Thinkpad X201 - HOT Core i5 2.4 Ghz ... only runs at 1.2Ghz without battery COOL

[DaveLembke]

I have a IBM Thinkpad X201 with 2.4Ghz Core i5 and laptop and the power supply were both getting VERY hot. I decided to remove the battery and see if it would run cooler if the laptop was running at the same time and not having to charge battery. Sure enough the laptop was cooler no longer 3rd degree burns to hands or lap at bottom of laptop and power supply was no longer feeling like it was going to melt into a pool of molten plastic. All temps back to how a laptop should feel for heat output...

 BUT I noticed that the BOINC Asteroids program that I run on it when idle now was saying that it was going to take 10 hours to complete 4 project tasks instead of normally less than 5 hours.

Its running Linux Mint 17.3 Cinnamon and I installed i7z to it and it shows the core use and frequency and sure enough its running full tilt 100% CPU use, but only at 1200Mhz. I shut it down and install battery and bring it back online and its back to 2400Mhz with i7z.

This is the strangest thing because I never expected a laptop to downclock itself if no battery was present. Looking online I saw some stuff about going into the BIOS and setting performance to MAX on AC power. Did that and whats strange is that before it boots the OS I can hear the fan racing and pumping out heat and as soon as the Linux Mint 17.3 OS loads it calms so not sure if its hardware or OS related as to the downclock to 1200Mhz from 2400Mhz when running without battery.

Thing is if I get the CPU back to 2.4Ghz without battery it might roast again so I'm sort of split in the decision to find a work around to where there is no battery having to charge adding load to the power supply and laptop while the laptop is also in use or just leave it be and let it run cooler at 1.2Ghz and take 2x longer to complete each work unit for the BOINC project.

Figured I'd share this hear to make others aware of this underclock feature that I guess Thinkpads do when powered with no battery present as well as get some input on suggestions of how to get the clock back to 2.4Ghz without battery or a way to fix the fact that both the laptop and its power supply roast when running full tilt at 2.4Ghz with this Core i5 520M.

*Saw that IBM has a battery recall for fires but the part number of my battery isnt included in the recall, however a part number of 42T4539 and 42T4648 there is a 42T46xx battery that is part of recall so thats why i wanted to do without the battery because maybe the problem runs earlier in production and not looking to have a Lithium Fire and lose the house etc.

** Other thing to mention is that the fan for CPU only kicks on when its roasting hot and needs to cool itself. I was looking around for a way to force the fan to run 100% of the time at full speed. Opened up the laptop and applied new arctic silver thermal compound thinking maybe that would help and no change. Tempted to snip the 5 volt wire and solder it to a 5 volt source on the main board to force it to stay on all the time when powered as for it goes through hot and cold cycles. If it was running all the time it would force max airflow across all components all the time and should level out the temp of the laptop at a cooler temp than 3rd degree burn hot and then fan kicks in to cool and then laptop not so burning hot to touch and then it heats back up again and  goes through this blistering hot and uncomfortably warm cycle when battery connected with or without wall power. Looking at CPU temp it rises into the 76C range and then drops to 55C when fan kicks on at 2.4Ghz. At 1.2Ghz the laptop is running between 43C and 55C. Not sure what the emergency shutdown temp is for the CPU in this small laptop but I think 80C is the starting temp threshold for some systems to alarm and shutdown at 85C. No option in BIOS to adjust or see what its shutdown temp is rated/set for. For the fact that the CPU reports the temps that seem correct the thermocouple must be ok for the temp monitoring for the laptop. I have heard of in systems before where a thermocouple is dead or not correct and it allows for systems to roast.

So thats all I have on this, looking for peoples feedback and suggestions on this one. Maybe its normal for the IBM X201 with Core i5 520M to roast, I see lots of complaints about people cooking their laps with them and people saying its not normal and others saying its normal and a flaw in this production line yet IBM isnt doing any recall about it and suggestions range from selling laptop to give the problem to someone else to shotgunning the problem by taking a chance at replacing main board which might fix the heat issue.

