Buzz/high pitch sound in headphone output

What does that mean?

 

Wow, the RMClock HLT-setting really removed the noise which annoyed me the most. Great find!

But my CPU speed is now constantly @ 100%. Why?

 

Sounds like you either didn't setup the performance on demand, or you dont have it set to use performance on demand.

If you left click on the round gear looking icon you can set the profile, if you dont performance on demand configured though, you will need to open the program and check a few boxes under profiles (P-States and Throttling, with all the speeds you want)

 

I'm not sure...I posted the only documented part I could find... lol...

Probably going to have to post on their forum for the answer..

 

Pyro, isn't there a MS Windows HotFix to help this situation? I believe the procedure goes something like this:

1) Call (yes you read it correctly) MS and request hotfix for issue 896256.
or
1) Go to AMDZone.com and search for hotfix (yes this is for both Intel and AMD).

2) Got into regedit click HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager, go to new and select Key.

3) Type Throttle for the new key name.

4) Right click Throttle click New and select DWORD value.

5) Type PerfEnablePackageIdle for the value name.

6) Right click PerfEnablePackageIdle and then choose modify.

7) In the Edit DWORD value box, type 1. In the Value data box, make sure that Hexadecimal is selected and then click OK.

8) Quite the registry editor and reboot.

I must give credit to Maximum PC 2006 page 40 "Dual Core Survival Guide" by Gordon Mah Ung.

I HAVE NEVER TRIED THIS SO TRY IT AT YOUR OWN RISK. Also backup your registry before doing anything.

This may ultimately be a better solution or it may do nothing I am not sure you are the expert Pyro on this matter not me

 

No this is different.. Using a program like RMclock overrides windows settings. What you are mentioning is when the OS is having a problem with the scaling.. What this person is having a problem with is by default RMclock doesn't set your profiles, and it put the person in unmanaged speed by default. They need to go in and configure the speed/power profiles for this to work.

 

I was referring to not using the RMClock application at all. If low power states are the problem than maybe the procedure I mentioned would work similar to using the RMClock program and the HLT setting. Again you know far better than I so I will shut up Also does this issue happen with the T7200 processor I was unable to hear the noise in my Compal. Maybe I am deaf as well as blind.

 

The tweak you mention is used to activate the default throttling tables that are sent to the O/S from the CPU itself. RMclock lets you set your own power levels and speeds, as well as gives you other options built into the processor like HLT states, control over the heat sensors keeping your machine from toasting or locking up, and a few more things.. More simply, the patch you mentioned gives the ability for windows to know the speed/voltage tables exist, and tells windows to use them.. RM clock lets you hand manipulate them (requires the patch you mentioned), plus more features windows doesn't enable for simplicity stake...

It's all about reducing points of failure.. MS got such a bad reputation years ago from play and pray, they are very conservative now when it comes to extended feature sets... They look at the computer as making it simple to start and let you grow by adding stuff to it.. (though somewhat limited exposure in some areas because of closed sources)

*nix and BSD's use most of the features by default if you give it the right kernel for your processor family, other features can be added to a kernel, just like in windows. Linux is more about exposing all possibilities so the user has more control without the extra fluff, while sacrificing simplicity and thus driving some people away.

 

I see well that makes sense, learning more and more every minute.

Thanks

 

It's the nature of the thing. HLT is (someone correct me if I'm wrong) a CPU instruction that tells the processor to sleep, more or less. What RMClock does is send lots of those instructions to the processor when it's not busy doing other things. Windows sees that as activity. RMClock should calculate it out of its figures though.

 

Yes, HLT disables sectors of the registers in the CPU that aren't in use. CPUIDLE, CoolOn, and a few other programs were designed a few processor generations ago to manipulate the HLT because of heat issues. This HLT command isn't being hammered into the processor though, it is sent with regular spaced out intervals to avoid the HLT command waking itself up. When activited, it does have a "wake on activity" that enables the processor's registers again if needed between polls.

 

Take some initiative Pyro? what the hell do you think i am doing by asking the question on this thread?

 

Nice attitude there slick..

What I said was take initiative to call them back, since you stated that you had contacted their tech support on the issue and hadn't heard back. There tech support isn't this thread... Of course, if that is the same attitude you shared with them, I'm not suprised they aren't trying to grab you the quick fix..

As for the attitude directly at me, you might as well save it for someone who cares, Everyone here is volunteering time, knowledge, and experience, and you are getting exactly what you invest. So keep up that attitude... I'm sure that will get you the replies fit for you...

