CPU 100%

Questions and Answers : Windows : CPU 100%
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Profile Seeriktus
Avatar

Send message
Joined: 4 Sep 04
Posts: 24
Credit: 211,921
RAC: 0
United Kingdom
Message 233744 - Posted: 19 Jan 2006, 21:16:23 UTC

I was just wondering if running my CPU on 100% all day(13-14 hours) will be causing it any damage.
ID: 233744 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 233785 - Posted: 19 Jan 2006, 21:38:11 UTC

No, your CPU is build to be able to run at 100%. All you need is adequate cooling.
My CPU runs the various projects at 100% 24/7, I just make sure the cooler gets dusted out once in a while. (about 3 times a year, with the cat around).
ID: 233785 · Report as offensive
Profile Pooh Bear 27
Volunteer tester
Avatar

Send message
Joined: 14 Jul 03
Posts: 3224
Credit: 4,603,826
RAC: 0
United States
Message 233815 - Posted: 19 Jan 2006, 21:55:26 UTC

I have 1 machine that has been nearly 100% 24/7 for nearly 2.5 years, no problems, yet.

Processors are tuned to run full all the time. We just never use that processing power, hence SETI and other projects of this sort.



My movie https://vimeo.com/manage/videos/502242
ID: 233815 · Report as offensive
Gareth Lock

Send message
Joined: 14 Aug 02
Posts: 358
Credit: 969,807
RAC: 0
United Kingdom
Message 233924 - Posted: 19 Jan 2006, 23:14:53 UTC
Last modified: 19 Jan 2006, 23:17:08 UTC

Yeah... 100% CPU usage shouldn't be a problem... If you're new to BOINC then check the CPU temperature for the first week or so. I've got one machine running 24/7 participating in several projects with no problems and another one that, although doesn't run 24/7 does do at least two-three hours or more per day, again without problems.

If you need any pointers regarding CPU/machine cooling then drop me an e-mail at "garethlock@yahoo.com" and I can return an attachment containing a text file the contents of which I beefed up using several active threads from different projects when the topic came up due to the summer months last year. The only thing I would suggest is that you don't run BOINC or any DC project on a laptop.

If you've been a "Classic" user but are new to the BOINC system, then you should have no problems since "Classic" held the CPU at 100% for extended periods.

Hope this helps...


ID: 233924 · Report as offensive
Profile Scary Capitalist
Avatar

Send message
Joined: 21 May 01
Posts: 7404
Credit: 97,085
RAC: 0
United States
Message 234089 - Posted: 20 Jan 2006, 5:36:17 UTC - in response to Message 233924.  

Yeah... 100% CPU usage shouldn't be a problem... If you're new to BOINC then check the CPU temperature for the first week or so. I've got one machine running 24/7 participating in several projects with no problems and another one that, although doesn't run 24/7 does do at least two-three hours or more per day, again without problems.

If you need any pointers regarding CPU/machine cooling then drop me an e-mail at "garethlock@yahoo.com" and I can return an attachment containing a text file the contents of which I beefed up using several active threads from different projects when the topic came up due to the summer months last year. The only thing I would suggest is that you don't run BOINC or any DC project on a laptop.

If you've been a "Classic" user but are new to the BOINC system, then you should have no problems since "Classic" held the CPU at 100% for extended periods.

Hope this helps...



Why would you suggest not running it on a laptop? Plenty of people here reference running Seti on their laptops and this is the first post I recall reading where it was discouraged.

Founder of BOINC team Objectivists. Oh the humanity! Rational people crunching data!
I did NOT authorize this belly writing!

ID: 234089 · Report as offensive
Profile RRipley
Volunteer tester

Send message
Joined: 18 Apr 01
Posts: 880
Credit: 23,005,843
RAC: 0
Germany
Message 234110 - Posted: 20 Jan 2006, 9:16:45 UTC - in response to Message 234089.  

Why would you suggest not running it on a laptop? Plenty of people here reference running Seti on their laptops and this is the first post I recall reading where it was discouraged.


