Built-in thermal sensor Brushless DC fan speed control Linear control and regulation of fan speed based on temperature Band automatic green shutdown mode operation

Features: Fan Speed Controller Computer and Server Power Supply Telecom Low Cost Ventilation

Application: low cost integrated fan speed control 10 mA) FSCTXXA-UH5 is a highly integrated low cost fan speed controller suitable for PC desktop, laptop and server power supplies and various devices requiring low cost ventilation systems. An internal thermal sensor attached to the lug is used to adjust the fan speed. A continuous analog voltage produces a temperature proportional to the label temperature at the output and linearly controls an external PNP transistor (fan transistor) in series with the brushless DC fan to vary its speed. The ON control terminal can be used to select two operating modes. If not connected (high impedance) by default, the ON mode is selected. In ON mode, the fan speed is adjusted according to the tab temperature of the FSCT by adjusting the VBE voltage of the PNP fan transistor. At low temperature, the fan transistor voltage is limited to about 5.5 V (for 12V fans) to ensure the minimum voltage across the fan is above its stall voltage and keeps running (see pic). This regulation is performed with fan supply voltage rejection high. Mode shutdown allows operation in green mode. As long as When the temperature remains the same, the fan stops running below the ton threshold. When the temperature rises above the ton, the fan is forced to turn on, and the speed is adjusted according to the same temperature as the mode on. When the temperature is below TOFF, the fan automatically turns off. The magnetic The size of the hysteresis is large enough to reduce on/off ringing during green mode operation.

Detailed description and application information: 1. Output characteristics and temperature 1.1 Fan speed control It is well known that brushless DC motors have quasi-linear speed-voltage characteristics. Therefore, the fan speed varies linearly with the voltage applied to its terminals. The fsct integrated circuit provides a temperature-dependent voltage at its output that is sensed by its tag. This output voltage decreases with increasing temperature, following two modes of operation. As shown in Figure 1, by connecting this output terminal to the base of the PNP transistor, the fan voltage will increase as the sensed temperature increases. The fan voltage is given by the equation: VV VV fan EB output = negative - + (1) Typically, for 12 V v fan + and 1 V veb voltage this gives: V V fan V output V () = -11 (2 ) can be noticed that the pnp works as a linear amplifier. This avoids electromagnetic interference and noise compared to the PWM control circuit. 1.2 The on/off mode selection diagram gives the algorithmic flowchart of the fsct behavior. First, the two operating modes are distinguished by pin signals. On mode: active when ON pin is high or not connected (due to internal pull-up current source "yes") mode off: active when ON pin is low. During mode turn-on, the output voltage follows a three-part characteristic according to its junction temperature (Tj): tj tol: In this case, vout is constrained to its minimum value (vol). During mode off, hysteretic control allows the system to switch to on mode only when the temperature exceeds the ton value. The fan spins back up if the system cools down enough that the temperature is below Fahrenheit. The figure summarizes the fsct output characteristic versus temperature pattern for the two operations. 1.3 Minimum Speed (Mode On) The On mode allows the user to ensure that the fan will always be on regardless of the ambient temperature. For low temperatures (usually below tms=37°C), the output voltage is clamped to voms. VOMS is set to a maximum of 5.5 V, so the minimum voltage applied to the fan is 5.5 V (according to Equation 2). This voltage is higher than the stall value of most fans and will then ensure that the control fan will always run when the mode is on, avoiding due to low temperatures. The annoying noise of the fan and repeated starts are also suppressed.

2. Hysteresis control (mode off) FSCT can be turned off by ON signal. When this signal is low, the output pin is high, i.e. the fan is turned off. This mode can save energy wasted by the fan when the output power is very low. For 12V supply voltage, turning off the fan saves 0.5 to 2.5 watts of power (for a 200mA 12V supply DC motor). During the mode off, the fsct does not lose temperature control; in fact, the fan turns on automatically if the temperature gets too high. This safety feature protects the power supply or semiconductors from accidental overtemperature of the device. To maintain energy saving benefits, the FSCT turns off the fans when the temperature is below the TOFF threshold. In practice, for constant heating power (e.g. power supply using FSCT (see picture): Case 1: Heating power is too low to keep TJ below t. Fan remains off. Case 2: Heating power is high enough to raise TJ liters High to over t. However, since the energy is low, TJ drops down to Fahrenheit and the temperature starts to rise again, to t. This will cause the fan to cycle on/off periodically. Case 3: Heating power is higher than case 2, Therefore TJ remains above TOFF. The fan does not turn off in this case unless the heating power is reduced. For example, with a 200 W computer power supply, using a FSCT17 device on a power semiconductor heat sink, an output power consumption of 25 W can be achieved (TJ will stabilize at about 60°C, i.e. below standard t). Case 2 can reach a power dissipation of 50 watts. Then, due to the large hysteresis value (30°C on/off cycle (see TP in the graph), it lasts about 15 minutes. This is long enough to avoid excessive fan startup cycles per hour. Case 3 can reach power consumption of 75 watts or more. For 75W, the power supply ambient temperature stabilizes at around 42°C. It should be noted that for case 3, due to The output voltage follows the same linear law as the ON mode. Then, the hysteresis control is more efficient than the simple on/off control mode. The mode OFF mode turns on 1 volt temperature (degree Celsius) Joule volume tolton time of flight management system

Figure: Temperature change with mode off. 3. Internal Temperature Sensor 3.1 Temperature Sensor Linear Response The fsct device has an internal temperature sensor. This sensor is directly determined by the properties of silicon. It is actually a voltage reference proportional to absolute temperature as it is a picture of the silicon thermal voltage "vt" (refer to the equation below).

