Motherboard fan headers let your PC adjust cooling as temperatures change. Correct connections and a sensible fan curve can keep everyday work quieter while providing stronger cooling under load. Motherboard fan headers and fan curve setup are closely connected: before adjusting a curve, you need to know which fan the header controls, how it changes speed, and which temperature it follows.

Identify the motherboard headers

Open your motherboard manual and find its fan-header diagram. The labels beside the connectors help you locate them, but the manual explains their electrical limits and control options. Similar-looking headers may behave differently.

• CPU_FAN: Normally serves the CPU cooler fan. The motherboard often checks its speed signal during startup and warns when that signal is missing.
• CPU_OPT: Often serves a second CPU cooler fan. It may follow CPU_FAN instead of offering independent control.
• SYS_FAN or CHA_FAN: Usually serves case fans. Each header may have its own curve and temperature-source options.
• AIO_PUMP or PUMP: Intended for a cooling pump. Its default output and current limit may differ from other headers.

Shut down the computer, switch off the power supply, and disconnect mains power before moving connectors. Align the fan plug with the header’s plastic guide and avoid forcing it. Route cables away from blades. A fan’s lighting cable belongs on its appropriate lighting connection, not a fan motor header.

Choose PWM or DC mode

A standard three-pin fan has connections for ground, power, and a tachometer signal that reports rotational speed. DC control changes its supply voltage to adjust speed. A standard four-pin PWM fan adds a dedicated control signal while maintaining its normal supply voltage.

Use DC mode for a standard three-pin fan and PWM mode for a standard four-pin PWM fan, where the header supports those modes. A three-pin plug can fit a compatible four-pin header using its guide, but that does not guarantee the header supports voltage control. Proprietary fan connectors need their intended controller or a compatible adapter.

Automatic detection can help, but confirm it works. Change the requested output in the firmware and check whether the reported RPM changes. A three-pin fan connected to a header operating in PWM mode may receive constant voltage and run at full speed. The curve’s output setting is a control request, not a guaranteed RPM; different fans respond differently.

Use splitters and hubs within their limits

A passive splitter allows multiple fans to share one header. The header supplies the combined electrical load, so add the fans’ specified maximum current requirements and compare the total with that header’s documented limit. Account for startup requirements where specified and leave headroom. Do not assume every header has the same rating.

A powered hub can obtain motor power from the power supply while receiving a control signal from the motherboard. Follow the hub’s wiring instructions and respect its total and per-port limits. Many powered PWM hubs supply fixed voltage, so they cannot regulate three-pin fan speed.

Typical splitters and basic hubs return only one fan’s speed signal. Connect a working fan to the designated sensing port. The other fans may receive the same control command, but the displayed RPM cannot confirm that they are all spinning. Check every fan visually during setup.

Handle pump connections separately

Follow your liquid cooler’s connection diagram. Some pumps receive power separately and use a motherboard lead only for monitoring or control. Use the manufacturer’s specified pump settings rather than applying a quiet case-fan curve. Arrange CPU_FAN monitoring according to the cooler and motherboard instructions so a missing speed reading does not hide an actual cooling problem.

Select the temperature source

Enter the BIOS or UEFI hardware-monitoring menu and locate fan control. Select one header at a time and identify its available temperature sources. CPU temperature is the usual starting point for a CPU cooler fan because it tracks the component being cooled.

Case fan controls may offer CPU, motherboard, VRM, or external sensor readings. Find out what each label measures. A motherboard sensor may respond slowly and may not represent heat near the graphics card. A CPU-based curve can remain relatively quiet during a graphics-heavy task even when the case needs more airflow.

If firmware lacks an appropriate sensor, maintain a dependable baseline case-fan speed and verify temperatures during your actual workloads. Compatible operating-system software may support graphics-card temperature or combined sensor controls. Establish a useful firmware curve first so cooling works sensibly before the software starts.

Set up the fan curve step by step

1. Record the starting settings. Note each header’s connected fan, control mode, temperature source, and existing profile. Save a firmware profile if supported.
2. Run calibration where available. A fan-tuning routine can help discover the connected fan’s operating range. Confirm the routine is appropriate for that device before applying it to a pump header.
3. Find a reliable minimum. Lower output gradually while checking RPM and visible rotation. A fan may keep spinning below the output it needs to start from rest. Verify restart behavior and set the minimum above that uncertain region.
4. Keep the lower section gentle. Use relatively steady output across normal idle and light-work temperatures. Steep changes around this region can turn small temperature fluctuations into repeated audible surges.
5. Increase output as sustained heat rises. Make the curve climb through the temperatures your system reaches under load. Its upper section should provide strong cooling before the relevant component approaches its documented thermal limit.
6. Save and verify. Check that each affected fan responds before moving on to workload checks in the operating system.

No universal curve fits every system. Fan behavior, cooler capacity, sensor location, and room temperature affect the result. Begin conservatively and change one header or fan group at a time so you can understand the effect.

Smooth the response and check operation

Look for step-up, step-down, response-delay, or hysteresis settings. Hysteresis introduces a temperature gap between increasing and decreasing output; delay controls affect how quickly the controller responds. Read the firmware’s help text because names and implementations vary.

A modest response delay can smooth brief temperature spikes. Allowing fans to slow down gradually can also prevent repeated ramping after a task ends. Avoid settings that leave cooling unresponsive to sustained heat. Use fan-stop features only where supported, verify reliable restarts, and follow CPU cooler guidance.

With the case closed, check idle behavior, a CPU-heavy task, and a graphics-heavy workload. Monitor relevant temperatures and RPM long enough to see whether they settle. Stop if a required fan stalls or temperatures approach component limits. If high fan output still cannot provide adequate cooling, inspect the cooler and airflow rather than relying on curve changes alone.

Reboot to confirm settings persist. If fan behavior changes after login, check whether software is overriding the firmware and avoid competing control utilities. Keep CPU fan warnings enabled where practical; adjust a low-speed warning threshold only after confirming normal operation. Save the finished profile and revisit it after replacing fans or resetting firmware settings.