You are not logged in.
Pages: 1
one weird issue i have is hardware related. it isn't easy to explain. basically i have a dell laptop, c610, and the CPU gets a bit noisy sometimes, varies from a little squeal occasionally to a flat out drone. In windows i mainly get it when i am running MSN messenger even if it is minimized. In linux it picks up when ever X is busy. Thing was it was awful with the arch default 2.6 kernel, the noise started full volume as soon as the boot started and stayed till i halted the system. so i did a reinstall with the 2.4 kernel and the sound is no more. now, i know it isn't up to you guys to fix this but i was wondering if it may be possible to narrow the cause down - possibly to a module, which can then be removed from a 2.6 self-build. i think it may be acpi related but i have no clue where to start - any ideas or similar experience?
Offline
i have a dell laptop, c610,
that's your problem :-) sorry, to tell you, but dell is maybe the cheapest/performance, but not at all cheapest/quality for hardware (the only hardware i would avoid is dell and acer - especially the older laptops from both)
are the sounds high frequency, but hearable? then the noise you get is most probably a problem on your mainboard (only-hardware-problem) - my old compaq armada 1750 had very similar things, and i took it in pieces, connected it together while in pieces and found with measurements, that the sound system is too close to the cpu and over induction, currents made in the cpu / around cpu lead to "funny sounds" in the wire that goes to the amplifyer and you "hear" the cpu working
out of ferrit (no idea if it is english word, it's a metallic cristall) i made a plate to put inbetween and the noise was almoust eliminated, but as the display broke in 2 months, i had to buy a new laptop
if i were you, i would ask dell directly, as it is a construction problem
The impossible missions are the only ones which succeed.
Offline
my old compaq armada 1750 had very similar things, and i took it in pieces, connected it together while in pieces and found with measurements, that the sound system is too close to the cpu and over induction, currents made in the cpu / around cpu lead to "funny sounds" in the wire that goes to the amplifyer and you "hear" the cpu working
out of ferrit (no idea if it is english word, it's a metallic cristall) i made a plate to put inbetween and the noise was almoust eliminated, but as the display broke in 2 months, i had to buy a new laptop
I consider that to be a fairly impressive feat of computer engineering... did you learn that in your molecular biology classes!?
Dusty
Offline
dp wrote:my old compaq armada 1750 had very similar things, and i took it in pieces, connected it together while in pieces and found with measurements, that the sound system is too close to the cpu and over induction, currents made in the cpu / around cpu lead to "funny sounds" in the wire that goes to the amplifyer and you "hear" the cpu working
out of ferrit (no idea if it is english word, it's a metallic cristall) i made a plate to put inbetween and the noise was almoust eliminated, but as the display broke in 2 months, i had to buy a new laptop
I consider that to be a fairly impressive feat of computer engineering... did you learn that in your molecular biology classes!?
Dusty
induction and electriciy i learned from physics, using a screwdriver and taking things to pieces and put them again together i learned by myself ;-)
often people say to me, that giving me something you can take appart is not a good idea if you want it having in the same shape, but hey! everything i've taken appart worked afterwards (well, one radio is still in pieces, but it works again and looks very "cool")
unfortunately a lot of modern hardware is too much of plastic and all the electronics in one chip, so that you cannot manipulate too much by hand - my scanner has 2 ic, one chip one motor, 3 gears and a lot of plastic
The impossible missions are the only ones which succeed.
Offline
i forgot to mention it was a well known dell problem but they aren't forth coming with an explanation. it appears to be "cross talk" between the component as you suggest but that doesn't explain why it is worse with some apps and ONE kernel (i have booted 5 others and 2 other OS's) if you could help explain that then that helps i'm not implying this is anyones fault but Dell's but any help is appreciated
Offline
i forgot to mention it was a well known dell problem but they aren't forth coming with an explanation. it appears to be "cross talk" between the component as you suggest but that doesn't explain why it is worse with some apps and ONE kernel (i have booted 5 others and 2 other OS's) if you could help explain that then that helps i'm not implying this is anyones fault but Dell's but any help is appreciated
about msn i dont know, but i assume the ONE kernel, that produces it is using the hardware "more" intensively or simply different
you can experiment with the kernel by recompiling it with different options for the kernel and see what happens, but this takes time :-(
The impossible missions are the only ones which succeed.
Offline
i still want to figure this one out - i would like to be able to use the 2.6 kernel series without losing my sanity - i am happy to recompile my own but i was wondering if there was an easy way to compare 2.4 and 2.6 stock kernel configs?
i been using bash scripts a lot recently so i figure there most be some standard app the will show me the differences between the files?
Offline
i still want to figure this one out - i would like to be able to use the 2.6 kernel series without losing my sanity - i am happy to recompile my own but i was wondering if there was an easy way to compare 2.4 and 2.6 stock kernel configs?
i been using bash scripts a lot recently so i figure there most be some standard app the will show me the differences between the files?
i use this:
console: diff
X: kompare
wild guess:
try disabling acpi
try disabling stepping for cpu
The impossible missions are the only ones which succeed.
