HARD DRIVE AND CPU TEMPERATURES ON A
2.80C GHz PENTIUM 4 PC SYSTEM
AND HARD DRIVE TEMPERATURES
IN EXTERNAL USB ENCLOSURES

Author:  Mike Boesen  

Last updated: 11 February 2008

1.  Background

This article is a follow up to an article I wrote for the PC Users Group magazine Sixteen Bits, describing applications that are useful in monitoring the temperatures of the CPU, motherboard and hard drives.  The most up to date version of that article is here: http://www.pcug.org.au/~boesen/temperature_software/temperature_software.htm ).   I used the applications when undertaking a number of tests on my PC to see how different circumstances affected the temperatures of hard drives, the CPU and motherboard.  This new article summarises the results of those tests.  I'll incorporate any improvements that people might suggest in an up to date version that is kept on my web site.

What lead me to become interested in temperatures was assistance that I gave to a number of PC users who were having problems with their PCs.  I felt that in a few problem cases the possible cause was overheating.  Initially, my initial concern related to hard drives and the CPU.  

To be better informed about this topic I undertook a number of tests with my own PC.  In these tests I observed the temperatures of the CPU, motherboard and hard drives under a number of conditions:  different fan fan speeds, fans on and off, drive mounted internally and drive in removable caddy, idling and loaded conditions.  I also made a number of tests of 2.5" and 3.5" drives in external USB enclosures.

Taking into account the tests and my experience with other persons' PCs and from information that is available on the web, I am now of the view that it is unlikely that CPU temperature problems will occur for the vast majority of PC users.  On the very few occasions that I have encountered CPU overheating problems it was because the CPU cooler had been installed inappropriately.  On the other hand, I suspect that hard drive temperatures may possibly be a cause of some problems and that the life of some hard drives are likely to be shorter than they might otherwise be. 

2.  How hot is too hot?

2.1  CPU temperature

Modern CPUs are built to operate quite efficiently even when they are quite hot.  For the Pentium 4 series for instance, Intel's stated maximum "case temperatures" are between 66 and 78 C.  The "case temperature" is the temperature measured at the centre of the CPU's heat spreader - NOT the temperature of the PC's case.  For my P4 2.80c CPU, Intel indicates that the maximum temperature is 74 C.   Table 2 in this Intel article about Pentium 4 processors provides details about maxima for various versions of the P4:  http://support.intel.com/support/processors/pentium4/sb/CS-007999.htm  An abridged version of the table is below in Appendix 1.  

CPU temperatures will change rapidly in accordance with how much work the CPU is undertaking.  A large range in temperatures for a CPU is quite normal and depends on the amount of work being undertaken, the ambient air temperature, the type and effectiveness of the CPU's cooling system and the ventilation conditions within the PC's case. 

In the absence of specific information about the manufacturer's maximum operating temperature for a CPU,  I would be concerned if a CPU's temperature exceeded 65 C when the CPU is under heavy load and the ambient temperature is 27 C or less. 

However, even if the manufacturer's maximum operating temperature is exceeded, a modern CPU such as the Intel Pentium 4 series is designed to slow itself down automatically, reducing power consumption and hence limiting the overheating to a tolerable level.  This is referred to as ''thermal throttling'' or simply ''throttling''.  When such CPUs get really hot - around 95 to 100 C they are designed to shut down completely.  Infrequent shutdowns of this nature should not affect the operating life of the CPU but frequent shut downs can affect the life of the CPU.  

If you want to see some nice charts showing throttling in action, have a look at this article:  
http://www.digit-life.com/articles2/p4-throttling/index.html   The testing conducted is described thus by the authors:  ''...we conducted a very simple experiment on two top Pentium 4 3.2 GHz CPUs, one with a Northwood core and the other with a Prescott.  With the CPU loaded at 100 percent, we powered off the cooler fan and watched the temperature rise and the performance decrease as time went by.''

