*The T60 Fan Thread*
Posted: Tue Dec 19, 2006 11:51 am
Hi guys. I've been looking into the fan noise issue on the T60 for quite some time now and decided to post my findings.
Before I begin I'd just like to let those who are interested know, the two model types I've experimented with are 2613-HKU and 2613-E... I can't remember the last 3 digits, but here are the specs:
One:
Core Duo, ATI X1400, 1GB RAM, 14.1" SXGA, Wireless/Bluetooth, etc.
Other:
Core 2 Duo, ATi X1400, 1GB RAM, 14.1" SXGA Wireless/Bluetooth, etc.
-The machines are identical except for their processors - the chassis numbers are even the same (2613).
The first thing we need to do is differentiate between the T60s with a Core (1) Duo and the newer ones with a Core 2 Duo. The machines are identical with the exception of the processor (and heatsink/fan which I will get into later). Both machines use the same motherboard, the difference is that the Core 2 Duo variant uses a newer BIOS that handles the 64-bit instruction set (BIOS version 2.0 and newer). From here on I will refer to the Core 1 Duo variants as T60A and the newer Core 2 Duo variants as T60B.
There have been a number of people complaining about T60 fan noise while there have been others claiming that the fan is silent. How can it be possible that with the same fan (I will get into this later) some experience silence while others feel as if they are running in 3D mode while then are in fact only on the desktop? Well, it would appear that the T60As are the silent ones and the T60Bs are the noisy ones. It is not due to the fact that Core2Duos are hotter than Core Duos (they are, but only by 1 or 2 degrees), it is due to the BIOSes that accompany these chips. Remember, the T60A and T60B have the same motherboard and accomplanying hardware (with the exception of the CPU) yet have different BIOS versions. The T60As have version 1.X and the T60Bs have version 2.X.
For those not familiar with TP Fan Control I will discuss how it works very quickly. You can let it run in BIOS control mode, which will allow the computer to decide which speed to run the fan at depending on temperature. Or you can specifically tell the machine which speed to run the fans at (varying from 0 to 7, 0 being off and 7 being max). However, in the end these settings are controlled by pre-determined settings in the BIOS. TP Fan Control is simply telling the BIOS what speed setting to run the fan at, not what RPM to run it at.
I have used TP Fan Control and found the following:
*T60A fans idle at ~1900RPM while T60B fans idle at ~3500RPM
*The lowest the underlying BIOS settings will allow the fan to be turned down to is ~3000RPM on a T60B (on a T60A I forget, but it doesn't matter as 1900RPM is barely inaudible).
*On a T60B, settings 1\2 run at ~3000RPM, settings 3\4\5 run at ~3500RPM, and settings 6\7 run at ~4200RPM.
As you can see, although there are 7 speed settings, the BIOS really only lets you choose 3: fast, faster, and max. This is hardly any variation at all, and I find it surprising that when idling (even in the BIOS screen) the BIOS sets the fan to run at "faster" 3500RPM speed instead of "fast" 2900RPM speed.
I forgot to mention, but both machines ran at the same temperatures - their CPU and GPU temps were the same, so it is not a matter of one being hotter and the BIOS kicking the fan speed up on that machine --- both machines have the same temperatures.
As I mentioned earlier, the heatsink/fans are different on the two variants. I've opened both A and B types, as well as taken a look at the IBM documentation and have see that the A types (with non-integrated graphics) use FRU 41V9932 and the B types (with non-integrated graphics) use FRU 41W6407. I called up IBM and asked for a replacement for my T60B heatsink/fan and they sent me the replacement part for a T60A. It looked the same with the exception of a different fan and thermal grease applied in a different orientation. The fan, while the same dimensions, had slightly different fan blades, but was still made by the same manufacturer and produced the same noise levels. The pre-applied thermal grease was applied in = orientation whereas it should have been | | orientation. This is due to the fact that the cores on the Core Duo and Core 2 Duo are orientated 90 degrees from each other.
I nonetheless, wiped the pre-applied thermal grease off, applied Arctic Silver 5, and installed the T60A HSF into my T60B. It fit fine (as the HSFs are dimensionally identical). I did notice that GPU temps (and as a result of them sharing the same copper block, CPU temps) rose a few degrees C with the new HSF. I believe this is due to the fact that the T60A HSF isn't held as snugly into the GPU as the T60B HSF because the hold-downs are different. I believe the stock T60B holddown works best with the T60B HSF and vice-versa for the T60A. So, it's not to say that one HSF performs better than the other - they are both perform the same, it's just that the T60B HSF works best with the T60B holddown and the T60A HSF works best with the T60A holddown. These holddowns give greater contact to those nasty thermal pads IBM uses on the GPUs.
In my quest to try and fix the fan noise on my own, I did a small experiment. Using a replacement HSF, I cut the positive fan power wire (it's the red one for those wondering, the negative is blue, and the RPM reporting one is orange **I take no responsibility for anything you do to your own machine**), and spliced a trimmer potentiometer (a variable resistor) in series with the fan. My goal was to increase the resistance and lower the fan speed until my idle RPMs hit the same levels as a T60A (~1900RPM). This did not work. The machine compensates for the resistance by increasing the voltage to maintain the RPM that it wants the fan to be running at. If I booted the machine with the resistor in place it would simply compensate and run at normal speeds. If I booted the machine with the resistor out and then swapped it in after boot up, the RPMs would drop for a moment, the machine would then compensate and raise the RPMs.
I then tried to see if the machine would even boot if I removed the RPM reporting lead, yet still had the fan's power leads connected -- it didn't like it and wouldn't boot. So the only option I see left for modding the fan is to get some sort of current source or voltage source and attach it to the RPM monitoring lead and supply the machine with bogus RPM reports while putting the resistor in series with the fan, allowing the fan to be run LOWER, yet tricking the machine into thinking the RPMs are normal. Of course, I don't know if this would work and don't really want to try and it will be a pain to figure out how to get it to work properly.
To conclude, the T60A machines are quiet and the T60B machines are loud. Both machines using very similar HSFs whose fans produce the same noise, and operate at the same temparature - yet the T60As fans idle at 1900RPM whereas the T60B fans idle at 3500RPM. This discrepancy must be due to the difference in BIOS - there has been a change to the fan control between BIOS 1.X and BIOS 2.X and the only feasible option to get T60B fans running as quietly as T60A fans is to get a BIOS fix.
I propose contacting IBM, informing them of this discrepancy, and asking for, at least, an option in the BIOS that allows advanced users to switch their fans to "T60A operation mode" which lets them idle at 1900RPM and adds more variance to the fan levels (0-7 on TP Fan Control). This should not have any adverse affects on the machine as temps will remain practically the same. (I should mention that Core2Duo mobiles are rated for up to 100C temperatures -- look on the intel tech sheets if you don't believe me).
To that extent: does anyone know how to go about contacting IBM/Lenovo and politely asking for a fix? I don't know about the rest of you, but I'm fed up with my T60B. Between the fan noise and the high-pitched noise coming from the GPU, this is one troubled machine.
---
No, I did not try to upgrade my T60A to 2.X BIOS and see if the fan noise increased, but I believe someone on the forums did so and said it did.
Yes, I did use a T60A HSF in a T60B and experienced the normal T60B noise level.
Maybe you haven't noticed it, but on a T60A and T60B I've noticed a high-pitch whine. I attributed it to the wireless card, but someone in this thread seems convinced that it is the GPU (http://forum.thinkpads.com/viewtopic.php?t=34761)
Before I begin I'd just like to let those who are interested know, the two model types I've experimented with are 2613-HKU and 2613-E... I can't remember the last 3 digits, but here are the specs:
One:
Core Duo, ATI X1400, 1GB RAM, 14.1" SXGA, Wireless/Bluetooth, etc.
Other:
Core 2 Duo, ATi X1400, 1GB RAM, 14.1" SXGA Wireless/Bluetooth, etc.
-The machines are identical except for their processors - the chassis numbers are even the same (2613).
The first thing we need to do is differentiate between the T60s with a Core (1) Duo and the newer ones with a Core 2 Duo. The machines are identical with the exception of the processor (and heatsink/fan which I will get into later). Both machines use the same motherboard, the difference is that the Core 2 Duo variant uses a newer BIOS that handles the 64-bit instruction set (BIOS version 2.0 and newer). From here on I will refer to the Core 1 Duo variants as T60A and the newer Core 2 Duo variants as T60B.
There have been a number of people complaining about T60 fan noise while there have been others claiming that the fan is silent. How can it be possible that with the same fan (I will get into this later) some experience silence while others feel as if they are running in 3D mode while then are in fact only on the desktop? Well, it would appear that the T60As are the silent ones and the T60Bs are the noisy ones. It is not due to the fact that Core2Duos are hotter than Core Duos (they are, but only by 1 or 2 degrees), it is due to the BIOSes that accompany these chips. Remember, the T60A and T60B have the same motherboard and accomplanying hardware (with the exception of the CPU) yet have different BIOS versions. The T60As have version 1.X and the T60Bs have version 2.X.
For those not familiar with TP Fan Control I will discuss how it works very quickly. You can let it run in BIOS control mode, which will allow the computer to decide which speed to run the fan at depending on temperature. Or you can specifically tell the machine which speed to run the fans at (varying from 0 to 7, 0 being off and 7 being max). However, in the end these settings are controlled by pre-determined settings in the BIOS. TP Fan Control is simply telling the BIOS what speed setting to run the fan at, not what RPM to run it at.
I have used TP Fan Control and found the following:
*T60A fans idle at ~1900RPM while T60B fans idle at ~3500RPM
*The lowest the underlying BIOS settings will allow the fan to be turned down to is ~3000RPM on a T60B (on a T60A I forget, but it doesn't matter as 1900RPM is barely inaudible).
*On a T60B, settings 1\2 run at ~3000RPM, settings 3\4\5 run at ~3500RPM, and settings 6\7 run at ~4200RPM.
As you can see, although there are 7 speed settings, the BIOS really only lets you choose 3: fast, faster, and max. This is hardly any variation at all, and I find it surprising that when idling (even in the BIOS screen) the BIOS sets the fan to run at "faster" 3500RPM speed instead of "fast" 2900RPM speed.
I forgot to mention, but both machines ran at the same temperatures - their CPU and GPU temps were the same, so it is not a matter of one being hotter and the BIOS kicking the fan speed up on that machine --- both machines have the same temperatures.
As I mentioned earlier, the heatsink/fans are different on the two variants. I've opened both A and B types, as well as taken a look at the IBM documentation and have see that the A types (with non-integrated graphics) use FRU 41V9932 and the B types (with non-integrated graphics) use FRU 41W6407. I called up IBM and asked for a replacement for my T60B heatsink/fan and they sent me the replacement part for a T60A. It looked the same with the exception of a different fan and thermal grease applied in a different orientation. The fan, while the same dimensions, had slightly different fan blades, but was still made by the same manufacturer and produced the same noise levels. The pre-applied thermal grease was applied in = orientation whereas it should have been | | orientation. This is due to the fact that the cores on the Core Duo and Core 2 Duo are orientated 90 degrees from each other.
I nonetheless, wiped the pre-applied thermal grease off, applied Arctic Silver 5, and installed the T60A HSF into my T60B. It fit fine (as the HSFs are dimensionally identical). I did notice that GPU temps (and as a result of them sharing the same copper block, CPU temps) rose a few degrees C with the new HSF. I believe this is due to the fact that the T60A HSF isn't held as snugly into the GPU as the T60B HSF because the hold-downs are different. I believe the stock T60B holddown works best with the T60B HSF and vice-versa for the T60A. So, it's not to say that one HSF performs better than the other - they are both perform the same, it's just that the T60B HSF works best with the T60B holddown and the T60A HSF works best with the T60A holddown. These holddowns give greater contact to those nasty thermal pads IBM uses on the GPUs.
In my quest to try and fix the fan noise on my own, I did a small experiment. Using a replacement HSF, I cut the positive fan power wire (it's the red one for those wondering, the negative is blue, and the RPM reporting one is orange **I take no responsibility for anything you do to your own machine**), and spliced a trimmer potentiometer (a variable resistor) in series with the fan. My goal was to increase the resistance and lower the fan speed until my idle RPMs hit the same levels as a T60A (~1900RPM). This did not work. The machine compensates for the resistance by increasing the voltage to maintain the RPM that it wants the fan to be running at. If I booted the machine with the resistor in place it would simply compensate and run at normal speeds. If I booted the machine with the resistor out and then swapped it in after boot up, the RPMs would drop for a moment, the machine would then compensate and raise the RPMs.
I then tried to see if the machine would even boot if I removed the RPM reporting lead, yet still had the fan's power leads connected -- it didn't like it and wouldn't boot. So the only option I see left for modding the fan is to get some sort of current source or voltage source and attach it to the RPM monitoring lead and supply the machine with bogus RPM reports while putting the resistor in series with the fan, allowing the fan to be run LOWER, yet tricking the machine into thinking the RPMs are normal. Of course, I don't know if this would work and don't really want to try and it will be a pain to figure out how to get it to work properly.
To conclude, the T60A machines are quiet and the T60B machines are loud. Both machines using very similar HSFs whose fans produce the same noise, and operate at the same temparature - yet the T60As fans idle at 1900RPM whereas the T60B fans idle at 3500RPM. This discrepancy must be due to the difference in BIOS - there has been a change to the fan control between BIOS 1.X and BIOS 2.X and the only feasible option to get T60B fans running as quietly as T60A fans is to get a BIOS fix.
I propose contacting IBM, informing them of this discrepancy, and asking for, at least, an option in the BIOS that allows advanced users to switch their fans to "T60A operation mode" which lets them idle at 1900RPM and adds more variance to the fan levels (0-7 on TP Fan Control). This should not have any adverse affects on the machine as temps will remain practically the same. (I should mention that Core2Duo mobiles are rated for up to 100C temperatures -- look on the intel tech sheets if you don't believe me).
To that extent: does anyone know how to go about contacting IBM/Lenovo and politely asking for a fix? I don't know about the rest of you, but I'm fed up with my T60B. Between the fan noise and the high-pitched noise coming from the GPU, this is one troubled machine.
---
No, I did not try to upgrade my T60A to 2.X BIOS and see if the fan noise increased, but I believe someone on the forums did so and said it did.
Yes, I did use a T60A HSF in a T60B and experienced the normal T60B noise level.
Maybe you haven't noticed it, but on a T60A and T60B I've noticed a high-pitch whine. I attributed it to the wireless card, but someone in this thread seems convinced that it is the GPU (http://forum.thinkpads.com/viewtopic.php?t=34761)