Offline
dp - don't :? me! that's not a wild guess - that's a goddamn well educated guess and you know it!
cheers!
i'm just comparing them in IE at the mo (at uni) and basically acpi is the saem in both kernels except somethings are modules in 2.4 which are built in in 2.6 - these are enabled in 2.4 without the noise
therefore it must be the stepping - there is only a few mentions of cpu period on the 2.4 config - but loads that looks related in the 2.6 config. how ill i figure out what is what? or should i just embark on compiling my own 2.6 kernel and figure it out when i get to the config stage?
this seems like the right section:
#
# CPU Frequency scaling
#
CONFIG_CPU_FREQ=y
# CONFIG_CPU_FREQ_PROC_INTF is not set
CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y
# CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE is not set
CONFIG_CPU_FREQ_GOV_PERFORMANCE=y
CONFIG_CPU_FREQ_GOV_POWERSAVE=m
CONFIG_CPU_FREQ_GOV_USERSPACE=m
CONFIG_CPU_FREQ_24_API=y
CONFIG_CPU_FREQ_TABLE=m
i have never compiled a kernel before but i am assuming that it's going to be easy to spot the CPU Frequency scaling disable option?
Offline
Been my experience that noises like this come from the fans. Since it seems to be processor related, it makes even more sense. My Dell speeds the fan up when it detects a higher processor usage, though at other times it dosen't seem to care. I can force it to happen by using acrobat reader though.
The newer Dell stuff is pretty good compared to others IMHO. But if you want quality hardware, better buy an i-book.
-Shawn
Offline
thanks for those thoughts - you may like to investigate the i8k utilities - they allow you to control the fans on dell notebooks
may be i have misunderstood but i know for certain that my issue is not related to the fans
Offline
I get the same noise from time to time. I believe that the hard drive is the culprit. When it spins up, the noise quits.
Shawn, I agree, If I could do it all over, it would be an apple. I regret buying a Dell. Bad bios, Strange noises, ... sigh.
Offline
I'm running arch on a D800 laptop now and before that, I ran arch on a C600. I never had that exact problem myself but I noticed increased HD or processor noise from my speakers on both laptops when I plugged it in to a wall outlett that had another power adaptor plugged in. The noise would go away if I ran on just battery. Other than that, arch has run excellently on both Dell's.
Offline
well, i recompiled my first kernel.! twice actually. i took the opportunity to compile in the gensplash framebuffer splash patches and they worked brilliantly! The splash screens just look really professional without any tweaking at all.
Only problem is that my noise prob was not solved - i disabled the CPU_FREQ stuff and made all the APM and ACPI the same as the 2.4 kernel but i still get the noise and it is really bad.
I was wondering if it might be the built-in ethernet and modem causing it? i am not sure if 2.4 detects them properly or if it has the right drivers or something? maybe 2.6 is doing that better - no idea. anymore suggestions?
to those of you who have made suggestions about hard drive spin-ups etc - i know what the problem is - i am just trying to solve it - but i appreciate yous suggestions
Offline
i'm more convinced than ever that my prob with 2.6 is related to the kernel detecting and powering up my built in winmodem - is this even a feasible suggestion? How can stop the kernel from initiating it? I would like to move to 2.6 - 2.4 is starting to get me down...
Offline
a-ha! is it the clock speed....! in 2.4 it is set to 100hz in 2.6 the default is 1000Hz - this can be changed without a kernel recompile, so as I have a 2.6 kernel at home am going to give it a try - w00t!
good old google - amazing how you can try a million times but if you aren;t using teh right terms you can fly right by it everytime! actually you do need a recompile - i lied
http://www.linuxquestions.org/questions … post963356
blimey ---> http://bugme.osdl.org/show_bug.cgi?id=2478
http://bugzilla.kernel.org/show_bug.cgi?id=3406
how did i miss all this?
i have now found two fixes:
clock down to 100Hz --> seems a good idea if it works
disable the C3 state before compilation - what is the C3 state? is this the same as disabling the acpi processor module?
will try and compile latest kernel with ck patchset now
Offline
all sort - i made this patch to go with the -ck patch set:
diff -Naur linux-2.6.9/include/asm-i386/param.h linux-2.6.9/include/asm-i386/param.h.new
--- linux-2.6.9/include/asm-i386/param.h 2004-10-19 05:53:24.000000000 +0800
+++ linux-2.6.9/include/asm-i386/param.h.new 2004-12-09 21:40:33.000000000 +0800
@@ -2,7 +2,7 @@
#define _ASMi386_PARAM_H
#ifdef __KERNEL__
-# define HZ 1000 /* Internal kernel timer frequency */
+# define HZ 100 /* Internal kernel timer frequency */
# define USER_HZ 100 /* .. some user interfaces are in "ticks" */
# define CLOCKS_PER_SEC (USER_HZ) /* like times() */
#endif
or you can download it from here
and the -nitro patchsets allow you to set the clock to 100, 200, 500 or 100 anyway - so that is a good bet - i have had success with 500 and 100
Offline
Pages: 1