Hey guys, I have Corsair RM550x that is turning on/off its fan in 65-70 min cycles when PC is on idle. When PSU fan is off I meassure 31W from wall and 21W when fan is off. Other PSUs are not like that and they are around 25W. Funny thing is that fan shouldnt even spin below 220W iirc and I found it hillarius that its using less power when fan is on. PC is working fine, but is this normal?

Edit: specs - MSI Z270-A PRO, intel G3930, 8gb ram, 2 sata SSD and no GPU

Edit 2:

Thank you all for helping me out!

I believe that no flow in case caused PSU overheating and due to lower efficiency at higher temp PSU was using more power. Enabling case fans with small positive pressure fixed the problem immediately and now its stable at 24W on idle. On the next diagram PC was idle for only short period of time, so you can also see power curve when it was running some services, but still “much” less than 30 W.

Edit 3: Before this post I asked Corsair about that, but got only bot level responses. Conversation was going in circle and then I gave up :D

  • Hotzilla@sopuli.xyzEnglish
    16·
    1 year ago

    10W is pretty much nothing, most likely just the consumption of the fan. Sounds like that the temps increase, which then starts spinning the fans, which decreases the temps, and stops the fans. If this is annoying to you, change the temperature curves to slowly spin on idle, so that air flows even on idle.

    Edit: didn’t realize that numbers are other way around, maybe the other fans cause the airflow when small load is on the system, and this increases the consumption, and adds little bit airflow. Would explain both

    • rambos@lemmy.worldOPEnglish
      3·
      1 year ago

      Only cpu fan is spinning and case fans are off. I can tell how much power it use just by looking at PSU fan.

      Btw, is it possible to adjust fan curve on PSU? That might be a solution.

      Im not worried about 10W and I also dont mind if fan would run constantly since its still silent, but feels like something is wrong and its still under warranty. Why would 550W PSU with 0 RPM mode overheat at 30W and start the fan. They dont provide fan speed curve, only fan noise curve, but still I believe that 0RPM should be up to 220W. And true that they dont even provide efficiency below 10% which is way above the power on idle.

      • Hotzilla@sopuli.xyzEnglish
        3·
        1 year ago

        To My knowledge you cannot control the PSU fan curve, it is most likely controlled by the output power, or heat. Fan spinning itself is not indication of overheating, it is part of normal operation.

        If your case fans are off, it might be that temp inside the case is increasing slowly, and the psu fan is the first to trigger. Case fans in my opinion should be always on in idle to provide base fresh airflow. I would change their curves first and see does this remove the psu fan cycle.

      • rambos@lemmy.worldOPEnglish
        2·
        1 year ago

        I forgot to add specs, intel 3930, 2 sata SSDs, 8 GB ram and no gpu. CPU is at 1% and <40°C

  • TerryMathews@lemmy.worldEnglish
    6·
    1 year ago

    OP, couple things:

    The trigger for that fan on signal is going to be heat, not wattage. Granted those things usually go hand in hand but there are a handful of reasons that could be responsible for the behavior you’re seeing:

    • Your PC is exhibiting an unusually high draw on one depreciated rail and pushing that part of the PSU near it’s limit. Legacy components can be good for this, -12V is a great example of a depreciated power supply voltage.

    • You’re seeing heat from something else soaking into the PSU. GPU is a good internal example. If you’re an audiophile, headphone amp is a good external one especially if it’s sitting on the tower.

    • PSUs are designed to allow other case ventilation to pass through them for cooling. You’ve probably got another non-PSU cooling issue - clogged filter, dead or dying fan, negative pressure.

    Lastly, it’s hard for us to say whether that power draw is appropriate without knowing more about your PC. In general, for a desktop with a dedicated GPU it’s in the ballpark but again hard to nail down specifics. It’s a big ballpark. Things like what idle states are supported and enabled can easily account for 5W.

    • rambos@lemmy.worldOPEnglish
      11·
      1 year ago

      Yeah sorry , its MSI Z270-A PRO, intel G3930, 8gb ram, 2 sata SSD and no GPU. Ill add that to main post. Thanks for input mate, its bit more clear already, but still weird tbh. All that could explain fan behaviour, but why is it using less power when fan starts.

      Its true, my case fans are off since Ive set them to start spinning when CPU reach 45C and that is almost never. I guess Its not good to have 0 air flow. I should change case fan curves and see

      • TerryMathews@lemmy.worldEnglish
        1·
        1 year ago

        MOSFETs are less efficient outside their thermal envelope. At the micro-scale you’re talking about, you could be chasing loss of efficiency due to heat.

        • rambos@lemmy.worldOPEnglish
          2·
          1 year ago

          PSUs are designed to allow other case ventilation to pass through them for cooling

          loss of efficiency due to heat

          Your tips led me to a solution, but some statements in other posts are also correct.

          I believe that no flow in case caused PSU overheating and due to lower efficiency at higher temp PSU was using more power. Enabling case fans with small positive pressure fixed the problem immediately and now its stable at 24W on idle. On the next diagram PC was idle for only short period of time, so you can also see power curve when it was running some services, but still “much” less than 30 W.

          Thank you a lot and have a nice day !!!

  • MooseBoys@lemmy.worldEnglish
    4·
    1 year ago

    I’m guessing you have a periodic background task that runs every 70 minutes or so. This briefly increases the system power draw, eventually leading to the system heating up and the fan kicking on. It’s probably just a coincidence that the task ends (reducing the power draw) around the same time the fan kicks on.