

Loading the battery draws around 20 W more power additionally, also in parts because less power if draw from the battery for operation (which usually happens even when plugged in). That means 57 W for an idle OS X desktop with dedicated GPU which leaves less than 30 W headroom for pushing the CPU and GPU to their limits (this is why you need to keep the battery plugged in to get full CPU power out of a Macbook).

what on earth do you mean here? are you saying that the CPU only reaches its highest clocks when plugged into power, because thats just not correct..

Two things to mention:

1. The power draw from loading the battery changes depending on the current load status of the battery (and maybe temperature, too). The closer the battery gets to full load the less wattage is used.

The battery is reloaded (red led on PSU) whenever it's load drops below 95%. Load between 95-100% will not trigger the loading mechanism (green LED).

once the battery is fully charged, it is no longer "used" or charged - saving the battery effectively.

also, i was under the impression that up until 80% capacity - the battery was being fully charged with full wattage. then after 80% it was trickle charged.

2. The average increase of around 5 W for using Parallels corresponds somewhat with the average higher power draw in Bootcamp (around 7 W). So it's not Parallels that draws the additional power, but Windows itself (apart from some CPU overhead especially for HD access and intense graphics).

vmware also does this i presume? wrt virtualisation, actual performance is never a 1:1 scenario, there is always overhead.

wrt bootcamp - the power management drivers musnt be efficient enough.

 

[Timur]

which translates into how much of a change in how long the battery lasts? 30mins?

Depends on what you compare it to. Pure idle time, yes, 30 minutes. But we are talking about PSU heat here, not battery time, and I listed things that add up to more load/heat on the PSU.

what on earth do you mean here? are you saying that the CPU only reaches its highest clocks when plugged into power, because thats just not correct..

No, read that paragraph again. I am saying that the CPU only reaches its highest clock when the battery is plugged into the computer (not the PSU).

With the former removable batteries the CPU would stay at its *lowest* clock when the battery was removed. That is because there are time when over 85 W are used (high CPU and GPU load). And without the battery supplying extra power the PSU would burn down. So Apple *deliberately* dimensioned the PSU too small for peak usage in order to make it smaller (cheaper?) for common use scenarios.

once the battery is fully charged, it is no longer "used" or charged - saving the battery effectively.

The Macbook always draws part of its load from the battery, even when its plugged in. With low load/idle scenarios it's only a few watts (see iStat Menu for reading of real numbers), with high load/peak scenarios it's a considerably larger amount. I can load my Macbook Pro well over 90 W over long periods, and yes, the battery can run dry after some time even when plugged into the PSU. But in most scenarios it seems like Apple found a good balance, but therefor the little plastic brick gets quite hot.

also, i was under the impression that up until 80% capacity - the battery was being fully charged with full wattage. then after 80% it was trickle charged.

Yes, the load current drops bit by bit the higher the charge. This is due to temperature and longevity considerations and works the very same with all mobile phone batteries (and is the reason why load mechanisms for these kind of batteries are more complex than the usuan NiCD and NiMH ones).

vmware also does this i presume? wrt virtualisation, actual performance is never a 1:1 scenario, there is always overhead.

Quite likely yes. Like before I was just listing several points where wattage can easily add up even when the computer seems to be idle.

wrt bootcamp - the power management drivers musnt be efficient enough.

Parts and parts. I suspect that some of the blame is to be put on Apple's bad ACPI/BIOS Emulation implementation, which leads to numerous other problems with Bootcamp as well (one reason why comparing boot-times of OS X to Windows makes no real sense on a Mac).

 

[DoFoT9]

No, read that paragraph again. I am saying that the CPU only reaches its highest clock when the battery is plugged into the computer (not the PSU).

oh. baha ok gotchya - yes im well informed of what apple has done there. just didnt read it properly.

The Macbook always draws part of its load from the battery, even when its plugged in. With low load/idle scenarios it's only a few watts (see iStat Menu for reading of real numbers), with high load/peak scenarios it's a considerably larger amount. I can load my Macbook Pro well over 90 W over long periods, and yes, the battery can run dry after some time even when plugged into the PSU. But in most scenarios it seems like Apple found a good balance, but therefor the little plastic brick gets quite hot.

i would love to see some proof of this. because on my CD MBP, even with 2 monitors, a bunch of USB devices and 200% CPU usage. the battery ALWAYS stayed fully charged.

Yes, the load current drops bit by bit the higher the charge. This is due to temperature and longevity considerations and works the very same with all mobile phone batteries (and is the reason why load mechanisms for these kind of batteries are more complex than the usuan NiCD and NiMH ones).

thats intersting to know.

Parts and parts. I suspect that some of the blame is to be put on Apple's bad ACPI/BIOS Emulation implementation, which leads to numerous other problems with Bootcamp as well (one reason why comparing boot-times of OS X to Windows makes no real sense on a Mac).

it was my understanding that the BIOS is pretty much only used for bootup to tell the OS what hardware is there etc, after that control is handed off to the OS itself to control everything. no?

 

[grapii]

I've been running the MBP mostly on battery and only plugging the charger in when the battery icon in the menu bar turns red.

Is this the reason why the adapter is getting too hot? as it has to charge the battery from nearly empty?

I always thought that leaving the PSU connected kills the duration of the battery? and that running the machine as long as possible on battery power then charging only until full (not left in), keeps the battery going for longer and keeps it healthy. Am I wrong?

 

[GoyoNeuff]

Take it to Apple store.

They are produced from different companies. I just saw that the i7 (returned) had one done by one company, and the new i5 has it from a totally different one. The finish is not the same tough, the i7 power brick was way better than the one that came with the i5. The plug you can unlatch from it had a better and compact fit on the one from the i7 than the one from the i5. I keept the i5 one.
Cheers @!

 

[Timur]

I've been running the MBP mostly on battery and only plugging the charger in when the battery icon in the menu bar turns red.

Is this the reason why the adapter is getting too hot? as it has to charge the battery from nearly empty?

Yes, because it draws lots of extra power to do both: power the computer and load the battery concurrently.

I always thought that leaving the PSU connected kills the duration of the battery? and that running the machine as long as possible on battery power then charging only until full (not left in), keeps the battery going for longer and keeps it healthy. Am I wrong?

Yes, you are wrong, sorry. Battery life is measured in number of load cycles. That means: the more you use the battery, the earlier it's dead.

Another stress factor is the extra heat building up during loading (and the emptier the battery the more heat is produced by loading). And the quicker a battery is loaded (aka the more current runs into it) the less load cycles it can survive. All laptop batteries come with their own logic circuitry that tries to strike a balance between fast load times and "save" load times.

The Mac's load mechanism is set so that the battery is not loaded all the time when the PSU is plugged in (see red LED vs green LED). But if you leave the PSU connected for some time you will notice that the battery load drops bit by bit and once it reaches around 95% the loading mechanism kicks in (and loads the remaining 5% relatively slowly).

You should keep the battery in some use so that the chemicals inside stay "in motion", but that usually happens with normal use anyway. Just plug the PSU off once in a while.

 

[Timur]

oh. baha ok gotchya - yes im well informed of what apple has done there. just didnt read it properly.

...

i would love to see some proof of this. because on my CD MBP, even with 2 monitors, a bunch of USB devices and 200% CPU usage. the battery ALWAYS stayed fully charged.

It's not a common situation. I just had my PSU loaded at 86-95 W running some 3D game on Windows. And if you check iStat Menu on OS X you will notice that there is a steady load on the battery (1-5 W) according to its readings. It's nothing to worry about, but just another reason why sometimes the PSU has to work harder.

it was my understanding that the BIOS is pretty much only used for bootup to tell the OS what hardware is there etc, after that control is handed off to the OS itself to control everything. no?

Yes and No. Once the OS (Windows) takes over it handles most things by itself and thus is not depending on the slow and faulty BIOS emulation. This for example is the reason why the Windows based Chkdsk runs in a couple of minutes while the emulated BIOS based one (that sometimes happens before Windows boots up after a crash) takes hours.

But some things still make use of the emulated BIOS layer. All 2008/2009 based Macbooks cannot use their ExpressCard slot on Vista/W7 for example (XP works), which is most likely because the emulated BIOS seems to handle IRQ resources for that port the wrong way (cards cannot get an IRQ for their operation).

Especially all power/energy based functions seem to be problematic: random freezes on my 17" MBP when allowing to put the HD to sleep, Windows reporting memory corruption everytime you put the MBP to sleep, random reports of ACPI errors on all Windows version (XP, Vista, W7) with Event Viewer stating a "Firmware" error, etc.

 

[penter]

To the OP,
Mine seems to be just as hot. I don't think it's uncommon.
These new MBPs seem to run hot in every way, I guess.