There are a number of possible causes of excessively high CPU temperatures.  My servicing checklist is this:

2.2  Hard drive temperature

Information available on the web about the effect of heat on hard drive reliability suggests that heat affects the reliability of the operation of a hard drive's components and its service life.  For instance, this statement is made by one author: '' Moreover, reliability and durability of these drives depends much on their operating temperatures. According to our research, increasing HDD temperature by 5°C has the same effect on reliability as switching from 10% to 100% HDD workload! Each one-degree drop of HDD temperature is equivalent to a 10% increase of HDD service life.''  ( http://www.digit-life.com/articles2/storage/hddpower.html )

The rotation speed of the platters in a hard drive has a very marked effect on temperature.  This is because the rotating platters in the drive generate heat through air friction and the faster they spin, the more heat is generated.  Therefore modern 7,200 rpm drives can generate a lot of heat that needs to be dissipated.  Heat is also generated by the spindle motor and other electronic components.  

Hard drive manufacturers provide information about the claimed minimum and maximum operating temperature for their hard drives.  These maxima vary depending on the brand and model of the drive.  If you really want to know what the maximum operating temperature is for a hard drive, you will need to do a Google search.  For example:

I suspect that the setting of maxima may be a little arbitrary and that a manufacturer would be unable to provide a really scientific basis for the temperature stated.  However, what seems to be generally agreed is that the service life and reliability of a hard drive will be affected by the temperature of the drive,  irrespective of whatever the claimed maximum operating temperature is.  My own rules of thumb (rule of thumbs?) are: 

If excessively hot drive temperatures are noted there are a number of possible causes.  Virtually all are related to inadequate air flow over the hard drive.  The importance of achieving such ventilation cannot be overemphasised.  To improve ventilation I'd recommend you consider these solutions: 

3. Testing for hard drives in the PC

3.1 Test conditions

The PC that I undertook my tests on had these components:

I monitored temperatures of my drives and CPU using the freeware application Motherboard Monitor 5.  (See details in http://www.pcug.org.au/~boesen/temperature_software/temperature_software.htm )

All my tests were conducted with the speed of the CPU cooler's fan slowed to 2,150 rpm (versus the maximum possible of 2,660).  Because of the efficiency of the Zalman cooler design and my use of two case fans I can run the cooler fan at this slower speed maintaining the CPU at an acceptable temperature, with virtually inaudible cooler fan noise.  

I monitored temperatures under various conditions:

My tests were undertaken in January 2006 (i.e. Summer here in the antipodes) and during days when the ambient temperature in the computer room was between 25 C and 27 C. 

3.2 Test results for hard drives in the PC

I won't bore you with the results of all the tests (the stuff below is boring enough).  However, here are the highlights, presented in summary form. 

3.2.1  CPU temperature results

I also undertook testing with the case fans running at slowed speed.  The results are consistent with the patterns in the above results, with temperatures being only about 2 C lower than those above.

3.2.2  Internal hard drive temperature results

I also undertook testing with the case fans running at slowed speed.  The results are consistent with with the patterns in the above results, but the cooling effectiveness is about 5 C less than with a fan or fans running at full speed. 

3.2.3  Removable hard drive temperature results

So:

I also undertook testing with the case fans running at slowed speed.  The cooling effectiveness is about 2 C less than with a fan or fans running at full speed.

4. Testing for hard drives in USB 2 enclosures

have undertaken a few tests of hard drives in USB 2 High Speed enclosures.

4.1 Test conditions

One test was undertaken with this configuration:

A second test was with this configuration:

A third test was with the same configuration for the 3.5" hard drive test except that the ventilation fan was enabled.

Both drives were exercised using a batch file application that wrote about 12 GB of data in 2,200 files to the drive, then deleted all the files.  These two operations were conducted continuously in a loop that ran until the drive reached its maximum temperature (after about 2 hours of continuous operation).   While this test may appear to be extreme, it is no more extreme than when making a full backup of a partition or drive using an application such as Acronis.

4.2 Test results for the 2.5" drive

With the stress test running the small drive in its enclosure reached a stable warmed up temperature as follows:

That means that the drive temperature was:

Assuming that the difference between the ambient room temperature and the drive temperature is a constant 17 C irrespective of the ambient temperature level, then the temperature of the drive would reach the manufacturer's stated maximum operating temperature of 55 C only if the ambient room temperature were to reach a very hot 38 C.  

I can't recall the temperature in my study ever exceeding 30 C and in that event the drive temperature would be 47 C - comfortably below the maximum operating temperature of 55 C. 