Well, I second Gareth's suggestion since many many laptops here at work fail due to heating problems in normal working environments. When a laptop now works a long time with 100% CPU, the cooling system might not be able to work properly.
ID: 234110 · Report as offensive
Developer

Send message
Joined: 18 Jan 06
Posts: 8
Credit: 2,660
RAC: 0
Germany
Message 234141 - Posted: 20 Jan 2006, 11:57:15 UTC

Actually the CPU runs all the time at 100%. It can't really stand still. The only question is what it actually does in this time.

If you have no programs running the CPU temds to some operating system tasks and spends most of the time waiting for something to happen. But the CPU is running all the time.

The CPU usage display is somewhat misleading because it only shows how much of the CPU's 'attention' goes to running tasks. There also is an operating system task wich does nothing noteworthy except constantly running in an endless loop. This task only is tended to by the CPU when there is nothing else to do because the CPU can't simply stop. If this task were included in the display, it would constantly show 100% and not be of much use. If you open the task manager you will quickly find it. It's the one which got about 98% of the CPU time.

If you run a program which constantly does something the CPU will not service this idle task. The bottom line is that you simply donate the CPU time which normally would have gone to waste.

Unlike mechanical devices a CPU does not get worn out by use. Theoretically it can go on working forever. Unfortunately the transistors produce heat while switching and running a task involves a lot more switching than idling in a loop. Therefore the CPU gets hotter and without adequate cooling it may take severe damage. It seems reasonable that notebooks are somewhat proner to overheating due to their compact construction. The coolers are smaller and the airflow may also be less then optimal.

Perhaps a look into your computer's BIOS may put you on the safe side. Some mainboards have settings at what temperature an alert will be sounded and some also have settings for an emergency shutdown. If you have such settings, then set them to a low temperature, but of course one which still is above the normal operating temperature.
ID: 234141 · Report as offensive
Gareth Lock

Send message
Joined: 14 Aug 02
Posts: 358
Credit: 969,807
RAC: 0
United Kingdom
Message 234144 - Posted: 20 Jan 2006, 12:07:03 UTC - in response to Message 234141.  
Last modified: 20 Jan 2006, 12:11:50 UTC

Actually the CPU runs all the time at 100%. It can't really stand still. The only question is what it actually does in this time.

If you have no programs running the CPU temds to some operating system tasks and spends most of the time waiting for something to happen. But the CPU is running all the time.

The CPU usage display is somewhat misleading because it only shows how much of the CPU's 'attention' goes to running tasks. There also is an operating system task wich does nothing noteworthy except constantly running in an endless loop. This task only is tended to by the CPU when there is nothing else to do because the CPU can't simply stop. If this task were included in the display, it would constantly show 100% and not be of much use. If you open the task manager you will quickly find it. It's the one which got about 98% of the CPU time.

If you run a program which constantly does something the CPU will not service this idle task. The bottom line is that you simply donate the CPU time which normally would have gone to waste.

Unlike mechanical devices a CPU does not get worn out by use. Theoretically it can go on working forever. Unfortunately the transistors produce heat while switching and running a task involves a lot more switching than idling in a loop. Therefore the CPU gets hotter and without adequate cooling it may take severe damage. It seems reasonable that notebooks are somewhat proner to overheating due to their compact construction. The coolers are smaller and the airflow may also be less then optimal.

Perhaps a look into your computer's BIOS may put you on the safe side. Some mainboards have settings at what temperature an alert will be sounded and some also have settings for an emergency shutdown. If you have such settings, then set them to a low temperature, but of course one which still is above the normal operating temperature.



An interesting insight, one which, if I had more time the other night, I might have investigated into further in my initial post... The name of this loop as quoted from Win XP's Task Manager is "System Idle Process" just out of interest, taking about 95-97% of my laptop's CPU cycles as I type this. My two towers do the actual crunching.


ID: 234144 · Report as offensive
Profile Beaver - 417 RCAF

Send message
Joined: 12 Jun 99
Posts: 2
Credit: 875,741
RAC: 0
Canada
Message 234526 - Posted: 20 Jan 2006, 23:22:51 UTC - in response to Message 233785.  

I have had computers running 24/7 since June of 1999 doing Seti, the only problem I ever had was my computer getting too out of date to run the more advanced versions of Seti, the computer I bought in 1995 just doesn't cut the mustard now.
ID: 234526 · Report as offensive
jenesuispasbavard
Volunteer tester
Avatar

Send message
Joined: 13 Sep 05
Posts: 49
Credit: 12,385,974
RAC: 0
United States
Message 234780 - Posted: 21 Jan 2006, 7:25:11 UTC

My Dell Inspiron 600m has been running SETI@Home since September 2005 all day, except in-flight. Nothing's happened yet. And I live in Austin, Texas; it can get pretty hot here.
ID: 234780 · Report as offensive
Developer

Send message
Joined: 18 Jan 06
Posts: 8
Credit: 2,660
RAC: 0
Germany
Message 234795 - Posted: 21 Jan 2006, 8:51:00 UTC - in response to Message 234144.  


An interesting insight, one which, if I had more time the other night, I might have investigated into further in my initial post... The name of this loop as quoted from Win XP's Task Manager is "System Idle Process" just out of interest, taking about 95-97% of my laptop's CPU cycles as I type this. My two towers do the actual crunching.



Yes, that's exactly the one. I did not use its name because on my system it is called 'Leerlaufprozess' which is German for exactly the same. When you boot your system it will start a few tasks but they all do not have a high priority and thus only get a small share of the CPU time now and then. If you look into the task manager at this time, the idle process will indeed get most of the CPU time.

The user starts tasks usually with 'normal' priority. This means the low priority tasks get their share now and then as before and the idle process gets nothing unless the user tasks are also idling.

Last, there are ways to set a process up to highest or 'realtime' priority. In this case all other tasks get the bare minimum of the cpu time and this high priority task hamsters all the rest. This is only rarely done. Of course it squeezes out a maximum of performance for this one task, allowing for example the fastest possible ( = real time) response to external events. But the prize is an almost unresponding system. Even moving the mouse cursor to click a button becomes an execise in patience.

I hope this little excursion in the operating system's management of multiple tasks has been interesting.



ID: 234795 · Report as offensive
Developer

Send message
Joined: 18 Jan 06
Posts: 8
Credit: 2,660
RAC: 0
Germany
Message 234798 - Posted: 21 Jan 2006, 9:21:28 UTC - in response to Message 234780.  

My Dell Inspiron 600m has been running SETI@Home since September 2005 all day, except in-flight. Nothing's happened yet. And I live in Austin, Texas; it can get pretty hot here.


Years ago I lived in El Paso. Thank god the buildings have air conditions. But hot air can cause problems, especially if the computer's cooler is too small or the fan to weak. Also if somebody overclocks his system there may be problems when the air temperature climbs.

Another huge problem may be simple dust. The fan sucks it into the computer and it accumulates in the cooler. When the ribs of the cooler are filled with dust the air cannot flow through it properly and the cooling becomes less efficient. Once in a while I take apart my computers and clean out the coolers since my old Pentium II machine once almost fried on a hot summer day.

ID: 234798 · Report as offensive
Gareth Lock

Send message
Joined: 14 Aug 02
Posts: 358
Credit: 969,807
RAC: 0
United Kingdom
Message 235611 - Posted: 22 Jan 2006, 9:26:26 UTC - in response to Message 234798.  
Last modified: 22 Jan 2006, 9:27:58 UTC

My Dell Inspiron 600m has been running SETI@Home since September 2005 all day, except in-flight. Nothing's happened yet. And I live in Austin, Texas; it can get pretty hot here.


Years ago I lived in El Paso. Thank god the buildings have air conditions. But hot air can cause problems, especially if the computer's cooler is too small or the fan to weak. Also if somebody overclocks his system there may be problems when the air temperature climbs.

Another huge problem may be simple dust. The fan sucks it into the computer and it accumulates in the cooler. When the ribs of the cooler are filled with dust the air cannot flow through it properly and the cooling becomes less efficient. Once in a while I take apart my computers and clean out the coolers since my old Pentium II machine once almost fried on a hot summer day.



Yeah... I'd go with that... Just to give you an idea as to what dust on the cooler can do... I have an Athlon64 FX53 usually used for gaming, but has a place on my BOINC profile 'cause it gets to do a little crunching when I'm on the net sometimes. Anyhow... Back to the point... Did a clean-out, taking the fan module off the heatsink to gain access to those fins and got a mini PC cleaning vac to them, filled it's small bag somewhat and screwed the fan back on... Anyway.. After powering on an hitting BOINC for two hours the machine was running 3-4 degrees cooler than before.
ID: 235611 · Report as offensive
Profile Seeriktus
Avatar

Send message
Joined: 4 Sep 04
Posts: 24
Credit: 211,921
RAC: 0
United Kingdom
Message 235619 - Posted: 22 Jan 2006, 10:24:31 UTC - in response to Message 233785.  

No, your CPU is build to be able to run at 100%. All you need is adequate cooling.
My CPU runs the various projects at 100% 24/7, I just make sure the cooler gets dusted out once in a while. (about 3 times a year, with the cat around).


How can I dust out my fan? particularly with laptops, the desktops seem to not need cleaning because they have larger fans
ID: 235619 · Report as offensive
Developer

Send message
Joined: 18 Jan 06
Posts: 8
Credit: 2,660
RAC: 0
Germany
Message 235822 - Posted: 22 Jan 2006, 18:15:22 UTC - in response to Message 235619.  

No, your CPU is build to be able to run at 100%. All you need is adequate cooling.
My CPU runs the various projects at 100% 24/7, I just make sure the cooler gets dusted out once in a while. (about 3 times a year, with the cat around).


How can I dust out my fan? particularly with laptops, the desktops seem to not need cleaning because they have larger fans


This is rather easy to do on a regular PC. You must simply unlatch the entire cooler from the CPU and unplug its power cable. Then you usually can easily remove the heat sink from the fan. The heat sink is just a ribbed piece of metal, so it's easy to clean. Just take care not to remove the paste on the underside which is used to ensure a good thermal connection to the CPU. Before reassembling everything it might be a good idea to apply some fresh paste (very little is sufficient because the heat sink will be pressed onto the CPU by the latches). You can get the paste at any store which sells electronic parts.

Disassembling a notebook is quite another story. Their construction is a lot more compact than a regular PC and it may be nessecary to remove quite a few of the notebook's components before you can actually get to the CPU. On top of that not all parts of a notebook seem to be as standardized as the components of a regular PC. Some parts like hard drives or CD ROMs are interchangable and others appear to be the manufacturer's own design. I would therefore advise against trying to disassemble the notebook without really sufficient need unless you feel up to the task.
ID: 235822 · Report as offensive
Gareth Lock

Send message
Joined: 14 Aug 02
Posts: 358
Credit: 969,807
RAC: 0
United Kingdom
Message 235934 - Posted: 22 Jan 2006, 21:07:59 UTC - in response to Message 235822.  

No, your CPU is build to be able to run at 100%. All you need is adequate cooling.
My CPU runs the various projects at 100% 24/7, I just make sure the cooler gets dusted out once in a while. (about 3 times a year, with the cat around).


How can I dust out my fan? particularly with laptops, the desktops seem to not need cleaning because they have larger fans


This is rather easy to do on a regular PC. You must simply unlatch the entire cooler from the CPU and unplug its power cable. Then you usually can easily remove the heat sink from the fan. The heat sink is just a ribbed piece of metal, so it's easy to clean. Just take care not to remove the paste on the underside which is used to ensure a good thermal connection to the CPU. Before reassembling everything it might be a good idea to apply some fresh paste (very little is sufficient because the heat sink will be pressed onto the CPU by the latches). You can get the paste at any store which sells electronic parts.

Disassembling a notebook is quite another story. Their construction is a lot more compact than a regular PC and it may be nessecary to remove quite a few of the notebook's components before you can actually get to the CPU. On top of that not all parts of a notebook seem to be as standardized as the components of a regular PC. Some parts like hard drives or CD ROMs are interchangable and others appear to be the manufacturer's own design. I would therefore advise against trying to disassemble the notebook without really sufficient need unless you feel up to the task.


Like I said in my post about the Athlon64 FX53... There is no need to remove the heatsink from the CPU, it just makes more work!! Most fans are held to heatsinks by four screws. Just shutdown the box as normal and switch off at the hard switch at the back of the PSU, failing that, switch off at the wall. DO NOT disconnect the power lead from the back of the machine or from the wall, as the earth connection serves to ground any static that might be present on you.

BEFORE YOU START WORKING AND REGULARLY DURING WORKING ON ANY MACHINE TOUCH ANY METAL PART OF THE CHASSIS, THIS MAKES SURE THAT ANY STATIC YOU ARE CARRYING ON YOU IS GROUNDED AND THAT YOU ARE AT THE SAME POTENTIAL AS THE ELECTRONICS YOU ARE WORKING ON TO PREVENT EXPENSIVE DAMAGE TO THE INTEGRATED CIRCUITS ON THE MOTHERBOARD AND EXPANSION CARDS.

You'll need a small phillips head or posidrive screwdriver (usually) to remove the four screws holding the fan and the heatsink together, remove the fan from the heatsink and remove the power lead connected to the fan's motor from the board or PSU or whatever it's connected to. (Mine takes it's power from a molex feedthrough connected to the PSU, whilst sending a sense signal back to the board.) Then all you need to do is take a mini dust vacuum (got mine for about 6 quid from the local Maplins store) and get rid of all those dust bunnies trapped between the fan and the heatsink. Might be an idea to get the vac to the underside of the fan too. Then screw the fan back to the heatsink and reconnect the cables. This method prevents the need for removing the heatsink from the CPU and having to clean paste off both the CPU and the base of the heatsink before applying new paste or pad or whatever.

Whilst I'm on the topic here, just adding extra is no good whatsoever!! The heat generated by the processor during normal operation changes the chemical composition of the thermal interface, thus bonding the heatsink to the CPU. This is a ONE TIME process!! Thus the thermal interface, be it paste or pad needs to be COMPLETELY removed from both CPU and heatsink and fresh applied, each and EVERY time you remove the heatsink from the CPU.

Thus my method has a lot less hasstle involved, don't you think??

P.S. I hope this last paragraph serves to correct you Developer, whomever you are.



ID: 235934 · Report as offensive
OzzFan Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Apr 02
Posts: 15691
Credit: 84,761,841
RAC: 28
United States
Message 236001 - Posted: 22 Jan 2006, 23:47:33 UTC

DO NOT disconnect the power lead from the back of the machine or from the wall, as the earth connection serves to ground any static that might be present on you.


I would beg to differ here. With many of today's motherboards still carrying a charge as well as their capabilities to stay on with a very small draw of current even though they are "off" (so as to allow a keyboard with a Power button to turn them on, or other similiar uses), leaving the system plugged in can actually cause more harm than was once thought.

The best method for grounding oneself is to purchase an anti-static wrist strap or simply touching one's feet to the ground (bare ground, nothing that would generate static electricity such as carpeting; bare feet, no socks, shoes or slippers to prevent proper grounding).

I've seen recent motherboards destroyed because of that false logic. I've always stuck to my method for years and have never blown a component.
ID: 236001 · Report as offensive
Gareth Lock

Send message
Joined: 14 Aug 02
Posts: 358
Credit: 969,807
RAC: 0
United Kingdom
Message 236070 - Posted: 23 Jan 2006, 1:50:16 UTC - in response to Message 236001.  
Last modified: 23 Jan 2006, 2:03:13 UTC

DO NOT disconnect the power lead from the back of the machine or from the wall, as the earth connection serves to ground any static that might be present on you.


I would beg to differ here. With many of today's motherboards still carrying a charge as well as their capabilities to stay on with a very small draw of current even though they are "off" (so as to allow a keyboard with a Power button to turn them on, or other similiar uses), leaving the system plugged in can actually cause more harm than was once thought.

The best method for grounding oneself is to purchase an anti-static wrist strap or simply touching one's feet to the ground (bare ground, nothing that would generate static electricity such as carpeting; bare feet, no socks, shoes or slippers to prevent proper grounding).

I've seen recent motherboards destroyed because of that false logic. I've always stuck to my method for years and have never blown a component.



Thanks for the correction, maybe what I should have made plainer was that the user might have to wait 10 mins or so for the caps in the PSU to completely discharge AFTER they switched the hard switch (usually a rocker type situated by the mains supply lead input on the back of the PSU) off, failing that, the switched wall outlet, thus isolating ALL power to the system internals whilst leaving the earth ground connected through the mains cable to whatever earthing hookup is configured at the distro board at the entry to the property.

Of course, if you want to make doubly sure that the mobo isn't getting any juice, you could first disconnect the ATX connector(s) from the mainboard before starting work! Then maybe you'd like to explore the possibility of removing the BIOS battery to completely isolate the board from any possible source of power!!

Other than that, I stick to my guns. I've gutted and put together many a PC at my local referbishing club this way without problems, I've also built my own boxes since the days of the 486DX2 without any problems in this fashion. You could go out and buy an anti-stat strap for a couple of quid. Indeed, I do have one which I use on the more lengthy tasks, but I still ground it by attaching the crocodile clip to a grounding dag on the chassis somewhere. Seems to be the most easily accessable grounding point.

Anyhow... The essence of my original post remains the same... Removing the heatsink assembly from the CPU opens up a whole new can o' worms that can be simply avoided by removing the fan from the top and working on the heatsink in situ as long as the machine is powered down.



ID: 236070 · Report as offensive
Developer

Send message
Joined: 18 Jan 06
Posts: 8
Credit: 2,660
RAC: 0
Germany
Message 236094 - Posted: 23 Jan 2006, 2:33:18 UTC - in response to Message 236070.  

DO NOT disconnect the power lead from the back of the machine or from the wall, as the earth connection serves to ground any static that might be present on you.


I would beg to differ here. With many of today's motherboards still carrying a charge as well as their capabilities to stay on with a very small draw of current even though they are "off" (so as to allow a keyboard with a Power button to turn them on, or other similiar uses), leaving the system plugged in can actually cause more harm than was once thought.

The best method for grounding oneself is to purchase an anti-static wrist strap or simply touching one's feet to the ground (bare ground, nothing that would generate static electricity such as carpeting; bare feet, no socks, shoes or slippers to prevent proper grounding).

I've seen recent motherboards destroyed because of that false logic. I've always stuck to my method for years and have never blown a component.



Thanks for the correction, maybe what I should have made plainer was that the user might have to wait 10 mins or so for the caps in the PSU to completely discharge AFTER they switched the hard switch (usually a rocker type situated by the mains supply lead input on the back of the PSU) off, failing that, the switched wall outlet, thus isolating ALL power to the system internals whilst leaving the earth ground connected through the mains cable to whatever earthing hookup is configured at the distro board at the entry to the property.

Of course, if you want to make doubly sure that the mobo isn't getting any juice, you could first disconnect the ATX connector(s) from the mainboard before starting work! Then maybe you'd like to explore the possibility of removing the BIOS battery to completely isolate the board from any possible source of power!!

Other than that, I stick to my guns. I've gutted and put together many a PC at my local referbishing club this way without problems, I've also built my own boxes since the days of the 486DX2 without any problems in this fashion. You could go out and buy an anti-stat strap for a couple of quid. Indeed, I do have one which I use on the more lengthy tasks, but I still ground it by attaching the crocodile clip to a grounding dag on the chassis somewhere. Seems to be the most easily accessable grounding point.

Anyhow... The essence of my original post remains the same... Removing the heatsink assembly from the CPU opens up a whole new can o' worms that can be simply avoided by removing the fan from the top and working on the heatsink in situ as long as the machine is powered down.




No arguements against precautions against static charges, even if this has been more severe a problem in the old days. You only had too look at some of the old CMOS chips to zap them.

And I also agree that it's certainly more work to get out the heat sink. But, without trying to make a science out of it, your dust certainly is not my dust. A little blowing or vacuuming just will not work.This would only remove the top layer, but below that there usually sits another layer which has been baked together by the heat and this may require some wiping and scratching.

An even more extreme example is a friend, who smokes one cigarrette after another. His power supply and also his cooler are covered with a sticky brown coat of dust and nicotine. My only idea was to take out his heat sink and scrub it with soap and water. Instead he did just what you say and removed only the fan. The sticky dust could not be blown out, so he started to scratch it out with a screwdriver - and wound up buying a new mainboard because the mechanical force of his cleaning had split the CPU socket into two parts and did not work anymore.

So let's just agree that your way is the best if there really is just loose dust in the heat sink and mine for the more severe cases
ID: 236094 · Report as offensive
Gareth Lock

Send message
Joined: 14 Aug 02
Posts: 358
Credit: 969,807
RAC: 0
United Kingdom
Message 236372 - Posted: 23 Jan 2006, 18:00:12 UTC - in response to Message 236094.  
Last modified: 23 Jan 2006, 18:03:55 UTC

DO NOT disconnect the power lead from the back of the machine or from the wall, as the earth connection serves to ground any static that might be present on you.


I would beg to differ here. With many of today's motherboards still carrying a charge as well as their capabilities to stay on with a very small draw of current even though they are "off" (so as to allow a keyboard with a Power button to turn them on, or other similiar uses), leaving the system plugged in can actually cause more harm than was once thought.

The best method for grounding oneself is to purchase an anti-static wrist strap or simply touching one's feet to the ground (bare ground, nothing that would generate static electricity such as carpeting; bare feet, no socks, shoes or slippers to prevent proper grounding).

I've seen recent motherboards destroyed because of that false logic. I've always stuck to my method for years and have never blown a component.



Thanks for the correction, maybe what I should have made plainer was that the user might have to wait 10 mins or so for the caps in the PSU to completely discharge AFTER they switched the hard switch (usually a rocker type situated by the mains supply lead input on the back of the PSU) off, failing that, the switched wall outlet, thus isolating ALL power to the system internals whilst leaving the earth ground connected through the mains cable to whatever earthing hookup is configured at the distro board at the entry to the property.

Of course, if you want to make doubly sure that the mobo isn't getting any juice, you could first disconnect the ATX connector(s) from the mainboard before starting work! Then maybe you'd like to explore the possibility of removing the BIOS battery to completely isolate the board from any possible source of power!!

Other than that, I stick to my guns. I've gutted and put together many a PC at my local referbishing club this way without problems, I've also built my own boxes since the days of the 486DX2 without any problems in this fashion. You could go out and buy an anti-stat strap for a couple of quid. Indeed, I do have one which I use on the more lengthy tasks, but I still ground it by attaching the crocodile clip to a grounding dag on the chassis somewhere. Seems to be the most easily accessable grounding point.

Anyhow... The essence of my original post remains the same... Removing the heatsink assembly from the CPU opens up a whole new can o' worms that can be simply avoided by removing the fan from the top and working on the heatsink in situ as long as the machine is powered down.




No arguements against precautions against static charges, even if this has been more severe a problem in the old days. You only had too look at some of the old CMOS chips to zap them.

And I also agree that it's certainly more work to get out the heat sink. But, without trying to make a science out of it, your dust certainly is not my dust. A little blowing or vacuuming just will not work.This would only remove the top layer, but below that there usually sits another layer which has been baked together by the heat and this may require some wiping and scratching.

An even more extreme example is a friend, who smokes one cigarrette after another. His power supply and also his cooler are covered with a sticky brown coat of dust and nicotine. My only idea was to take out his heat sink and scrub it with soap and water. Instead he did just what you say and removed only the fan. The sticky dust could not be blown out, so he started to scratch it out with a screwdriver - and wound up buying a new mainboard because the mechanical force of his cleaning had split the CPU socket into two parts and did not work anymore.

So let's just agree that your way is the best if there really is just loose dust in the heat sink and mine for the more severe cases



Yeah... What I didn't say however, is that I have one of those high-pressure air-duster cans to loosten the stubborn stuff, then just use the brush attachment on the vac to kick it to touch. Though... I would say that if you can do it without removing the heatsink, then do so. However, if the crap is so baked or gummed on that my method doesn't work, then by all means, remove the heatsink... Just remember to COMPLETELY remove ALL traces of thermal interface material and apply fresh before re-applying the heatsink to the top of the processor.

My point was this... THERMAL INTERFACE, BE IT PASTE OR PAD, CANNOT BE RE-USED AFTER THIS CHEMICAL CHANGE HAS TAKEN PLACE.


ID: 236372 · Report as offensive
1 · 2 · Next

Questions and Answers : Windows : CPU 100%


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.