I got this laptop given to me by a person who upgraded to a Core i7 and larger laptop, and didnt want to pay their local dump $20 to dispose of it. When asking why throw about this laptop, the one and only complaint they had when they gave me it was that it runs HOT! But not hot enough to shut itself down. Budget to correct for this if buying parts is a solution is less than $50. Its not a critical computer and I have many computers. End result may be just to run it at 1.2Ghz and leave it be. Undecided with that but willing to try out solutions with it. 

[BC_Programmer]

For the battery issue, this is inherent to Thinkpads, as you've found. The battery is basically a critical part of the power delivery system and without it the capabilities are a bit lobotomized:

The power adapter and battery together form the power delivery system for your notebook. Removing the battery reduces its ability to support heavy loads, so the notebook throttles back the maximum CPU speed. In some cases, changing to a higher wattage power adapter may eliminate this throttling. There is a higher risk of system corruption or damage through accidental disconnection of the power adapter.

source.

Allegedly, a higher wattage supply fixes this.

The other issue- the temp issue- I think is more interesting. I have a T550 (2.6Ghz i7-5600U) and can't say I've ever noticed high temperatures. I don't run any distributed computing software however. Just as a test I'm running Prime95 on it now, but after a half hour it still only feels what I'd call warm (both the brick and the laptop).

You don't mention it but have you verified that it's not say clogged with dust inside? Also Some search results I f ound where others have the same issue with that model have claimed the problem was resolved by removing and re-applying thermal paste, as well.

[DaveLembke]

You don't mention it but have you verified that it's not say clogged with dust inside? Also Some search results I f ound where others have the same issue with that model have claimed the problem was resolved by removing and re-applying thermal paste, as well.

I typed a lot in the initial post... sometimes I type TMI, but best to cover everything at the get go so everyone knows where Im at with it. I guess below got burried in my lengthy read. Here is the section that I covered looking for issues to the heat issue. Its all clean inside and fan spins correct RPMs as tested with speedfan when i had Windows 10 trial on it for a short while.

** Other thing to mention is that the fan for CPU only kicks on when its roasting hot and needs to cool itself. I was looking around for a way to force the fan to run 100% of the time at full speed. Opened up the laptop and applied new arctic silver thermal compound thinking maybe that would help and no change.

I didnt catch the part about running a higher wattage supply. Going to see if I have a higher wattage Thinkpad supply available with the same voltage. I picked up some IBM thinkpad power supplies at a yard sale fund raiser at local church for $3 this summer, maybe one of those is higher than the 65 watt 20 volt thats connected to it now.

I figured at $3 even if just one of the power supply bricks works it was worth it. I have also used other make power supply bricks with off brand laptops by chopping the cord end off of a dead power supply and adding it to a healthy one to get laptop working again of similar amp/wattage and voltage. I was given a Gateway laptop that owner said it was dead. Tested the power supply and no output but rated at 19 volts. I had a HP that was given to me where the 19V power supply was good but the laptop was actually junk and so I used that power supply for the gateway laptop, so that the Gateway came back to life on the HP power supply with Gateway plug end added to it with correct polarity and shrink tubed connections and electrical tape. 

Thanks BC for checking into this and pointing out the higher wattage supply as a possible solution maybe for powering it without battery.

[Geek-9pm]

A room air conditioner is more cost effective.
Compared to purchase of  new high performance laptop.
Keep the room as cold as you can bear it.
Wear a coat.
Drink hot chocolate. 

[patio]

Really Geek ? ?

[Geek-9pm]

Yes, Patio.
A 15,000 BTU window model is under $200 at some retail outlets.
In a small room  it can keep the ambient temp down to 60 F or 16C and will keep the laptop cool. Put the laptop in the air flow.  It really works.

[BC_Programmer]

An air conditioner doesn't address the problem. The problem is that the laptop gets hot. The laptop is going to get hot regardless of the ambient temperature. It's like suggesting a rowboat to address a flooded basement.

[DaveLembke]

Well as wild as this might sound... I am going to add extra airflow to the laptop and see if thats a solution to running it full tilt.

