For a while now, I’ve had need of one ‘over-the-top’ instrument cooling fan suitable for heavy use. I reached the end of a lengthy search as I chanced upon a couple of useful PC CPU fans and PC Case fans. Since the computer hardware store had them on clearance, I managed to add all of them to my shopping cart!

Within a week the packet reached my doorstep. Naturally the first step was to open the carton to see the contents. After breaking the security seal and opening the box, I quickly found one 4-wire Intel CPU fan on top. Most surprisingly, that’s exactly what my first pick was!

Now I’ve got the key part down, time to scale my basic idea up. But first, I will put my attention entirely on the 4-wire fan. And while this post is primarily focused on its speed control, the cognition took on can be utilized later in certain power electronics projects. Before we go ahead, we need to learn the technical aspects of a 4-wire fan.

4-wire fans explained

A 4-wire CPU (motherboard) fan is a standard brushless dc (BLDC) fan, but has four wires extending from the casing. Each of these wires serves a specific purpose. Note that Chassis and CPU fans may use either 3-pin or 4-pin power connectors. 3-pin connectors are usually used for the smaller chassis fans with lower power consumption, while 4-pin connectors are usually used by CPU fans with higher power consumption.

The simplest PC fan requires just two wires – one wire provides power (12V), and the next wire is ground (0V). In a 3-wire fan, the extra (third) wire is the ‘tachometer’ signal (Tacho) wire that indicates how fast the fan is running. Here, the fan speed is typically controlled by increasing or decreasing the voltage over the power wire, as in the case of a common 2-wire fan. However, the 4-wire fan is a little different as it has the fourth ‘control’ wire to input a signal to the fan that commands it to speed up or speed down. Simply, a 3-wire fan reports its speed, and a 4-wire fan lets regulate its speed by an external signal, usually coming from the motherboard.

Technically, a 4-wire fan is called as “4-Wire Pulse Width Modulation (PWM) Controlled Fan”. Intel developed connection standards for 4-wire fans. Following is the connector pin out and wire color code of a typical 4-wire fan. Signal description is included as well.

Here, GND is 0V and +12V is the positive supply for the fan. Sense (tachometer/tacho) delivers two pulses per revolution of fan. Tacho output is an open-collector output thence requires one external pull-up resistor. Control (PWM) is the input for PWM pulses. Base PWM  frequency is 25kHz but it’s acceptable from 21kHz to 28kHz.

Intel specification (July 2004, Rev 1.2) defines the intended operation of a fan that implements the Pulse Width Modulation (PWM) control signal on the 4-wire fan interface. The following requirements are measured at the PWM (control) pin of the fan cable connector:

• PWM frequency: Target frequency 25 kHz, acceptable operational range 21 kHz to 28 kHz
• Maximum voltage for logic low: VIL = 0.8 V
• Absolute maximum current sourced: Imax = 5 mA (short circuit current)
• Absolute maximum voltage level: VMax = 5.25 V (open circuit voltage)

Below you can see a specific and detailed portion of a bit old official motherboard schematic where the fan interface depicted clearly.

The fan operating voltage shall be within the range 12 V ±1.2 V. Look, the tachometer signal has a pull-up to 12V (maximum 13.2V), and  there’s a potential divider which finally sets a logic-level output (high) voltage approximately 2.8V. Likewise, the PWM input of the fan has a pull-up to motherboard logic (5V or 3.3V? – don’t know). According to Intel, if no control signal is present the fan shall operate at maximum RPM.

It’s worth noting that many motherboards provide insufficient options to customize automatic fan control or simply don’t go below certain duty cycles and speed levels. Using add-on controllers such as the Noctua NA-FC1 (see below) can lower the motherboard’s PWM control curve in order to achieve truly silent operation, even with high-speed PWM fans.

It’s time to go forth

Our prime focus at present is on using easily available components to build a 4-wire fan controller to automatically and/or manually limit or boost the rotation speed of the cooling fan in a power electronics project, rather than making an adapter circuitry to pick off the existing motherboard PWM signals. We can dismiss the Intel specifications from our minds for a little while.

It may be preferable to have more thoughts on integrating a temperature sensor and setting a temperature threshold to have the fan ramp up as needed to keep the system cool. That means an efficient go up to reduce the noise produced by the fan without affecting performance when called for. Here, I found a neat circuit idea for controlling a temperature driven fan with PWM. The circuit can be adapted for 4-wire fans but it has some serious limits as pointed out by the designer www.paulvdiyblogs.net/2015/08/controlling-temperature-driven-fan-with.html

So, how to regulate the fan speed?

As you might noticed, I have the 4-wire fan (E97379-003) which is supplied with boxed Intel processors.

Here’s the initial run report of my 4-wire fan tested without the control (PWM) signal input.

• DC Input: 12V/2A (regulated)
• Current Consumption: 300mA
• Tacho Signal (Pulled up to 12V by 4.7K Resistor): 160Hz (50% duty cycle)
• Tacho Voltage: 6VDC
• PWM Signal Input: Floating (no external connection)

Nevertheless, the tachometer signal output denotes a ‘free-run’ RPM of 4800 (160Hz x 60 Sec /2) which is very close to the most discussed free-run RPM rate of 5000!

Just a note– There are many different types of fans and ways of controlling them.  A 2-wire fan is controlled by adjusting either the dc voltage or pulse width in low-frequency PWM. A 3-wire fan can be controlled using the same kind of drive as for 2-wire fans – variable dc or low-frequency PWM. The difference between 2-wire fans and 3-wire fans is the availability of feedback from the fan for closed-loop speed control. The tach signal indicates whether the fan is running and its rate of speed.

The 4-wire fans have a PWM input, which is used to control the speed of the fan. Instead of switching the power to the entire fan on and off, only the power to the drive coils is switched. The PWM signal drives the fan directly; the drive FET is integrated inside the fan. Since the PWM drive signal is applied directly to the coils of the fan, the fan’s electronics are always powered on, and the tach signal is always available. This eliminates the need for pulse stretching which’s ineluctable in 3-wire fan PWMs (Excerpts from Analog Dialogue 38-02, February 2004).

I missed that joy!

The first mishap came when I simply tried to regulate the fan speed with the help of my trusty variable frequency-variable duty cycle 5V PWM module (Tuned to 25kHz/0-100%). My fan model wasn’t happy to be PWM-controlled i.e. there’s no significant change observed in the rotational speed of the fan even though the duty cycle swept across its scale at every point. I doubt that the fan I picked is designed to accept a different control/PWM signal. And in the specifications sheet, we can see that the fan speed is indicated as 1200-2500 RPM. I’m not sure why it is not stated as 0-5000 RPM? If you can pinpoint the reasons behind that, please shed some light through the comments section.

Here are a few random notes I’ve made after seeing several online discussions on today’s 4-wire PWM fan control thoughts (you’re never too old to learn):

• 4-wire fans with control input provide the ability to regulate the RPM of the fan without changing the input voltage delivered to the fan it. A variable RPM allows the cooling rate to be adjusted to meet demand, silencing the fan and saving energy when full speed is not required
• The standard control signal is a square wave operating at 25 kHz, with the duty cycle determining the fan speed. Typically a fan can be driven between about 30% and 100% of the rated fan speed, using a signal with up to 100% duty cycle. The exact speed behavior (linear, off until a threshold value, or a minimum speed until a threshold) at low control levels is manufacturer dependent
• In certain new builds the control input has an internal TTL-level pull-up resistor tied to 5V/3.3V.
• In some fans, the PWM signal drives the fan directly because the drive FET is integrated inside the fan
• In some fans the speed control input accepts a voltage signal instead the PWM signal
• Certain fans calls for a ‘reverse’ PWM , or inverted PWM control signal input Similarly, some fans looks for a Push-Pull PWM output driver instead of the standard Open Drain PWM output driver
• Certain fans demands 12V PWM signal instead of the 5V PWM
• Certain fans fan has a custom firmware that accepts something other than the standard 25kHz PWM
• Delta Electronics formulated an advanced PWM circuit control utilizing soft switching, resulting in a smooth current signal to their Delta 4-wire fans. This thermal technology further reduces vibration and noise while continuing to provide better performance, high reliability, and cost-effective thermal cooling solutions
• Intel’s E97379-001 fan is equipped with automatic speed correction circuitry and adjust its speed to the maximum regardless of the input voltage. This renders the majority of basic fan controlled useless as their speed can only be controlled via a PWM signal from a motherboard.
• Thermal speed controlled fans with a built-in thermistor are designed that the fan itself contains a thermistor. The temperature of the air flowing through the fan motor is detected and the fan’s speed changes automatically according to changes in that temperature. In some fans with Thermistor & PWM control, like the FHS-A9025S20 (Delta), the fan will go to temperature control speed if the control signal is disconnected

At the time I couldn’t decide if they were speaking from prior knowledge, or just making an educated guess. But the sure thing is that my 4-wire fan probably doesn’t work with the old school trick. Not sure yet whether I want to go to the trouble to design a ‘composite’ controller circuit for my fan yet.

A total failure!

But I don’t quit easily, so after a little Googling I ordered one cheap 4-wire fan controller board/module from Banggood. The controller board/module chosen has a temperature sensor and plenty of other utile features (https://www.banggood.in/DC-12V-Four-Wire-Thermostat-PWM-Fan-Speed-Controller-Module-p-997793.html). Next week’s article will describe how I took apart this pretty nice module. It’s not meant to replicate the exact electronics, but maybe understand some basic principles to build your own universal fan controller.

So that’s the plan – test my fan with the prebuilt 4-wire fan controller board/module. At first, I trialed the fresh fan controller with an old 4-wire PC cooling fan lying around and it worked like a charm!

The next attempt was to drive my new 4-wire fan (E97379-003) through the controller module but alas the fan runs (again) at a steadfast speed around 5000 RPM, and not responded to the standard speed control stimulus. This is the approximate RPM readout (sorry for the blurry grey snap). And the above scope capture was taken while the speed controller’s PWM output is at its minimum (20% ) but RPM of the disobedient fan is at its maximum (5000)!

I couldn’t figure out how to understand its cryptic speed regulator scheme. Bogged down!

The most scandalous cheat of its time!

Since I was highly thwarted,  I decided to take a breakdown of the fan in hand without due thought just to clear my suspicion. Yes, it’s right – the fan is a dirty cheap Chinese knock off without a pulse width modulation control circuitry. The PWM input wire (blue) has no role in the play as it’s merely soldered to a vacant copper track of the circuit board (see below).

Furthermore, you can see that the fan driver chip is a 4-pin chip CR211 which’s a very cheap 2-phase dc motor driver for 3-wire BLDC fans. How is that?

https://p.globalsources.com/IMAGES/PDT/SPEC/398/K1139513398.pdf.

A temporary cessation

I’m speechless!

Counterfeiting online is nothing new of course, especially when it comes to Chinese electronics. To unsuspecting consumers, counterfeit goods can appear legitimate as long as the seller’s description looks legit. Anyway, always keep an eye on commodities that are retailing at a much lower price than the wholesale price of legitimate genuine products, even if it’s sold through a trusty online storefront. Be vigilant!

I have one more article to cover about this topic. It’s another bit of electronics that made me curious! So stay tuned and thanks for reading!

At the bottom, an annotated photograph that points the pin out of a typical 4-wire Intel CPU cooling fan.

…