 

Hello all !

I'm new on this forum, and I'm french (not my fault ! ), so sorry for my bad English !

I have a HP dv9051 (T5500, 1.66) that produces the same annoying noise !
As a lot of users seem to have...
[On a french forum, people who have a T7200 based notebook told us their machine are not affected by this bloody noise]

I tried RMClock, it removed the buzzing sound, but I don't know how to configure the settings properly, to avoid the processor to heat up (and make the fans annoying me more than the noise I tried to stop !)

The best and easiest solution, IMHO, would be a patch from a software/hardware manufacturer (MS, Intel, ...) but it seems unlikely, so if someone has the skill to tell me what would be the right settings for my system, I would be over the moon !
Some of you seem to be very competent and reliable (I'm obviously not !), so maybe there is hope !

Thank you very much !
Jerome, France.

 

Try this guide: http://winhlp.com/rmclock.htm
With it should get you in the right direction. If the guide isn't enough, ask specific questions. I've already posted everything I have set in this thread, so maybe go back and look for that? Don't forget the readme, Dmitri put alot of effort into detailing that document so you could read it and figure stuff out! =P

 

Ok, i have recieved my new compal HEL80 now, and at first i couldn't manage to find the noise. but then suddenly to my dismay i discovered it on the headphone mini jack when the machine had awoken from hibernation. A well i thought, at least i can just avoid hibernation and reboot to avoid the noise. But alas one day upon normal boot the noise was there, after a little trial and error i found that disabling the biult in webcam removed the noise, weeeee. But then today the noise returned (on the headphone output, i have not yet heard it on my external soundcard via the firewire port on this machine)

So i read up a little on this thread and hurray! the HLT setting in RMclock relieves me of this tooth gritting annoyance. however, the cpu fan is now allways running, (and i have set RMclock's profiles thingy to performance on demand) A little tutorial on how to correctly configure RMclock and a little more info (if relevant) on that noise filtering setting you mentioned pyro9219 would be extremely appreciated, and i must say i'm impressed buy the guys/girls presenting all this useful information to this thread, thanks!

On a further note, it seems to me that the most obvious and annoying noise is the supposedly from the CPU (i guess this because RMclock fixed it for me). However through the headphone output i can also hear what seems to be HD noise (because the pitch and presence coincides with spin up of HD at startup ) and also what seems to be noise from the graphics processing unit (noise that coincides precisely with changes in the gui and nothing else (resizing windows and the likes...) RM clock fixed the first of these and for me that was the loudest, most consistant and most annoying noise.

 

@Pyro: This is brilliant !!! Thank you so much for the fast and useful answer !!! I'll have to explore those informations deeply and carefuly, but it sounds nice... even if it's only for AMD ! ;-)

I still don't understand why MS or Intel didn't release a patch yet ! I can't imagine they're not aware of this problem !

@Daniel: have a look at the link given by Pyro... It seems that the "HLT" setting in RMclock is not the perfect idea (I thought it was too, but it heated up my system and the fans ran non-stop).

I'll let you know if it works for me !

 

What's for AMD? Ignore the guide saying how to cool AMD, thats just the only guide I could find for RMclock.. It fully supports all Intel and AMD chips.

Originally these CPUidle, CPUwaterfall, CoolON, RMclock etc were all designed to enable HLT states because traditionally, AMD processors have been VERY overclock friendly, and HOT... Before we have speed throttling CPU's this was the trick we used to conserve a little power and lower the heat (if used with thermal controlled system fans you could save more power, and have a quieter system)

The HLT isn't supposed to be a fix either, just to clarify, its just something that I know reduces CPU noise, and was more or less a test to see if it helped. Like I've stated a few times, I dont have this noise, so I just take educated guesses at things to inform you guys to either help me track the problem, or to eleviate the symptoms.

As for the CPU fan... I highly recommend updating your bios using information from this thread... http://forum.notebookreview.com/showthread.php?t=90406
Please make sure to ask if you have questions about how to do this, there is risk of becoming a person without their notebook for a while till its fixed. I'll also toss in a general warning that states "if it isn't broke, don't fix it"...

This will give you the latest bios, which will change your fan settings for the better (it wont be on/off all the time, and your system should run cooler also) It also covers a few keyboard issues, as well as fixes the battery time remaining.

For those of you who need a little extra help to get going, these are all of my RMclock settings once and for all (I list drop down box selected values, and I only state boxes I have check marks in)

**********************************************************
RMclock 2.2 settings: (Use 2.2 because lower Vcore's are enabled, Lower Vcores means more battery time and less heat)
Management Tab -
Allow multi-step transitions
Sum of CPU core loads
Keep current profile
Alert when CPU thermal throttling enabled
Restore on management off
Restore on application exit
FID/VID found at startup
Show taskbar
Start minimized
Run automatically

Profiles - All the OS power management is defaults on every tab. (Standby is Never)
Performance on Demand when AC
Performance on Demand when Battery
6x 7x 8x 9x 10x all checked Range .950V - 1.175V
Adjust intermediate VID's
No changes to No Management
Power savings on both tabs: Index 0 for PST/ODCM(use both)
Maximal performance on both tabs: Index 4 for PST/ODCM(use both)
Performance on demand Battery: Index ALL PST/ODCM(use both) and Level 7
Performance on demand AC: Same as battery but on BALANCED

Advanced CPU -
All checkboxes on processor Tab
No checkboxes on Platform Tab
**********************************************************

Other then that, please take a few minutes and read the documentation Dmitri includes in your startmenu for RMclock. There is loads of information about what each check box does, and why it is there, and why you should use or not use it.

As for Intel being aware.. I work at Intel, so yes, we are aware, and if you have read enough posts know I've been trying to gather more information to make a formal report. The issues are not with the processors though, its the fundamental problem with notebooks, AMD chips do the same thing.. The faster we push the clocks (osilating crystals) the more EMI the systems have, and as time goes on stuff is shrinking more and more which means less protected and insulated circuitry. This problem is as old as computers themselves though.

If you really want something to think about, your system has components that are considered "good" with as much as 20% variance because of cheaper materials in the doping process. Thank the disposable technology revolution for this!

 

I'm posting this seperate so hopefully if anything, it gets attention after that long thread...

notice I dont use the HLT state...

Feel free to try my combination as I listed, maybe that will have a "fixed effect", or try the bios update I mention, my system is cooler now then it was with the default biios I got.

 

Read up a little? Shame, read it all! =P You missed where at some point we identified atleast "two" sources for noise

What OS are you using that you get noise from resizing windows and whatnot? Vista with glass? Linux with Compiz( or whatever they call it now)
If you are using an OS with hardware accelerated desktop, that brings me back to theorizing the fault is in an IRQ from the video device, otherwise, you got me there, I'm not sure.

Webcam would be the same... IRQ sharing problem.

Interesting thing about "Plug and Play" OS's is that you can reinstall everytime and potentially get different hardware addresses for everything everytime(within scope of what the device allows). This could be a major cause as to why some unlucky few are having these problems, while many of us dont.

I know in the olden days a soundcard sharing IRQ with either the mouse, or the HD would drive you nuts because you would hear noise all the time when either the HD spun or the mouse moved across the screen. Most motherboard guides come with the IRQ sharing table so you can see where to put stuff more effectively.

 

Wooo !!!
Pyro, you're fantastic ! Thank you so much !

Sorry, I didn't realize you were working at Intel !
Even if it's not Intel's responsablity, I hope they will sort it out (or someone else, it doesn't matter, as long as a patch is released), it will be less risky than bad settings for RMClock... Nice to hear you're gathering informations about that issue !
I know HP released a BIOS update recently, I'll try it very soon !

Cheers !

Jerome (France) ... sorry for my bad English !

 

@Pyro:
Is there any reason for this setting "No checkboxes on Platform Tab"

I myself are still looking for the optimal setting (and program) to get as little fan noise as possible.

Personally I don't bother about the sound in the headphones, it's too low to really irritate me.
For the record, I now again use RMClock to do the power management. I already use SpeedswitchXP but it is outdated, also used the normal windows profiles for a while.

My setting for RMClock (CPU is a merom T7400) are:
Management: Defaults
Profiles: Defaults and then for Profile Performance on demand I changed to Power saving level 7 for battery. Also notice that Use Throttiling (ODCM) is disabled (Default setting). Also read on the official RMClock forum that ODCM is useless? Anone want to comment on this?
Advanced CPU setting: All checked on both Processor and Platform, except the Chipset-Based CPU Throttling settings.

In idle mode the fan now runs for 20s and the mean time between 2 fans startups is 2m50s.

I think the fan startups, shutdowns are completely controlled by the bios depending on CPU temp?

BTW, why is Performance/Power Saving Preference grayed out in all profiles except performance on demand?
Edit: I think I already know the answer: this setting is only relevant for modes with more than one P-state.

 

You answered that last question yourself :]
You are also correct in that until RMclock is loaded, you are running on bios defaults. Also meaning that there is no throttling occuring, which also means heat while booting... This makes sense though if you think about it, imagine waiting for XP to boot at 400ish mhz =P (I've dont this...)

ODCM ~ There are some bad posts on that forum about ODCM, but they contain valid information. Basically the point most people make that while the proc is having enough activity to pull it from the lower states, it will always be at 100% ODCM. ODCM off will yeild you only the power states based on your multipliers, for my chip that means 1ghz is as low as it goes. Using ODCM I can drop it down to nearly 330mhz. Now the average user has music players, browsers, and god knows how many tray and background apps eating up their system, so this means realistically you wouldn't drop below 1ghz. My system though, I've got all extra processes turned off, I dont have any high activity software running, so getting to the lowered 330ish mhz isn't all that uncommon for me.. infact, for the average duraction of writing this msg about ODCM I'm hovering around 450-700mhz, and while I'm gone at work the machine sits at a nice 330ishmhz.

Just a note on CPU throttling. It's deals with changing how fast the CPU crystal is occilating. Faster it goes, the more power it needs not only for the action, but the faster you start and stop something the more loss of power you have. In theory, you could "throttle" a cpu from 0-100% smoothly, but crystals that accurate are VERY expensive, and dont make sense to buy. Instead they give you approximate values that are gapped enough to be useful. This is also because the more speed changing and whatnot you do has an impact on the life of the CPU. However, this isn't an issue because by the time you would wear out a CPU from throttling you passed time for an update, and probably dont care (5+ years)

If my system is on and in WindowsXP, I've got Teamspeaks client open, RMclock, Gaim, the wireless utility, omnipass, and my Nostromo Controller open. I normally hover around 1ghz if running music from winamp(I have all the frills turned off, no graphs or bars on the music displayed.. aka, all cpu stuff is turned off because I dont look at the window, I minimize it and use keyboard shortcuts or the trayicons I installed as a plugin)

As for the platform tab, I've decided they weren't benefitial for my needs so I dont use them (Read the documentation on the advanced tab settings) One of the options you should see is Chipset based CPU throttling, and I dont want that, I want my user activity to determine what speed the CPU will run at.

Also remember that Intel designs these chips to work in thin clients (and they keep hoping consoles) that may be on proprietary OS's, or with limited bios options and some of these features exist on the chip so that simpler options exist for their OS's or firmware to control. For instance a specialty processor that goes into a gaming system may only want to go into single 50% mode while not running a game, because throttling might be to big a risk for performance while in the gaming. By using a chipset to control this, you can use the I/O of the hardware itself to determine whats happening (A spinning DVD could keep the game at 100%, while the game itself may be playing a video from memory and drop a throttled system low, meaning it would have to speed up once the video ends to get back into the game, causing a point of possible delay)

A real world example is that once my games are loaded, I rarely stay at fullspeed. Guild Wars and BF2142 run around 1ghz if no loading content on my system.

 

Be careful with what you are asking... As I've stated, the problem isn't in the OS, its in the EMI and electrical noise generated from components that are litereally crammed together and stacked on each other. Software doesn't fix the issue... I'll repeat

SOFTWARE DOESN'T FIX THE ISSUE

What software can do, is lower power going through the system by slowing it down (or flat out lowering VCORE), which effectively lowers the noise and EMI... It's not a solution to ask the users that to fix their brand new systems that are blazing fast, that they need to run it as slow as possible.

(My first dual core was an AMD X2 +3800... I overclocked it to 2.4ghz, just to install RMclock and go HEY, THIS IS GREAT! I can run it at 1ghz while surfing the net! and while I'm not loading content I can game at about 1.2-1.6ghz!)

To put this in perspective, all of you that have less then 2gb of ram, I could say you need to upgrade, because the swap file usage that you have is causing the EMI that makes your noise problem...(my swap is hardly ever touched) HD's create massive amounts of EMI because they are high power (when compared to the rest of the system) and drive electromagnet motors.

 

Please read through more of this thread before posting questions. Poor pyro9219 keeps answering the same questions when almost all of this has been discussed already in this thread. It's not fair to make him keep answering your repeat questions.