I ran it with battery installed and it runs in a Turbo mode at 2.530Ghz for the 2.4Ghz Core i5 CPU and pumps out serious heat. Since this laptop is really just for crunching data, I am going to either alter its case removing the bottom section and drill out a section and attach one of these fans to it to add airflow far an beyond what it was even intended to have. This should draw in cooler air across the main board and pump out the heated air out the bottom of it. ( OR ) I will attach this fan to the exhaust vent of the laptop to draw additional airflow through the heatsink which just like the vacuum hose sucking on the vent cooled the CPU and GPU and caused the clock to climb back up as it appeared that it was thermal throttling when hot.

**NOTE: Do Not Try This At Home with Vacuum Cleaner hose connected to side of laptop to draw away heat... I only did this because I was testing to see what would happen if I cooled it with extreme airflow and willing to take the risk of a static discharge killing the system as well as the small fan inside was likely being spun up to very fast RPMs from the couple thousand CFMs etc. Luckily no static discharge and no damage to fan which still spins and pumps out excess heat.


http://www.newegg.com/Product/Product.aspx?Item=9SIA0ZX2C28713&cm_re=slot_fan-_-35-129-025-_-Product

Watching the CPU frequency with i7z as the laptop starts to roast it starts to downclock ( thermal throttle ) from the Turbo speed of 2.530 Ghz to around 2.1 Ghz and as it gets even hotter to 1.8Ghz where it levels off at and hot air coming out the side vent hot enough to burn your skin if left there. When I took the vacuum cleaner to the cpu exhaust vent to draw air through it to suck away the heat the clock climbed back up to 2.53Ghz Turbo clock for the Core i5. Removed the Vacuum and as it heated up it started to downclock again to thermal throttle.

With new thermal compound already added after removing the old and still having heat issues, Im thinking adding additional CFMs of airflow for just a cruncher laptop that is more of a tabletop, should do the trick. I have some extra 12VDC power supplies that I can wire the fan into so the fan is separately powered.

[Calum]

Maybe its normal for the IBM X201 with Core i5 520M to roast

Mine doesn't, if that helps at all.  It gets no warmer than any other laptop in its class, certainly not uncomfortable to touch under normal usage, and even under extended full load it doesn't get to the point that it's too hot to touch, max CPU temp is in the mid/high 80s.  Prior to re-pasting it would hit the low 90s and the fan would be much louder, it's a lot quieter now and somewhat cooler under full load, the difference in normal usage is more pronounced.
Also, for the PSU, there's very little you can do about how hot it gets as that's just a function of the power being drawn.  You could potentially look for a higher wattage charger which should run a little cooler if it bothers you.

[DaveLembke]

Prior to re-pasting it would hit the low 90s and the fan would be much louder

hmm... mine only runs the fan full tilt at boot when the BIOS splash screen shows. Then the fan shuts off. System boots and then the fan turns on again for a bit and shuts off and then as it gets warmer the fan eventually stays on constant, but never gets to the same speed that it did at the initial boot when the fan ran full tilt. Laptop never shuts itself down on a thermal or fan issue of not spinning etc. 

Gonna see where the extra airflow gets me with this. The laptop to me has no real value to it since I got it for free as it was given to me and I just want to use it as a cruncher because the Core i5 that it has does really well for less than 50 watts of power draw as measured by a kil-a-watt device and how fast its able to crunch data for BOINC.

Thanks for everyones input on this one. 

[Geek-9pm]

New  thermal compound is not better than tooth paste. It does nothing.
Try this:
Stick some thermal compound in your mouth...
Does it get cooler?

Air flow does make a difference.

[DaveLembke]

New  thermal compound is not better than tooth paste. It does nothing.
Try this:
Stick some thermal compound in your mouth...
Does it get cooler?

Geek ...  While most of your stuff is good, some is out there, this is just like... why post this.  I think you need a "DO NOT TRY THIS AT HOME ... IM A TRAINED PROFESSIONAL" under your statement that you have put thermal compound in your mouth before to suggest this. 

The thought of that is about as nasty as the trick I saw on youtube with someone swapping the tooth paste on someone with diaper cream and filming their friend brushing with diaper cream. Zinc Oxide is not good for the mouth, and in this case most thermal compounds use a metal like aluminum which isnt good for mouth.

[BC_Programmer]

New  thermal compound is not better than tooth paste. It does nothing.

Some thermal compounds conduct heat away from the heat spreader faster than others. The main reason for repasting is less to use a better thermal paste and more to simply replace the existing paste, as over time thermal paste can dry up, bubbles can form, etc that reduce the amount of heat that can be conducted to the Heatsink (and away from the CPU). repasting ensures the contact is good.

Toothpaste is better than nothing when it comes to thermal compound, but only in the same sense that Saran Wrap is better than nothing as a  form of birth control.

[Geek-9pm]

No argument with BC.

Zinc Oxide is not good for the mouth, and in this case most thermal compounds use a metal like aluminum which inst good for mouth.

You are right. I should not have said that.

Back to topic.
 It is very hard to improve the air flowing a laptop computer. Thermal paste makes no significant improvement. Whatever was used by the OEM  is near optimal.
To  reduce temps in a laptop some effort to use external cooling is needed. Very high volume of air might damage the fans inside the laptop. The alternative is to use cool air forced into the laptop with moderate pressure.

In other words, put the laptop in a small area shared with air conditioner. Allow the output of the air conditioner to blow cool air into the offices of the laptop.
Links:
http://www.tomsguide.com/answers/id-1790527/laptop-cooling-gaming-temp.html
Gamer cools down his laptop.

http://gadgetwise.blogs.nytimes.com/2009/05/19/10-ways-to-cool-down-your-laptop/
10 Ways to Cool Down Your Laptop - The New York Times

This has been widely documented elsewhere. I did NOT originate idea.
For a MacBook:
http://computers.tutsplus.com/tutorials/keeping-your-cool-with-apple-mac-operating-temperatures--mac-52077
 

Air Conditioning
The converse to sitting in the sun, working in an air conditioned office means a lower ambient temperature which means that your Mac is not soaking up lots of external or ambient heat.

So, I was not joking about the air conditioner.   

[Salmon Trout]

New  thermal compound is not better than tooth paste. It does nothing.

This toothpaste thing has been going around for years. I think one of the original sources was a thermal compound comparison at Dan's Data in 2002. They compared a number of compounds available at that time - Arctic Silver, Arctic Silver 3, Cooler Master PTK-001 and HTK-001, Nanotherm Ice and Nanotherm Blue, and just for laughs, a brand of toothpaste, and a food spread called Vegemite (the site is Australian). They also tried no compound at all and a generic unbranded sachet of zinc oxide gel often found included with heatsinks. The test setup was a solid copper 'heater block' with a heating element inside to mimic the CPU. On this was stuck a Thermaltake Volcano 6Cu+ cooler, with the various compounds (or no compound) in between. With a dual-input thermometer they could see the ambient air temperature, the temperature of the probe inside the heater, and the difference between the two (the delta-T). Thus they could see how the cooler was managing to get rid of the heater's output, without the result being polluted by changes in the ambient temperature. Divide the delta-T by the heater wattage, and you get a number in degrees Celsius per watt (°C/W). The lower the number, the better.

The results were that no goop at all (bone dry) gave a result of 0.66 °C/W, while all the heatsink compounds gave results between 0.48 and 0.56 (the spread of values is practically negligible - they all performed the same in practical terms). Then for laughs they did the tooth paste and vegetable spread, 0.47 (toothpaste) and 0.48 (Vegemite). So toothpaste was better than nothing at all, and about as good as the thermal compounds they chose. This is probably the source of the oft-repeated allegation that "toothpaste works as well as thermal compound". In fact you can probably find any number of goops in your kitchen, bathroom or garage* that would work about the same for a while, and that's the point that it is missed or ignored by the armchair warriors. In a week the toothpaste was beginning to dry up. Someone on Super User said they'd been using a rebuild with toothpaste on their Thinkpad T42 for about 3 years, with no problems, but they could have been bullsh*tting. Those Thinkpads did run cool.

*These have all been tried since by others: toothpaste, antiseptic gel, dielectric silicone, hair styling clay, peanut butter, raspberry preserves, anti-seize lubricant, high temperature grease, lip balm, liquid gel deodorant. Peanut butter is terrible. Some of them smell weird when they get warm. None of them last.