Computer slows to 1% speed randomly and at startup

the_tool_man wrote:

On Apr 10, 1:03Â*pm, "123Jim" wrote:

Thanks I could not remember the name of the modes

Here's why it happens and how to change
it:http://www.microsoft.com/whdc/device...e/IDE-DMA.mspx

123Jim:

I investigated this, but everything is set to use DMA.

Regards,
John

Have you tried downloading a linux live cd and running the pc using that
just to rule out a hardware problem.

"the_tool_man" wrote in message
...
On Apr 10, 8:57 pm, Paul wrote:

Is the slowdown only icon related ?

WinXP can have a couple problems. I've heard of "icon cache corruption",
where a database file needs to be deleted, so Windows can rebuild it.
In addition, sometimes people need to bump up the size of the icon cache.
You might try a Google on "icon cache" and see what pops up.

Paul- Hide quoted text -

- Show quoted text -

No. I was just using the icon refresh as an example. It happens
every time at startup, and then randomly occurs during use of the
computer. I have found no pattern or trigger to make it happen.

I have tried placing a small fan blowing into the front of the case,
but this made no difference.

I downloaded Motherboard Monitor, but it doesn't support my
motherboard (or my motherboard doesn't have any temperature sensors).

Please keep the suggestions coming!

Regards,
John.


You could try this:
http://www.softpedia.com/get/Tweak/S.../BootVis.shtml
"BootVis is a performance trace visualization utility that you can use with
Windows XP systems."

It shows you what is being loaded during boot and how long it is taking ...
so you might be able to spot what it is that's causing the delay at boot
time.

For example, bootvis tells me my boot is 'done ' in 75 seconds, but it takes
164 seconds before my computer is ready for use and idle. Explorer.exe
kicks in at 68.41 seconds. Also my 'driver init' time is 37.19 seconds and
starts at 4 seconds in.

Update:

I downloaded CPU Stability Test 6.0, which let me do a CPU warming
test. With the task manager performance tab visible, I can start the
CPU test, and within 15 seconds, watch the scroll rate of the graph
essentially come to a stop. I have to click on the stop button for
the test and wait about 90 seconds before the computer responds and
the test stops. The computer cooling fan never speeds up above idle
the whole time. I have repeated this several times. I am convinced I
have a thermal management problem with the CPU. I visually checked
the motherboard for another header for the fan, but there isn't one.
I'm going to try hot-wiring the fan and see what happens.

Regards,
John.

On Sat, 11 Apr 2009 17:03:29 -0700 (PDT), the_tool_man
wrote:

Update:

I downloaded CPU Stability Test 6.0, which let me do a CPU warming
test. With the task manager performance tab visible, I can start the
CPU test, and within 15 seconds, watch the scroll rate of the graph
essentially come to a stop. I have to click on the stop button for
the test and wait about 90 seconds before the computer responds and
the test stops. The computer cooling fan never speeds up above idle
the whole time. I have repeated this several times. I am convinced I
have a thermal management problem with the CPU. I visually checked
the motherboard for another header for the fan, but there isn't one.
I'm going to try hot-wiring the fan and see what happens.

Regards,
John.


While it does seem possible there is a thermal problem,
usually a CPU will not heat up enough to crash or
downthrottle within 15 seconds if the only problem is the
fan not spinning up. In fact, a system can run for 15
seconds with the fan totally stopped in most systems and not
have risen much, perhaps 5 to 10C which on a properly cooled
system would tend to leave it still below the throttling
threshold.

If you have some heatsink grease, I'd pull the heatsink off,
clean off the original thermal interface material (may
require a petroleum solvent if it's the melted waxy type,
and the heatsink may come off much easier if you try to
remove it shortly after turning the system off so the
interface pad is still warm, OR reheat it with a hair dryer
or heat gun set _very_ low.

Examine both surfaces to be sure there are no
irregularities, then apply only a tiny, less than a grain of
rice sized blob on the center of the CPU then reinstall the
'sink.

CPU-Z will show you the clock speed of the CPU while you are
running the Stability Test or other stress tests.
http://www.cpuid.com/cpuz.php

HWMonitor "might" show you accurate temperatures and/or
voltages, or it might misread them, but if you compare the
temps before a stress test and during it, you should see CPU
temp rise significantly either way if it is being reported.
http://www.cpuid.com/hwmonitor.php

Update:

So much for jumping to conclusions. I was convinced it was a cooling
issue, but hotwiring the fan made no difference. Go here, if you're
interested:

http://groups.google.com/group/alt.c...de82627?hl=en#

In my search for info on the proprietary fan connections, I stumbled
across an online Dell service manual for my computer. It made
reference to a built-in diagnostic utility on a hidden partition of
the HD. I booted to that partition, ran the utility, and selected the
most comprehensive test. After 4 hours of testing, I learned two
things.

1: The fan speed controller works. There was a specific test for
speeding the fan up and seeing how fast it spins after a second or
two. It does not test whatever hardware measures the temperature, so
I guess thermal issues aren't entirely ruled out. However, see
learning number two...

2: During two of the lengthy HD tests, it found errors. During the
read test, it found Error code 0F00:0244 Unrecoverable data error or
media is write protected. During the verify test, it found Error code
0F00:1A44 with the same description. During both tests, it found
these errors on about 50 blocks (same ones both times). Also, the
sequence of sounds made by the HD when these errors occurred were
identical to those that happen when the computer slows. Conclusive?
I don't know.

ActiveSmart and other HD testing utilities I have still say the HD is
fine.

Regards,
John.

"the_tool_man" wrote in message
...
Update:

So much for jumping to conclusions. I was convinced it was a cooling
issue, but hotwiring the fan made no difference. Go here, if you're
interested:

http://groups.google.com/group/alt.c...de82627?hl=en#

In my search for info on the proprietary fan connections, I stumbled
across an online Dell service manual for my computer. It made
reference to a built-in diagnostic utility on a hidden partition of
the HD. I booted to that partition, ran the utility, and selected the
most comprehensive test. After 4 hours of testing, I learned two
things.

1: The fan speed controller works. There was a specific test for
speeding the fan up and seeing how fast it spins after a second or
two. It does not test whatever hardware measures the temperature, so
I guess thermal issues aren't entirely ruled out. However, see
learning number two...

2: During two of the lengthy HD tests, it found errors. During the
read test, it found Error code 0F00:0244 Unrecoverable data error or
media is write protected. During the verify test, it found Error code
0F00:1A44 with the same description. During both tests, it found
these errors on about 50 blocks (same ones both times). Also, the
sequence of sounds made by the HD when these errors occurred were
identical to those that happen when the computer slows. Conclusive?
I don't know.

ActiveSmart and other HD testing utilities I have still say the HD is
fine.

Regards,
John.

You used the CPU Stability test and reported that the behaviour occurred
when you used the cpu heat test. I don't think that test would trouble the
hard drive at all. I imagine a very small amount of data would be swapped
between the processor and memory, If that, possibly it swaps data between
processor and cache, although I can't be sure about that. That leads me to
think about the processor again.

Someone suggested that you check that the heat sink is correctly attached to
the processor, by repasting with thermal paste. (in this thread or the other
one). Did you try that?

I would have thought however that if the heat ink was not contacting the
processor correctly the computer would just turn off rather than start
processing in some reduced state as you have mentioned. However maybe your
processor does that, and it might be worth a try.

cheers

the_tool_man wrote:
Update:

So much for jumping to conclusions. I was convinced it was a cooling
issue, but hotwiring the fan made no difference. Go here, if you're
interested:

http://groups.google.com/group/alt.c...de82627?hl=en#

In my search for info on the proprietary fan connections, I stumbled
across an online Dell service manual for my computer. It made
reference to a built-in diagnostic utility on a hidden partition of
the HD. I booted to that partition, ran the utility, and selected the
most comprehensive test. After 4 hours of testing, I learned two
things.

1: The fan speed controller works. There was a specific test for
speeding the fan up and seeing how fast it spins after a second or
two. It does not test whatever hardware measures the temperature, so
I guess thermal issues aren't entirely ruled out. However, see
learning number two...

2: During two of the lengthy HD tests, it found errors. During the
read test, it found Error code 0F00:0244 Unrecoverable data error or
media is write protected. During the verify test, it found Error code
0F00:1A44 with the same description. During both tests, it found
these errors on about 50 blocks (same ones both times). Also, the
sequence of sounds made by the HD when these errors occurred were
identical to those that happen when the computer slows. Conclusive?
I don't know.

ActiveSmart and other HD testing utilities I have still say the HD is
fine.

Regards,
John.

So the conclusions are

1) Replace hard drive :-)

2) SMART test isn't as good as we thought.

Your priority now, is to save your data.

Paul

On Apr 22, 7:09*am, "123Jim" wrote:

Someone suggested that you check that the heat sink is correctly attached to
the processor, by repasting with thermal paste. (in this thread or the other
one). Did you try that?

I haven't replaced the thermal paste. Maybe I it's been the HD all
along, and when I pulled the heat sink out I created a thermal issue.
Lack of contact with the heat sink might explain why hotwiring the fan
made no difference. I'll get some thermal paste and re-do the heat
sink again, just to make sure.

Regards,
John.

On Apr 22, 7:19*am, Paul wrote:

So the conclusions are

1) Replace hard drive :-)

2) SMART test isn't as good as we thought.

Your priority now, is to save your data.

* * Paul

Thankfully, I did a full backup three weeks ago (just in case
Conficker showed up).

Regards,
John.

Somewhere on teh intarwebs the_tool_man wrote:
On Apr 22, 7:19 am, Paul wrote:

So the conclusions are

1) Replace hard drive :-)

2) SMART test isn't as good as we thought.

Your priority now, is to save your data.

Paul

Thankfully, I did a full backup three weeks ago (just in case
Conficker showed up).

Have you tried Hard Disk Sentinel? There's a trial version I believe. I
swear by it and would like to know what it says about your disk.

Cheers,
--
Shaun.

"Build a man a fire, and he`ll be warm for a day. Set a man on fire, and
he`ll be warm for the rest of his life." Terry Pratchett, Jingo.