The Boltzmann constant absolute temperature K is compared with vbe, and the positive temperature coefficient of this sensing method is +2mv/°C. Sensing method (-2mv/°C, sometimes used for thermal protection) due to its better accuracy (low effect of process dispersion). This signal is then processed to provide the desired output voltage range. The internal sensor allows the user to not use a negative temperature coefficient thermistor (NTC). Thus, the user gets rid of the Joule effect, which interferes with the temperature measurement due to the NTC bias current. Furthermore, the fsct response is linear with temperature. This simplifies thermal studies and heatsink ratings for power components or microprocessors. NTC thermistor users also need to add a fixed resistor in order to get a linear thermal response from it a bit like a sensor. Linear behavior is also only guaranteed over a limited temperature range. 3.2 IPPAK Mounting Notes First, it should be noted that the label is directly connected to the GND pin; then, when the FSCT is glued to the heatsink. If this heat sink is at a different voltage than ground, you must add an electrical insulator between the tab and the heat sink. There are also many benefits to using a non-isolated through-hole package like ippak compared to ntc bulbs. In fact, NTC does not provide such a flat area as IPPAK. The user needs to add some glue to make sure to increase the thermal resistance and response.

Two components should be used to improve the heat exchange between the FSCT die and the heat sink, the temperature must be monitored. These components include: thermal interface pads to reduce the effect of voids on thermal impedance and to ensure electrical isolation (if needed) a clip to push the IPPAK onto the heatsink and then reduce the interface thermal impedance. Depending on the heatsink type, several clips are available: saddle clips for elongated heatsinks (see picture); clevis clips for thick heatsinks (see picture) special-shaped heatsink clips.

3.3 Temperature measurement error First of all, the time constant between the temperature changes on the outside of the ippak copper chip, the silicon chip is in the range of several hundred milliseconds. Since the temperature phenomenon is very slow for the target application (the temperature of the mosfet heat sink typically increases with the rate of 1°C per second (in the power supply), the fsct is able to react immediately to overheating events. In addition, the extremely low junction-to-case thermal resistance ( 3°c/w) to minimize temperature measurement errors. In the calculations below, we consider both the thermal resistance of the package and the heatsink tab interface (Fig. Graph of the interface: IPPAK heatsink interface. There are several One company offers adhesives and release materials for the interface between the electronic device and the heat sink. These interfaces are available with dedicated label footprints. For ippak packs, users can choose from SOT-32, SOT-82, or even to -126 or to -220 packs).

These interfaces provide very low thermal resistance. For example, the Sil Pad® 800 series Bergquist designed for low cost and low installation pressure exhibits a typical thermal impedance of 0.45°C.in/w. At 10 psi below normal atmospheric pressure (1 atm=1013 hpa=1013 x 100 x 14.5 x 10-3=14.7 psi), this impedance can increase to 0.92°C.in/w. Additionally, the mounting clips will apply 15 to 50 N of force. This will result in a pressure of 25 to 200 psi. In this case, the thermal impedance varies from 0.6 to 0.29°C.in/w for the Sil Pad® 800 series. For example, let's assume fsct and heatsink. We only consider the surface area for heat exchange. This surface is typically equal to: S = 4.7 x 5.1 mm² = 0.037 sq inch This creates an additional resistance: rthc - h = 1/0.037 = 27°c/w Then, the maximum power dissipated in the FSCT, is used for maximum The output power, given by the equation:

4. Temperature/voltage slope change The output voltage-temperature characteristic of fsct is designed to be suitable for most applications in the PC power supply field. The user's advantage is in realizing intelligent temperature regulation. However, some applications require dedicated temperature regulation features. The figure provides an example of a solution that allows changing the PNP base voltage and temperature. This schematic requires only a single voltage amplifier (DIL8 package example) and less than 10 complementary resistors. This schematic maintains the applied constant minimum voltage (VOMS) below the transmission temperature. In fact, the u1a op amp subtracts 5.1V from VOUT (thanks to the D1 zener, see VREF). This means that the new base-to-ground voltage (VNEW) remains the same only if you are at a voltage higher than 5.1 V (VOM is in mode on or VCC is in off mode) (see diagram). So, if you want to increase the voltage to temperature ratio exactly, i.e. Vout will drop faster as TJ increases, the diagram should implement 13 schematics. Indeed, the voltage at opamp U1b as a follower) output is: