PSU Buzzing noise when Aircon turns on?

Discussion in 'Electronics & Electrics' started by interlopr, Dec 17, 2014.

  1. interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    So, I recently just finished building a new PC, I noticed when ever my Aircon unit is turned on, My PSU emits a loud buzzing noise.

    It only happens when the AC is on.. Any idea whats going on? PSU issue? or something more serious maybe with the house wiring?
     
  2. OP
    OP
    interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    Just tried the PC with different outlets in the house, Directly plugged into the wall and via anti surge boards, Still the same noise when ever the AC is turned on..
     
  3. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    Is the new AC unit one of the "inverter" kinds i.e. does the compressor and/or fan spin at variable speeds?

    If so does the frequency of the buzz change as the motor(s) in the AC change speed?

    Is there any interference in say your TV or any other appliance?
     
  4. OP
    OP
    interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    It's a Fujitsu inverter, no change of the frequency just a solid buzz sound regardless of AC settings.

    No other interference on any other appliances that's why I'm certain it's my computers power supply..
     
  5. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    To confirm that you could try it in a different house. It could just be co-incidence that the PSU went screwy at the same time the A/C was installed.
     
  6. OP
    OP
    interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    I think you misunderstood what I said :tongue:

    Power supply is the new unit, Not the AC. Have had multiple PC builds in this house and none have never had any buzz noise like this one.

    I just don't get why when ever I turn the AC on, It makes my power supply in my computer buzz.

    Here's a short video of the buzzing noise

    https://www.youtube.com/watch?v=5UbzI1iO0lk

    EDIT: Just to be clear, It only buzzes when the Aircon is ON, It does not buzz when my Aircon is OFF
     
    Last edited: Dec 18, 2014
  7. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    There's only two things having the A/C on can do:

    Cause the mains voltage to the PSU to drop.
    Impose harmonics onto the mains.

    The former seems the most obvious.

    That'd be annoying. My old NAS developed that kind of noise, it's usually caused by a loose inductor / transformer. I fixed my NAS PSU with some varnish as it was out of warranty. As your PSU is new I'd get it repaired / replaced, it's quite possible the problem developed in transit.

    If you wanted to test my theory try pluging in something else that draws a fair amount of current e.g a heater.

    Also if you've got a multimeter check the mains voltage with A/C on and off.

    [edit] Reason I'm suggesting you try to determine how the A/C is causing the fault to come to light is if you're going to return it under warranty it helps a lot if you can describe what causes the problem else some tech might just plug it in and conclude there's nothing wrong with it.
     
    Last edited: Dec 18, 2014
  8. OP
    OP
    interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    Thanks for your help, I'll be sure to grab a multimeter and check.

    My AC has a setting on it for it to just blow Air, And when that is on, The buzz doesn't happen, Only when Heating or Cold air is applied.

    I'm going to head over to a friends house tomorrow and test my system there and load up there house to see if I can replicate the issue
     
  9. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    Air only means the compressor isn't running and that's what draws all the power.
     
  10. LINUX

    LINUX Member

    Joined:
    Sep 18, 2001
    Messages:
    3,062
    Location:
    Dudley, Newcastle
    When the mains voltage drops the PSU will draw more current, it is a constant power device (ideally). This would explain why it buzzes (more?) when the ac is on. It may also indicate iffy wiring in your house, but that's something a sparky should comment on, I'm an engineer.

    One thing you can do is a load test. Unless the AC is, like, halving the mains voltage (it won't be) you should hear the buzzing with Furmark+hyper-pi running. If that doesn't cause the buzzing I would suspect the AC's inverter causing electrical noise but you'd need an oscilloscope to measure it easily.

    Also: if you record the buzz and do a spectrum analysis in Audacity or similar it will provide information. If the buzz is not a 50Hz harmonic (100/150/200Hz etc) then it is more likely to be a noise issue than a voltage drop one.
     
  11. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    Nothing likely to make a sound inside a SMPS runs at mains frequency.
     
  12. LINUX

    LINUX Member

    Joined:
    Sep 18, 2001
    Messages:
    3,062
    Location:
    Dudley, Newcastle
    The input RFI chokes would (although I'll admit that a lot of PSUs don't have them); the switching frequency would be high enough to be described as a whine, not a buzz.

    Edit: The buzz is several peaks around 100Hz, they're spaced apart by about 3Hz (97Hz, 100Hz, 103Hz) so my initial guess is that the input RF choke's windings are slapping against each other at every current peak:


    Click to view full size!


    On the linear plot you can see the noise around 6kHz, this noise is dominated by a bunch of peaks, all 100Hz apart.