 

I could setup my own "Ask Pyro" thing here and link to my paypal for donations... Ya know... Xmas is around the corner!

 

I couldn't find solid answers for the following questions regarding the noise from the headphones on merom cpus:

Is the noise present on merom versions of hel80?

If yes, is it always present, or only when the power cable isn't connected?

If it's always present, did you find an almost-perfect solution that works without decreasing performance?

 

There actually are answers to these questions in this thread. It just takes some reading to get to them all.

Yes, it is. It's not very loud, so perhaps it is not noticed by most, or it is not present on every system. I had to look for the sound for a while before detecting it on my system.

There are 2 different noises being discussed here.

#1​

One is the CPUs voltage making a high-pitched noise when on battery. Undervolting the CPU while on battery reduces the amount of power going through the voltage regulators and eliminates this noise. If you do not undervolt, the sound is not very noticeable, I only heard it when placing my ear right next to the keyboard.

#2​

The other noise is the one heard through the headphone jack. I also feel this is a minor issue, but for some it is a problem. We have concluded that it is related to EMI created by the HD. Power + spinning = electronic noise. This is common in many electronic systems, and if you were to look around, you'd find there are many different notebook systems suffering from the same problem. It is too costly for manufacturers to tweak every system to eliminate every single little noise. pyro9219 has suggested that insulating the HD may eliminate this noise, but to date nobody has tried it. It seems that by playing with some of the processor settings in RM clock that some have found ways to minimize the noise, so read through that part of the thread if you want to experiment. I personally found no noticeable change in the already minor sound with my own testing.

Like I said, the noises are so subtle that most do not even notice it. If you look at this thread, there are only a handful of HEL80 owners who have said they hear it. I had to look for it, and pyro9219 doesn't hear anything on his system.

This post should bring you up to date if you didn't bother to read through the thread.

 

Thanks Chris (or chrisyano), nice of you to sum up the thread for those of us that are in our exam period

however your post differs a little from my experience

heres what ive found so far.
on both the systems i've tested, the cpu related noise (#1) is present in the headphone output and it actually gets louder when the pc has its power cord plugged in. in addition, on my systems headphone output i can hear HD noise, some kind of GUI related noise, and a sparse intermittent crackling sound that only appears on one the left channel (and i have made sure final noise is not due to any headphone cables or even the plug itself, this is noise coming from within the PC and caused intermittently by the PC)

Also the Cpu related noise, on "somekindofwonderful"'s system was present via the firewire port as well as the mini jack port for the headphones and was very reduced in amplitude when the power cord was disconnected from the HEL80. but the last and possibly most important issue i want to bring fully to light although it has been touched on is that, even with the sound chain closed somewhere between laptop and speakers / PA system, the noise is still apparent and quite loud when the powercord is plugged in to the HEL80. all i can think of is that the HEL80 is varying the voltage drawn from the adapter at the frequency corresponding to the buzzing sounds pitch. meaning that it is effectively sending this buzzing sound as EMI over the power cables to any hardware plugged in. this theory supports the discovery i recently had that different combinations of audio hardware plugged in to different sockets gives different results regarding the presence of this buzz.

an example of what i mean:

the other day i was routing the analog output of another computers analog output (headphone output) through my external Firewire soundcard, (plugged into my HEL80) as soon as the jack plug was inserted the notorious high pitch buzz was present. My first thought was that this was noise coming from my freids laptop headphone output, but then i noticed that the sound corresponded to mouse movements window resizing and so on on my hel80, and when i unplugged the power cord to my hel80 (disconnecting it from the power mains that the other laptop was also plugged into) the noise ceased.

is this also a normal EMI occurance for notebooks (cos i never had an issue like that with my stationary PC) or is this a problem that should be adressed ?

sorry for the crap grammar, but its late and i have an exam to study for.

Daniel

 

The reason you don't hear this sort of noise with your desktop is that there is enough room in the desktop to keep EMI emitting components far enough away from the sensitive ones that may pick up some of the noise.

 

i was referring more to the noise that seems to be carried via the electrisity mains. I would have thought that what with all the components being in some way connected to the power supply, and the laptops power subbly being some 2 meters or so from the laptop whereas the desktop PC has the power supply built into the chassis (and in my shuttle computer it is rather cramped up against the other components), there would be less chance of the power adapter carrying EMI to the mains on my laptop. but correct me if i'm wrong.

 

I'm a little pressed for time at the moment, but a dirty answer for your statements (most of which I already answered by the way)

Unplugging power is lower "noise" Unless using a program like RMclock to force the noise, your system automagically enters a lowered power state when you use the battery. Lowering power is lowering noise.

Normal Computers place the HD's away from the exposed circuitry so you dont notice these problems, try holding your HD while its spinning against your sound card and see what happens (please dont do this and take my word for it =P)

The smaller the hardware gets and the more we cram it together, the less shielding we are using to insulate the hardware. This is just the nature of notebooks/laptops.

As for the HD being the source, I've just pinned that out for two major reasons, location in the chassis, and the fact that HD's have been known to cause EMI issues way back to the original spindle drives.

As for hearing the noises, I've tried and tried, earphones, built-in speakers, and my 5.1 set.. I can't get anything out of mine that isn't supposed to be there. As Chrisyano said, I've suggested insulating the drive using 3m EMI shielding tape (used in alot of medical computers around sensitive equipment), but I dont think anyone seems to be bothered by the noise enough to actually try that. Cheaper route is to build a shell out of aluminum foil, this needs to be done slow and with care though, not like wrapping cold pizza.. You dont want either the 3M tape or the foil touching any exposed electronics, but you do want it to completely surround the drive. Possibly even the audio ports.

I also mentioned IRQ's several times. Try looking up what I said about those for your "gui" and "mouse" related noise.. That is also nothing new, and most motherboard manuals even mention this issues, thats why the supply diagrams so you can put sensitive equipment (soundcards) on dedicated IRQ's. Major components in your system have IRQ's, but the number is limited so they get shared.

As for that, this is my last "re-post". I'm spending far to much time resaying stuff that could be found with a little time on this thread. If I was meant to restate it each time, the forum would dump my posts at regular intervals

 

I'm not aware of any power supply that doesn't do some sort of noise filtering on the rails. The rails are also shielded somewhat by the plastic housing, but as I said in my last post, its reduced for space and heat limitations.

I remember some of those shuttlepc lots that had powersupplies that exploded because the capacitors were incorrect... that was fun.. I think fry's electronics sold most of those and had a hell of a recall.

 

Ok, thanks for all the help guys.

I have read all the posts, and dug deep into the annoying world of IRQ sharing several times before, and as far as i have found there is little to be done when it comes to deciding the IRQ of my one and only FW connection. (currently my IEEE 1395 connection has IRQ 20 wich is as i understand a virtual IRQ that is probably in actual fact shared with one of the 16 real irqs, i was thinking that a possible way of giving it a new IRQ would be through a PCMCIA Firewire controller, but MY PCMCIA controller is sharing IRQ16 with my graphics card, network card, usb controller and card reader) i have also seen bioses before that allow reconfiguration of IRQ and in most cases that at least causes windows to rearrange things alittle (though rarely in the manner one tries to set in the bios) We'll check what "somekindofwonderfulls" usb and FW controllers are sharing IRQ channels with and maybe that will whed some light on the differences between our to identically equipped HEL80s

Again thanks for all the help, i dont mean to get you to repost lots of stuff.
What worries me though, is that "somekindofwonderful" is experiencing the noise frequently using firewire and usb devices. This ,as far as im concerned shouldnt happen. And the only noise im looking at removing is the CPU noise, the rest is low enough to live with. I have the exact same machine as "somekindofwonderful" (tested in the exact same context with the exact same eqiopment in the exact same room) yet our predicaments differ (for the time being).

ill do some more experiments after my exams and try to determine if in fact the machine is causing noise on the power mains. Being live musicians that use laptops this is a rather serious issue for me and "somekindofwonderful" and unwanted noise coming out of our audio equipment in any circumstance (especially recording) is highly undesirable.

Ill get back with the results of my mad scientist experiments sometime after the 11th

woopipoo

 

What about this worries you? EMI could be entering through one of the bridges, through the electrical rails, through the CPU socket area. Anywhere on the PCB on the motherboard could be the entry point, especially if their are contacts or saudering points exposed... The noise could be internally generated from the occilating of the voltage regulator, meaning that it would travel through the processor to the rest of the system.

Basically, pick any peice in the system that has voltage passed through it, and start checking for interference there

Have you tried lowering the voltage of your cpu to remove your sound?

 

I know this is a somewhat silly question, but are you referring to performance level 7 or power saving level 7?

 

Oops! Power saving level 7. It basically holds off from upping the speed until a little more time has gone by to keep the speed and voltages jumping all over the place while on battery. If enough time goes by and the processor demand lowers, it never enters the higher power states, saving battery time.

 

Thanks for clearing that up. Now I've been reading up a bit about RMclock, and it seems ODCM actually increases power consumption (and thus, increases heat) in some cases with the newer generation processors (yonah+). One example would be this post from the RMC boards: http://forum.rightmark.org/topic.cgi?id=6:797

Have you tried both with and without ODCM and had better experiences with it enabled, or have you just been using it from the start?

 

I've tried both on and off, but heat isn't what I've been watching for... I mainly go for battery life.

Realistically, if there is a difference, it must be only minutes which could also be caused by slightly different usage while on battery... I just turned off ODCM to mess with it again, but from what I've noticed its just a speed change...

 

All righty, with the help from pyro's post a couple pages back and the extra help I got above, my rig is now running 10c cooler at full load than before (68c -> 58c), which keeps the fan from stepping into it's highest mode . My idle temperature also went down by 3-4 degrees (48-49 -> 45).

I haven't taken my Hel80 anywhere on battery with the new undervolt in place yet, but I expect it should add at least a couple of minutes extra. As a guideline for other people pushing their T2500's, I can say mine BSOD'd out on me when I tried 1.0v (at 12x multiplier), so I decided to raise it to 1.05 (again on 12x, letting the program auto-adjust the intermediate VIDs) and after running Prime95 for 6 hrs, it still hasn't crumbled, which leads me to believe this is pretty stable.

Again pyro, thanks for your help and I hope what little info I could provide here will inspire more HEL80 owners to try this neat little "mod"

 

Just an update, I've been trying since yesturday without using ODCM and I dont see any real difference other then when I set to power saving at 1ghz its still realistically useable compared to 350ish mhz. Power gained is maybe a few minutes, like I said, its hard to tell because usage isn't in a control environment. I've also been running my proc with a max of 1.05v as well since yesturday. My 20 minutes of CPU-BurnIn put me at 57C, while under normal usage I'm sitting at 42C (both readings were using performance on demand)

As for the highest fan speed, I've not had that since... well, it only does it when I first boot..

 

I've been using RMClock for a while now and I seems to me use more than average CPU time for my 'mostly idling' applications running in the background.
And Is it me or does it sometimes report wrong FID and VID values? In CPUZ they always seem correct.
Other than those two comments, thumbs up!
Any comments?

 

OK, I don't understand the first thing you said. As for the FID and VID, they are most likely correct for what the values you told RMclock to use. Remember, the job of RMclock is to change the VID and FID, I'd say check your settings. CPUZ reports the same values for me as RMclock,while RMclock is running.

 

That I think it uses too much cpu%

 

What makes you say that? mine doesn't register any usage... Turn off HLT state, only use that if you are trying to reduce noise.

 

That is not on, try monitoring it with process explorer and compare the CPU time used with other background applications.

 

As I already stated, mine uses 0%-1% unless the window is in focus while graphs are up. Even with the graphs, its only goes up 1-2% max... If you literally are refering to like minutes, or hours of time compared to other processes, then get rid of the program if it bothers you, the program pretty much constantly has to poll the processor to adjust power and speed states, as well as get performance data from the core(s). There isn't a way to manage the CPU without taking a little time from it.. but like I said.. time isn't important, its demand.. which is at max that I've seen 2% while the interface is open. If 2% is killing your system's ability to do everything else, then again, get rid of it. If you feel comparing an active monitoring application like RMclock to something like the print spooler that goes idle when not in use, get rid of it. Nobody here told you to use it...

If you don't trust it... don't use it.. go use NHC.. that way you can load the entire .NET framework to manage your processor. Otherwise, let windows manage it, and you will never know how much time its taking since it will be managed by the grand XP kernel which is hidden from the reporting services.

 

I'm actually a very happy user of RMClock now that I know most of it, but I thought that was one thing that it could be improved upon.
Don't be so defensive...

Indeed .NET is really disgusting. I refuse almost any .NET application.

 

Sorry if i am posting something previously discussed but that noise from my hgl30 was fixed after uninstalling my windows battery drain fix ... anyway if there is a workaround without uninstalling this fix i would be very grateful.

 

Could you specify which fix you uninstalled? That is the first time this has been mentioned in this thread.