I also tried the stress test with the enclosure balanced on its thinner edge so that there would be better ventilation.  The drive temperature was about 4 C less than when the enclosure was flat on a book.   However, to get that 4 C benefit, I'll need to make a little stand so that the enclosure is stable when positioned on its edge.  Two folded bits of aluminium bar would do the trick.

With or without such a stand, the temperature performance of this 2.5" hard drive should be acceptable irrespective of what ambient temperatures I am likely to be experiencing.

As an aside, I left the drive plugged in at the cessation of the stress test.  With the drive still connected to power but with no disk read/write activity it cooled down after some time and the temperature stabilised at a level that was 9 C lower than the hottest temperature noted during the stress test - that was only 8 C warmer than the ambient temperature.

4.3  Test results for the 3.5" drive with the ventilation fan disabled

With the stress test running continuously, and the enclosure's ventilation fan disabled the large drive reached a stable warmed up temperature as follows:

That means that the drive temperature was:

Assuming that the difference between the ambient room temperature and the drive temperature is a constant 26 C for all ambient temperature levels, then the temperature of the drive would reach the maximum recommended operating temperature if the ambient temperature were to reach 34 C. 

The temperature in my study has never exceeded 30 C and in that event the drive temperature would be 56 C - close to the manufacturer's stated maximum operating temperature of 60 C.  The small margin leaves me with reservations about the wisdom of using such an enclosure on a very hot day if it has no exhaust fan. 

It should be noted that even if the drive is just idling (that is no read/write activity occurring at all), I found that its temperature stabilised at a surprisingly warm level - 22 C hotter than the air in the room.

4.4  Test results for the 3.5" drive with the ventilation fan operating

I tried the same stress test for the 3.5" drive with the USB enclosure's exhaust fan enabled.  The drive reached a stable warmed up temperature as follows:

That means that the drive temperature was:

Assuming that the difference between the ambient room temperature and the drive temperature is a constant 11 C for all ambient temperature levels, then the temperature of the drive would reach the maximum recommended operating temperature only if the ambient temperature were to reach an impossibly hot 49 C. 

The temperature in my computer room has never exceeded 30 C and in that event the drive temperature would only be be 41 C - way below the manufacturer's stated maximum operating temperature of 60 C.  This large margin demonstrates the value of having an enclosure that has an exhaust fan if the drive is a large 3.5" drive. 

5.  Conclusions

5.1 Caveats

Keep in mind that my conclusions reflect results for a system that:

5.2  Conclusions about my own PC

1.  CPU temperature: 

2.  Internal hard drive under heavy load: 

3.  removable hard drive in a caddy when under heavy load: 

4.  In the light of the above findings, I am now confident that this hardware configuration will ensure that the temperatures for my internal drives and the CPU will never exceed the maxima, and there will be very little fan noise:

5. I am confident that the temperature of the 2.5" hard drive that I tested in a non-ventilated USB enclosure would not be a concern, even if it were running under stress such as that which would occur if the drive were used to make a drive backup or clone and on a hot summer's day.

6.  I have concerns about the appropriateness of using the 3.5" hard drive that I tested in a USB enclosure with its ventilation fan disabled.  While it would be OK if the ambient temperature were say, about 22 C, on a hot summer's day with an ambient temperature of say, 30 C, the temperature of the drive would be about 56 C - close to the maximum operating temperature specified by the manufacturer.

7.  With fan ventilation of the enclosure enabled, I am confident that the temperature of the 3.5" hard drive that I tested would never be an issue, even if it were under stress such as that which would occur if the drive were used to make a drive backup or clone and on a hot summer's day.

5.3  Matters that other PC users may wish to consider

Here are some matters for others to consider in respect of their PCs:

Mike Boesen


Appendix 1Boxed Intel Pentium 4 Processor Thermal Specifications

Source:  http://support.intel.com/support/processors/pentium4/sb/CS-007999.htm

Note:  This table is an abridged version of the original table in the source.

Processor Core Frequency (GHz)
Processor Package
Maximum Case Temperature (oC)
Maximum Recommended Fan Inlet Temperature
(oC)
Processor Thermal Design Power (W)
1.30
423-pin OOI
69
40
48.9
1.30
423-pin OOI
70
40
51.6
1.40
423-pin OOI
70
40
51.8
1.40
423-pin OOI
72
40
54.7
1.40
478-pin FC-PGA
72
40
55.3
1.40
478-pin FC-PGA
72
40
55.3
1.50
423-pin OOI
72
40
54.7
1.50
423-pin OOI
73
40
57.8
1.50
423-pin OOI
73
40
57.8
1.50
478-pin FC-PGA2
73
40
57.9
1.50
478-pin FC-PGA2
73
40
57.9
1.50
478-pin FC-PGA2
71
40
62.9
1.60
423-pin OOI
75
40
61.0
1.60
423-pin OOI
75
40
61.0
1.60
478-pin FC-PGA2
75
40
60.8
1.60
478-pin FC-PGA2
75
40
60.8
1.60A
478-pin FC-PGA2
66
40
46.8
1.70
423-pin OOI
76
40
64.0
1.70
423-pin OOI
76
40
64.0
1.70
423-pin FC-PGA2
76
40
63.5
1.70
478-pin FC-PGA2
76
40
63.5
1.70
478-pin FC-PGA2
73
40
67.7
1.80
423-pin OOI
78
40
66.7
1.80
423-pin OOI
78
40
66.7
1.80
478-pin FC-PGA2
77
40
66.1
1.80
478-pin FC-PGA2
77
40
66.1
1.80A
478-pin FC-PGA2
67
40
49.6
1.90
423-pin OOI
73
40
69.2
1.90
478-pin FC-PGA2
75
40
72.8
2
423-pin OOI
74
40
71.8
2
478-pin FC-PGA2
76
40
75.3
2A
478-pin FC-PGA2
68
40
52.4
2.20
478-pin FC-PGA2
69
40
55.18
2.26
478-pin FC-PGA2
70
40
56.0
2.40
478-pin FC-PGA2
70
40
57.8
2.40A
478-pin FC-PGA2
69.1
38
89
2.40B
478-pin FC-PGA2
71
40
59.8
2.40C
478-pin FC-PGA2
74
40
66.2
2.50
478-pin FC-PGA2
72
40
61.0
2.53
478-pin FC-PGA2
71
40
59.3
2.53
478-pin FC-PGA2
72
40
61.5
2.60
478-pin FC-PGA2
72
40
62.6
2.60C
478-pin FC-PGA2
75
40
69.0
2.66
478-pin FC-PGA2
74
40
66.1
2.80
478-pin FC-PGA2
75
40
68.4
2.80A
478-pin FC-PGA2
69.1
38
89
2.80C
478-pin FC-PGA2
75
40
69.7
2.80E
478-pin FC-PGA2
69.1
38
89
3
478-pin FC-PGA2
70
38
 81.9

3E

478-pin FC-PGA2

69.1

38

89

3.06
478-pin FC-PGA2
69
38
81.8
3.20
478-pin FC-PGA2
70
38
82.0

3.20E

478-pin FC-PGA2

73.2

38

103

3.40

478-pin FC-PGA2

70

38

82

3.40E

478-pin FC-PGA2

73.2

38

103

 


Appendix 2.  Rear case fan

Rear case fan is 120 mm Antec.  Note the grille behind the fan's blades.  Fan is attached to case using zip ties with silicon pads behind each corner of the fan to minimise any noise due to vibration.  The Zalman cooler is in the foreground.  Another view of the Zalman cooler is shown in Appendix 4.  

rear_fan.jpg


Appendix 3.  Grille for rear case fan

Fan's grille is on the right.  Original punched hole port for this fan has been nibbled out.  Not elegant, but is effective.  The smaller grille is on the Zalman power supply unit. 

fan_grille.jpg


Appendix 4.  Zalman cooler

The cooler consists of the heatsink (the thing with all the fins) with a integrated fan the centre.  The black object on the left is the rear case fan.   ASUS P4P800E-Deluxe motherboard..

cpu_cooler


Appendix 5.  Drive bay

The drive bay has space for 5 drives, only one of which is occupied by the 120 GB SATA drive.  The ANTEC case is extremely well designed.  Note the rubber grommets for mounting screws.  The drive cage is removable. 

The orange object is a 120 mm fan at the front of the case.  This is a slow-revving Nexus 'Real Quiet' fan which I am currently testing in place of my normal 120 mm Spire fan.  The fan is attached to case with zip ties and silicon pads under each corner. 

drive_bay.jpg