    Click to view full size!


    I can't be certain without having the PSU on a workbench, but my educated guess is that 100Hz nature is from the input current waveform's DC rectified peaks (this is why swichmode PSUs have rubbish power factor) and the 5-6kHz component is around the natural frequency of the choke that's being excited by said current peaks. ie: if you "hit the choke with a hammer" it would vibrate at around 5.5kHz for a short time.

    I don't know what would be modulating the peaks around 100Hz. It's characteristic of intermodulation distortion, when you put two sinusoids at frequencies f1 and f2 into an amplifier and measure the output at integer multiples of the difference of these frequencies. Eg: a amp driven at 100Hz and 103Hz would produce (hopefully small) peaks at 106Hz, 97Hz, 109Hz, 94Hz etc. What the hell is modulating your PSU at 3Hz I've got no idea, it could be a distortion product in your recording gear or something related to the way the PSU was being loaded, dunno.
     
    Last edited: Dec 19, 2014
  13. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    Problem is the recording is of every sound coming from the PSU and one should probably apply an A weighting curve to properly see the more annoying part of the spectrum.

    Many annoying sounds from inductors / transformers can come from the winding itself rather than the core. If you want to see this in action if you look carefully at the coiled filament of an incandescent lamp you maybe able to see it vibrate. The general fix is anything that'll stop the wires from vibrating e.g. varnish, epoxy, hot snot, cold snot, silicon. For power transformers with laminated cores loose laminations or just magnetically induced vibration can cause a pretty distinct hum but that's not very annoying.

    PS. probably not a great idea to whack a ferrite core with a hammer, they can shatter with sharp shards flying off :)
     
  14. LINUX

    LINUX Member

    Joined:
    Sep 18, 2001
    Messages:
    3,062
    Location:
    Dudley, Newcastle
    Certainly my analysis is very hand wavy, OCAU forums aren't exactly IEEE Spectrum :p.

    But since the buzzing only happens (or is only noticed) when the A/C is on and the buzzing has a strong 100Hz component it points the finger at a voltage drop issue. I wouldn't bet money on it, but I would whip out the meter and measure it before thinking/testing other possibilities.

    Another consideration is that if the voltage is dropping when the OP's wiring is loaded to ~10A (I'm guessing the A/C is ~2.4kW?) it *may* indicate a wiring problem which *may* be causing a connection to spot heat which could lead to some kind of catastrophic failure like a fire.

    Lots of low probability dots to connect, but since a fire is a possible outcome a risk matrix analysis would lead someone to have things checked by an electrician.
     
  15. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    9,242
    Location:
    4109
    Large DC offset caused by a faulty inverter drive ? Another hand wavey plausible moment :)
     
  16. OP
    OP
    interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    Okay, so I tested my system at my brothers house, and my parents house and wasn't able to recreate the issue.

    Does this mean something funky is going on with my houses power?
     
  17. desertstalker

    desertstalker Member

    Joined:
    Feb 5, 2005
    Messages:
    386
    Location:
    Melbourne
    Or with the AC unit
     
  18. RobRoySyd

    RobRoySyd Member

    Joined:
    Jan 24, 2008
    Messages:
    8,248
    Location:
    Sydney
    As there's no magic smoke coming out of the AC unit I'd be checking mains voltages and earthing first. That's a pretty basic kind of test. If that reveals nothing THEN look elsewhere.
     
  19. OP
    OP
    interlopr

    interlopr Member

    Joined:
    Sep 3, 2011
    Messages:
    216
    Location:
    Sunshine Coast QLD
    This may sound silly, but could someone guide me on how I go about correctly testing my mains voltage with a multi meter?
     
  20. LINUX

    LINUX Member

    Joined:
    Sep 18, 2001
    Messages:
    3,062
    Location:
    Dudley, Newcastle
    The mains voltage is measured between the active and neutral (the top two sockets): http://www.physics.usyd.edu.au/teach_res/inst/safety_files/image009.gif

    Your multimeter will need to be set to AC volts, denoted with the sine wave symbol: http://www.pedalpowergenerator.com/images/ac-dc-meter-symbols.jpg

    Naturally the meter needs to be set to a high enough range, unless it's auto-ranging.

    There's also the note that you should keep your fingers the hell away from the meter probe tips. To be *really* safe you can insert the probes into an off power point and only switch it on when your fingers are clear. I'd do this if you only had a cheap meter; I wouldn't trust my life with cheap probes (low risk but sever consequences etc).

    If the meter probes are too short to make contact inside the power socket you'll need new probes, don't do anything stupid to get around this.
     

Share This Page

Advertisement: