I have exactly the same problem- it's good to see your thorough dissection of it!
I have 1st gen MBP core duo 2Ghz, 15.4 screen. It was my wifes and when she upgraded I put the max RAM of 2gb in and put in a new Momentous hybrid drive. It kept crashing during OS installation and would only install fully if I took the battery out.
generally very pleased with the performance but just like you anything that involves moving multiGb files or processing video (eg rendering) makes the temp ramp upuntil it hits 120c and cuts out. Interestingly it never does this with the battery removed - I am not sure if this relates to increased cooling or some fault when the battery in (when I failed to load the os (above) there was a warning mesage that equated to battery error). So I bought a new battery- same problem..
I took it to a Mac store in London, (because I blamed it on a new battery I had just bought) they checked it in the Genius bar and said everything working fine, but "it should not be happening", but it was after all a "vintage mac"
When I do intense taks I remove the battery, but have to tape the power lead to the desk, otherwise if I knock out the mag connector all work is lost.
I don't want to be forced to buy a new mac as it is otherwise a very capable machine. When I changed the HDD it was very clean inside. Mind you I haven't got as far as changing the heatsink thermal compound.
I wonder if my very fast HDD is adding to the problem- any ideas?
Does your mac run cool without the battery in?
I have a similar 'vintage mac', 17" 2.16 core duo which runs fine with the battery inside all the time.
What I have noticed the MBP runs hotter that my newer Unibody based on temperature readings that I have taken on the outside surface of the machine and that the 'normal' fan speed is 1000 rpm compared to the 2000 rpm of the Unibody. The conclusion I have come to is that it is simply a hotter machine and I have doubts that the HDD that you mention is the cause of your problem.
Thanks for that.
I have no problem with it being a hot machine, can even be useful in the winter to keep my legs warm. The only problem is the CPU reaching 120c and cutting out. This is a nuisance both for loss of work, and also a concern that there may be damage occurring if it is getting that hot each time before it switches itself off...
The temperature sensors are shutting your MBP down so that it will not commit suicide.
How fast are the fans running? You might try a SMC reset. Run an Apple Hardware test.
If this is the case under 'normal' use, not CPU intensive tasks, then there must some problem. If you have the mechanical aptitude, I would remove the keyboard and inspect the interior for dust and whatever. If nothing of consequence is found then I would take to the Apple store for analysis.
Can't remember the fan speed but they don't seem to be getting quicker at all even with intensive stuff. I have used SMC fan control but this needs 6000rpm to stop it getting to critical temp and then it sound like a noisy fan heater rather than a mac.It is an option but not v good when working with audio..
Exactly like the original poster, this Mac runs cool most of the time but moving multiGB files or video rendering will make the temp ramp up (and sometimes cut out) within 2-5 minutes. If it doesn't get to cut-off temp, then it cools very quickly as soon as said task has finished. Basically I can get away with rendering a short bit of video. Anything more substantial I need the battery removed or fans roaring!
The inside of the mac was dust free when I fitted HDD, dust free again when I recently checked it and the recent Genius bar Mac hardware test was all good.
I still wonder what the relevance of the battery removal is- it doesn't get hot at all with serious tasks and battery out. As I stated earlier- when originally loading Osx it kept cutting out (I now assume this was overheating)- the install log came up with some funny code- when I did a google search some forum response was that was a battery error code.
The macshop were kind enough to give me a new battery "just in case" but this has not helped..
Thanks for your help
You should install the Fan Control system pref and set it to what I discussed above. The widget iStatPro will tell you the results of temperature and cooling fan speed. I doubt if the new HDD added much heat. It likely added a little, but not enough to be the cause if the over heating issues. I believe that the manufacturing proves for applying the heat sink grease was less than optimal when the early machines were made (I read quite a bit about this)-they just globbed it on where it should have been used sparingly and a thin layer left after installing the main board. Over year of operation and many thermal cycles (getting hot then cold) this heat sink grease has maybe baked and solidified and no longer conducts heat into the chassis as well as it used to. I also think that the expansion and contraction of the boards has loosened up the thermal connection somewhat. Tighten all the main board screws like I did. It helps a little. It may be that they had to glob on the heat sink grease thicker than they wanted because of the movement of the boards during heating and cooling would lift the processor off the chassis too much for a thin layer of heat sink grease to accommodate.
