Noise from computer

[odjwizardo]

hi all am new to this forum thing so please bear with

here is my problem i have the
amd 6 core cpu
cool master 1000w psu
4 gig ram
ati 6950 hd
main bored is a gigbite bored

i am using the comp for my Home DJ'ing and have a mixer and a 2000w amp but when the amp is plugged in the sound card (jack plug) there is a very clear sound that changes with cpu work loads

even in mute the sound dose not change in volume

after a bit of hunting round with a mic and head phones i found the source of the interference and and was shocked to find it was coming from the cpu+heat sink even just moving the mouse is enufe to change the sound of the interference

so here is my question how can i stop the sound card picking up the interference from cpu ????

 

[MEgamer]

mmm more grounding needed somewhere.

 

[anwaypasible]

the mixer doesnt have a ground lift switch?
maybe you need a 'professional' usb soundcard with a ground lift switch.
and
maybe the usb soundcard will work without the noise.

appears to be the onboard soundcard with the problem.
and you could get an isolation transformer.. but finding them for a soundcard output isnt the easiest thing to do.

 

[odjwizardo]

i did think of that and i all so found a 10mv difference from the CPU heat sink to ground i have now got an optical decoder and connected it up this stopped the noise getting to the mixer just fine but the mic leads and leads to the amp are still picking up the noise (the leads are balanced xlr)

 

[MEgamer]

if the balanced are still picking up interference, thent hat means the inteference must comes from some source. if the leades up connected between the amp, and teh mixer, is the noise not coming out from the mixer???

you could also try power conditioning, every single device you have.

 

[odjwizardo]

i have put a scope on the amp and the signal is digital in nature the mixer is analog when the PC is off all channels on the mixer are clean only when the PC is on dose the noise come throw.

iv been reading other forums relating to sound problems and you need to bear in mind that the amp is 2000w and will pick up radio signals very well if not grounded i wish i could post a video/sound clip so you can hear it for yourselves

i make and test my own leads all leads are double screened

 

[anwaypasible]

PC soundcard had a problem with analog.. meaning there is probably a problem with the digital too since the ground can be the same.
not saying you shouldnt have tried with the digital output..
only that the problem could remain because of the soundcard.. and the problem could come back once again with the optical decoder.

why not simply go get one of those creative emu usb soundcards?
it has balanced xlr inputs and balanced trs outputs.
the balanced trs output appears to be a 'stereo' plug connector, with the typical three different spaces (one for ground, one for left, one for right)
but used as a balanced connection instead.
xlr is three prong.. so i dont see why you cant find a trs to xlr adaptor.

doesnt appear to have a ground lift switch.. but that doesnt mean it needs one.
especially since you could run your mic through the device and straight to the amplifier.

the mixer you have should really have something dedicated to ground loops?
maybe the power conditioner will help.
maybe the motherboard design is so bad that you need to remain isolated from it.

i mean..
maybe you add a resistor or something to the heatsink for trying to get the voltages to match.. and maybe your motherboard hates you for it.

 

[odjwizardo]

the optical decoder is noise free if i run it from its own power adapter 240v to 5v but if i run the thing from the usb port the noise comes back so i will stick to the 240v adaptor
i had thought of grounding the heat sink but i don't know what would happen to the cpu or main bored if i did that

to brake things down this is whats going on

pc to mixer clean signal

mic to mixer picking up noise cable is screened (two mic's both picking up noise one more than other as one is further away from the heat sink but noise is same type)

with both mic's unplugged all Chanel's clean noise free

from mixer to amp left and right have noise same noise as mic's had the leads run past the cpu heat sink

if pc is off all Chanel's are clean (Mic's to)

the noise signal is so strong its knocking out my net work 1GBPS to 100MBPS and that's with cat6e i have cat6a on its way hope this will fix that

the hole system is in a u10 case

 

[anwaypasible]

maybe the motherboard isnt all that bad to itself.. but is radiating a lot of electrical noise.
your computer case might be reflecting that noise back onto the motherboard like a mirror.
and
maybe your power supply isnt as noise free as it should be.


generally..
the reason these noises happen is because there is too much electricity in one spot and not enough electricity in the other spot.
you either raise one or lower the other (whichever is safer)
but
the alternatives to fixing the problem would be discovering why the electricity is higher in one spot and lower in another spot.

to say that the part on the motherboard is junk.. that is one conclusion.
to say that the circuit trace is too small or the resistance is wrong.. that is another conclusion.
to say that the power supply is injecting noise into the motherboards circuits.. that is another conclusion.
to say that the computer case is amplifying the problem.. that is another conclusion.


i would suggest removing the motherboard from the computer case and putting the motherboard on some cardboard or something so there is no chance of things being wet ..or.. no chance that two of the metal pins on the bottom are touched together.
this might mean glass.. plastic.. a piece of fabric.. cardboard.
nothing wet.. and nothing metal.

with the motherboard out of the computer case.. plug it all up and turn it on to see if the noise is still there.
if it is, then you know it wasnt the computer cases fault.
then you look at the power supply or the motherboard itself.
when the power supply is injecting dirty electricity into the motherboard, you could be killing all of the parts plugged into the motherboard because of the dirty electricity.
so you really should discover.

the funny thing about this situation..
when you hear the noise from the computer processor.. that electrical problem might not exist unless you plug something up into a weak area of the electrical circuit.
to say that the whole thing runs okay without being touched, and when you plug something in.. extra electricity is needed, but there isnt anything there to provide the extra electricity.
like i said.. it could be a junk circuit trace.. and this means you could solder a wire from a power pin to the source of the weakness.
if the piece doesnt wish to accept full power with the extra load from the plugging in of the mixer, then the piece will fail.

junk parts on the motherboard are simply that.
and if the power supply is feeding the motherboard dirty electricity.. good parts can look like junk.

 

[odjwizardo]

i have bench tested my pc as i do with all my builds before i put them in to my case
i have tested the psu on a scope and all is as should even when running at near full power the psu is 1000w cooler master pro
this noise has been with me sines the beginning but can only be herd with head phones which were plugged in the my pc speakers
the sound/noise changes with cpu % usage like opning a web page playing games or just scroling the mouse

 

[anwaypasible]

hard to say the impedance characteristics of the motherboard are anything like the oscilloscope.
and you say you tested the reliability of the computation.. but didnt say anything about testing the audio with the headphones.
i can guess that some front panels are easier to remove than others.

anyways..
noise from a soundcard that goes in tune with the cpu is nothing new.
to say it happens to people somewhat often.. those people had to clean it up too, or live with it.
i am willing to bet those people bought an aftermarket soundcard that has higher quality power filtering capabilities to prevent the noise.
or
maybe the aftermarket soundcard has made a specific attempt to block electrical noise.

truth be told.. those soundcards dont have the ground issues.
they can be forced to make noise when too much is connected.
but
that usually means a receiver, a cable box, a television because of the cable box, the graphics card to the television.
things of this nature can cause a problem.. but i didnt experience an audio hiss, what i got was a horizontal 'hum' on my television screen.
seems like connecting the television to the computer was one step too much.

 

[odjwizardo]

found out something you may find interesting i found the noise round my cpu so i took a chance with my main bored with my scope had to set it to pic up round the 3khz rage to see it but i found the path of the noise on test pions on my pc and its every were how ever when i put a smaller heat sink on the noise was much less on the bored and on the sound out but is dose not affect the working of the pc as the digital signal is emuown to the noise so to stop the noise i have built a stainless steel faraday cage round the cpu heat sink the noise is all but gone seams the bigger the heat sink the more noise radiates out over the bored people with water cooling should not have problems of this kind its taken some time but all channels are now clear

 

[anwaypasible]

does that mean the heatsink is somehow attaching itself to the ground of the motherboard?
like.. when the bolts go into the motherboard holes to hold the heatsink down firmly?
i could imagine that being why the different size heatsink is changing the amount of noise.. because there is less metal on the smaller heatsink, there is less of a circuit fluctuation.
maybe you could completely eliminate the noise with plastic washers to cover the metal around the bolt hole?

or

do you think the heatsink is simply acting like an antenna somehow and distributing the electrical noise from the heatsink down to the processor?

if it isnt the bolt holes causing circuit connectivity..
maybe it is the heatsink touching the processor?
i know the top of the processor is metal.. and if it electrically touches the heatsink, that would be like a circuit connection.
question is..
would things be more stable if those two pieces didnt touch electrically.. or is it better to let them touch electrically?
because..
maybe you could switch thermal pastes to something that is less electrically conductive.
arctic silver has metal in it.. and that should mean it is electrically conductive.
the ceramique stuff is also electrically conductive, since we have ceramique heaters and peltiers.


makes me wonder if the heatsink was acting like an antenna and picking up the noise from all of those points you tested.. and then injecting that noise by touching the processor plate.. or because the motherboard has metal rings around the bolt holes for ground.
you said the steel faraday cage around the heatsink has almost completely eliminated the interference.
but
i wonder if the heatsink was actually being an output rather than an input.. and the noise from the heatsink was being captured by all of those places you tested (if not more places).

is the harness of the heatsink plastic that holds it down firmly?
maybe the different bolts have a different electrical conducitivity and is coming into contact with the electrical circuit.
and if the harness is metal.. that would say more.

watercooling isnt all that bad.
as long as you can tighten the hoses as tight as you possibly can without the metal tubes collapsing.. the whole thing should remain water tight and not be scary anymore.

how did you manage to keep the heatsink cool?
is there a fan on the top of the cage or ..?

 

[odjwizardo]

the heat sink was not made for my main bored and hat to buy a kit to fit it the heat sink is here http/www.overclockers.co.uk/showproduct.php?prodid=HS-005-PL&utm_source=froogle to test my theory i put the heat sink that came with the cpu and yes there is a an electric connection to the cpu as with most mosfets and transistors the case is often + in Volts and i did find a 10mv signal there so may be a thin glass plate which is old school the only problem with that is that if the glass is not flat or the heat sink not flat/cpu the it could crack not good as for the clamp its all cast alu with big springs and bolts to hold it down id not want to put any plastic round the bolt holes just in case of static build up its not likely but better safe than sorry

the cage i made is a mesh so the fan is inside the cage to and air flow is not affected