new gtx 280 card, psu makes whining sound now

[dmc919]

never heard it before. its a 600watt ultra xfinity. when i play games or run a benchmark, the psu makes a whining sound. i thought i had enough power for this card.

[Rich'[ard]]

i've read about this before. are you sure it's your PSU, because i think it's the GPU that's making the noise. it's something to do with the chip, not the fan.

EDIT: ok i'm not so sure now, after i found nothing about whining GTX280s on google lol.

EDIT 2: found something. is your card oc'ed over 650mhz?
http://www.overclock.net/nvidia/425834-new-gtx-280-makes-very-high.html

[diaz]

Could happen that the card's heatsink isn't fastened properly, or could also be PSU like you say.

-D

[dmc919]

it's the evga superclocked at 621mhz, i used a tube from my ear to the case to see where to sound it coming from.

[sup3rcarrx8]

My GTX260 has the same problem under full load. It could definitely be the card, but i just learned to deal with it since I have my speakers on when I'm gaming anyways.

[AngryArtichoke]

A few cards I've built have exhibited problems like this in their voltage regulators. While the whine was annoying, it didn't actually affect the performance (not always the case).

Usually the whine comes from a poorly designed switcher, high frequency oscillations in a voltage regulator (not a problem), an improperly attached component (usually an inductor), or a dying cap in the case of power supplies. I'm sure there are other threads floating around here that go into why some components whine.

Is there an better versed EE around here that can elaborate on the whine a bit more?

[sup3rcarrx8]

I would record the sound if I still had my digi cam =/

[Super Nade]

Is there an better versed EE around here that can elaborate on the whine a bit more?

Not an EE any more but here is my 2c...:)

The whine is usually from an inductive element in an RLC circuit. If the conditions are right, some of the radiated EM energy couples to metallic elements which re-radiates sub-harmonics of the switching frequency that fall in the audible range.

[RollingThunder]

Not an EE any more but here is my 2c...:)

The whine is usually from an inductive element in an RLC circuit. If the conditions are right, some of the radiated EM energy couples to metallic elements which re-radiates sub-harmonics of the switching frequency that fall in the audible range.

S/N,

Can it be fixed by a layman with proper instructions from you?

[diaz]

Not an EE any more but here is my 2c...:)

The whine is usually from an inductive element in an RLC circuit. If the conditions are right, some of the radiated EM energy couples to metallic elements which re-radiates sub-harmonics of the switching frequency that fall in the audible range.

S/N,

Can it be fixed by a layman with proper instructions from you?

RLC circuit = Resistor Inductor and Capacitor circuit / kind of a resonator circuit.

EM = Electro-Magnetic

"Couples to metallic elements..." = Electron flow from the electro-magnetic energy is transfered to nearby metallic elements, which in return react to the electron flow by dissipating that energy through resonance / or vibration, based on the element's natural resonant frequency. Since that vibration creates sound, there is the inevitable sound wave creation. When a sound wave is created, it has higher order, and lower order harmonics. These are simply higher or lower frequency reproductions of the original sound, but fractioned in frequencies. The "sub-harmonics" he is refereing to are most likely the lower order harmonics (lower frequency) that fall in the audible range (hearing range), which is under 20KHz.

-D

[Super Nade]

S/N,

Can it be fixed by a layman with proper instructions from you?

I'm not sure. I have not tried it on an SMPS unit. The first thing I would suggest would to try and dampen the mechanical aspect by adding some glue between the filter caps. Electrically eliminating the noise would in principle entail changing the resonance frequency. This can be done by either changing the resistance of the loop (re-touching solder joints) or by replacing a passive element or two (caps or inductors).

[diaz]

I'm not sure. I have not tried it on an SMPS unit. The first thing I would suggest would to try and dampen the mechanical aspect by adding some glue between the filter caps. Electrically eliminating the noise would in principle entail changing the resonance frequency. This can be done by either changing the resistance of the loop (re-touching solder joints) or by replacing a passive element or two (caps or inductors).

Sometimes clocking to a different frequency could solve the issue, unless its a power filter... then what you said is the only way.

-D

[Super Nade]

Do you mean, clocking=>more current=>changes duty cycle of the switcher=>changes resonance condition? That is a very nice idea. I have not had a problem like this on computer equipment so I can't say I have tried this.

[diaz]

Do you mean, clocking=>more current=>changes duty cycle of the switcher=>changes resonance condition? That is a very nice idea. I have not had a problem like this on computer equipment so I can't say I have tried this.

Nothing wrong with trying. Also disable CPU and PCIe spread spectrum to narrow the frequency band / duty cycle fluctuations as much as possible...

-D

[hank123]

I have made old dual core intel chips scream before when I oced them.
Now I know why.

Thanks nade and diaz

[jmdixon85]

My 8800GT made some strange whining noises when it was new. Quite as a mouse now tho. Strange eh.