Mine has been running pretty well since I adjusted the fan cut in speed temperature down and lowered the speed at which the fans reach full rpm. I have had a couple of uncommanded shut downs over the past three months. One of them occurred when the machine got stuck in a continuous loop trying to communicate with the Airport Extreme Base Station which wasn't responding for some reason. I looked at the log and it shut off after attempting this many many times...like hundreds. Seemed like a software glitch and it hasn't happened again. It successfully runs Time Machine backups now, which it couldn't do before I adjusted the fans speeds.
It's not ideal, because I never went in and removed the main board and re-did the heat sink interfaces to the chassis, but it's working well enough to not tear it apart. It's a whole lot snappier with the 7200 rpm drive and extra RAM. Good enough for now, although you can't install OS 10.7 on these Core 2 Duo machines and eventually that will force us to a new computer.
Why the removal of the battery would have a cooling affect on the MBP is a mystery to me. I have pushed this MBP in the past and have had runaway processes on rare occasions where the fans have roared, but it has never shut down on its own accord. Maybe in this regard I have been lucky and you have not.
You might consider putting it on a couple of thick aluminum plates for a heat sink when you press the MBP (as described earlier in this post). It won't hurt to try and it might mitigate your problem.
The fan control system relies on temperature sensors at various points in the unit. If you try the Fan Control system pref and set it to set the Base fan speed level to 2800 rpm, set the lower temp threshold to 110F at that speed, and set the upper threshold temp to 160F where the fan would reach 6000 rpm. Install the widget iStatPro and look at all of the temperature readings. If one of them isn't reading or is way off from the others (way out of range high or low) you may have a bad temperature sensor that screws up the temperature control system
Thank you both for your further advice. I have downloaded fan control - this is much better than SMC fancontrol that I had been using because of the threshold settings.
I copied your settings and did a test render in Adobe premiere that was fairly intensive and took about 25 minutes to run. The temperature crept up but so did the fans, soon reaching 6000 but the highest the temp got to was 101C on the CPU sensor. The HDD stayed cool at about 35c, so did the heatsinks and the GPU never went above 42C (these have never got hot).
I took the battery out and did the same render. The temps were much the same EXCEPT the CPU temp never exceeded 49C (ie 50degrees cooler). The fans hovered at around 3000RPM which was much more audibly acceptable.
The ambient temp is 21C, the mac is on a glass table so removal of battery can't really improve airflow that much. It isn't hot around the battery compartment; the RAM compartment is only slightly warm.
This takes me back to my original suspicion that there is some glitch to do with battery monitoring that overloads the CPU. (this fits in with the battery error log message I got when it overheated when 1st trying to install the OS). Of course it may be a glitch peculiar to my machine, but the batteryless test suggests that this is not primarily a cooling issue. The Apple store didn't have a clue about this and just kept saying my hardware and battery were all fine.
I am really pleased you have helped me get intensive tasks done with battery in situ thanks to fan control. Thank you again!
I am also pleased that this mac is capable of doing all I need cooly (with battery out). I can put off a new MBP for some time!
I would be fascinated to know if battery removal makes other 1st generation MBPs run much cooler. And of course if anyone has explanation or workaround for the (possible) battery issue- let me know.
Great that you got it thru with the fan control. The 101 deg C is about how hot mine gets after cranking for a while with the fans going full on. Thankfully for most of the stuff we do on that machine it doesn't get that hot. But we do some movie editing on it and that pushes it.
I'm going to try mine doing something really intensive with and without the battery. The RAM test software keeps the CPU percent way up for extended periods. I haven't tried that one since I adjusted the fan control thresholds, so I'm not sure it will make it all the way through it. It'll be an interesting test. I'll let you know what happens
I finally had a chance to run Rember (a RAM test app) that uses 50-55% CPU continuously. I ran it with the battery both installed and removed. Mine behaves pretty much like yours. It went up to 225-227 deg F with thye battery in and the fans went to 6000 rpm. I didn't let it finish the test to see if it would make it all the way through without an over heat protection shutdown. While the test was running, I took the battery out and the temperature started falling immediately. It settled out at about 125 deg F. I put the battery back in and the temperature went right back up again to the 225 F. I then just unplugged the power cable to the MBP. The temperature didn't go down any, so it's not the battery charging portion of the power supply adding excessive watts to the machine.
So the battery just being in the machine causes results in a severe temperature rise. I watched the % CPU usage with and without the battery in, and it didn't change significantly (if at all) , so I'd say the heat rise isn't due to battery monitoring. The temperature of the battery itself never really changes much when the CPU temp rises and falls, so I don't believe the battery is adding any heat to the unit. I felt it when I took it out and it was barely warm.
I really have no idea what is happening, but I don't believe our batteries are defective. My replacement battery is an original Apple battery rebuilt with new cells. The machine behaved the same for temperature with the old original equipment battery.
If I had to guess what the issue is, I would say that the power supply design is such that when then battery is installed and reserve power is available, the power supply changes modes of operation when the CPU percentages go above a certain point and the result is considerably more heat from the power supply conducts into the cooling path. When the battery is removed, and the reserve power is not available (charging power volt-amps are always way less then the battery volt-amp ratings), then perhaps the power supply doesn't go into this other mode that results in higher heat loads. But this is pure conjecture. I don't know why they would design it that way.
Thanks for checking that- you seem to have an identical situation to me- this has happened with a worn-out mac battery in situ and also with a brand new apple branded battery and a 3rd party one.
Very interesting to hear someone has the same problem and glad I haven't ripped it apart to add some more heat sink compund!
I too don't know what the reason for this problem is- it would be good if someone at Apple could suggest a solution although it's unlikely several generations of MBP later.
So I'll carry on with the fans turned up* for most work and and occasionally remove the battery and tape the power supply to the table for good measure (when I'm doing intensive audio stuff)
Thanks again for all your thorough input- Do you reckon the CPU works "harder or faster" with the battery in (hence more heat) than out?
*my wife says the Mac sounds like a PC with the fans roaring!
pgoodwin and mickybannon: The reason your machines get hotter with the batteries in place than with the batteries removed is that the CPU clock speed is automatically and drastically throttled back — by 1/3 to 1/2, depending on model — as soon as the battery is removed. Apple has designed this to happen because it's possible for the machine to demand more power than its 85W AC adapter can provide, and under those conditions the battery is drawn upon for the additional power. Without a battery in place, the power adapter (and consequently the computer) would instantly shut down as soon as power demand exceeded 85W, with the result that any unsaved work would be irretrievably lost. To prevent that, Apple cuts back the CPU's power demand by downclocking it, bringing the maximum total power demand of the machine below 85W. With older models like yours it was a real rarity for power demand actually to reach 85W, but the precautionary downclocking in the absence of a battery happened nonetheless. With today's MBP's and today's extremely greedy games, it's quite commonplace for power demand to exceed 85W and MBP batteries to discharge even while AC adapters remain plugged in. And temperatures under those conditions go sky-high, too.
Thanks eww. That makes perfect sense.
I have a sneaking suspicion that the heat sink compound interfaces are just degraded somewhat from what their originally shipped condition was and the result is that the small positive thermal margin the design had to begin with is now a negative margin when the CPU is really crunching.
I just have to summon up the ambition to tear it apart and re-do those interfaces. Quite a few hours of work.
pgoodwin, thank you for the post with the service manual link.
mickybannon, I have the same problem.
Thanks to everyone for an interesting post. I have a 2006 2.0 GHz Core Duo MacBook Pro with third-party batteries and heatsinks with fresh thermal grease, polished heat tube plates, and even copper shims on all three chips and new putty on the middle one. Like eww said, my machine is slightly more sluggish when the battery is removed.
Someone else had a similar problem, but didn't have a solution here: https://discussions.apple.com/message/9037094#9037094
Someone here appears to have found a way to make a CPU run quickly without the battery: http://forums.appleinsider.com/t/115180/run-a-macbook-at-full-clock-speed-withou t-a-battery
If anyone has any good suggestions, I would be fascinated to hear them.