AND9093 - Using MOSFETs in Load Switch …

1 Semiconductor Components Industries, LLC, 2014 February, 2014 Rev. 11 Publication Order Number: AND9093 /DAND9093/DUsing MOSFETs in LoadSwitch ApplicationsIntroductionIn today s market, power management is more importantthan ever. Portable systems strive to extend battery life whilemeeting an ever increasing demand for higher switches provide a simple and inexpensive method forthe system to make the appropriate power managementdecisions based on which peripherals or sub-circuits arecurrently in use. load switches are found in notebooks, cellphones, hand held gaming systems and many other load Switch is controlled by the system, and connectsor disconnects a voltage rail to a specific load . By turningunused circuitry off, the system as a whole can run moreefficiently.

2 The load Switch provides a simple means topower a load when it is in demand and allows the system tomaximize Switch BasicsA load Switch is comprised of two main elements: the passtransistor and the on/off control block, as shown in Figure 1. Example load Switch CircuitP-channelLoad SwitchOn/OffControlVINLoad+ The pass transistor is most commonly a MOSFET (eitherN-channel or P-channel) that passes the voltage supply toa specified load when the transistor is and P-channel ConsiderationsThe selection of a P-channel or N-channel load switchdepends on the specific needs of the application. TheN-channel MOSFET has several advantages over theP-channel MOSFET. For example, the N-channel majoritycarriers (electrons) have a higher mobility than theP-channel majority carriers (holes).

3 Because of this, theN-channel transistor has lower RDS(on) and gate capacitancefor the same die area. Thus, for high current applications theN-channel transistor is Using an N-channel MOSFET in a load switchcircuit, the drain is connected directly to the input voltagerail and the source is connected to the load . The outputvoltage is defined as the voltage across the load , andtherefore:VS+VOUT(eq. 1)In order for the N-channel MOSFET to turn on, thegate-to-source voltage must be greater than the thresholdvoltage of the device. This means that:VGwVOUT)Vth(eq. 2)In order to meet Equation 2, a second voltage rail isneeded to control the gate. Therefore, the input voltage railcan be considered independently of the pass of this, the N-channel load Switch can be used forvery low input voltage rails or for higher voltage rails, aslong as the gate-to-source voltage VGS remains higher thanthe threshold voltage of the device.

4 The designer mustensure that the device maximum ratings and the safeoperating area of the MOSFET are not Using a P-channel MOSFET in a load Switch circuit(as in Figure 1, the source is directly connected to the inputvoltage rail and the drain is connected to the load . In orderfor the P-channel load Switch to turn on, the source-to-gatevoltage must be greater than the threshold :VINwVG)Vth(eq. 3)At minimum, the input voltage rail must be greater thanthe threshold voltage of the selected pass transistor(assuming the gate voltage is 0 V when the load Switch isturned on).The P-channel MOSFET has a distinct advantage over theN-channel MOSFET, and that is in the simplicity of theon/off control block. The N-channel load Switch requires anadditional voltage rail for the gate; the P-channel load switchdoes not.

5 As with the N-channel MOSFET, the designermust ensure that the device maximum ratings and the safeoperating area of the P-channel MOSFET are not NOTEAND9093/ Switch Control Circuit ConsiderationsThere are multiple ways to implement the on/off controlblock in a load Switch circuit. This section will cover onecontrol circuit example for the N-channel and one for theP-channel load 2. N-channel Example Control CircuitN-channelLoad SwitchVOUTVINLoadQ1R1 ENVGATE+ + Figure 2 shows an example load Switch control circuit foran N-channel pass transistor. A logic signal from the systempower management control circuitry turns the load switchon and off via a small-signal NMOS transistor, Q1. WhenEN is LOW, Q1 is off and the pass transistor gate is pulledup to VGATE to keep it turned on.

6 When EN is HIGH, Q1turns on, the pass transistor gate is pulled to ground, and theload Switch turns off. Resistor R1 is selected so thatmilliamps of current or less flow through R1 when Q1 is standard range is 1 kW 10 additional voltage source, VGATE, is needed to keepthe gate-to-source forward biased. As expressed inEquation 2, the gate voltage must be larger than the sum ofthe output voltage and the threshold voltage. This may beundesirable for systems that do not have an extra voltage 3. P-channel Example Control CircuitP-channelLoad SwitchVOUTVINLoadQ1R1EN+ Figure 3 shows an example load Switch control circuit fora P-channel pass transistor. As with the N-channel example,a logic signal from the system power management controlcircuitry turns the load Switch on and off via a small-signalNMOS transistor, Q1.

7 When EN is LOW, Q1 is off and thegate is pulled up to VIN. When EN is HIGH, Q1 turns on, thepass transistor gate is pulled to ground, and the load switchturns on. As long as the input voltage rail is higher than thethreshold voltage of the PMOS transistor, it will turn onwhen EN is HIGH without the need of an additional voltagesource. As with the N-channel control circuit, resistor R1 isselected so that milliamps of current or less flow through R1when Q1 is on. A standard range is 1 kW 10 both control circuit implementations, the small-signalNMOS transistor, Q1, can be integrated into the samepackage as the pass ConsiderationsEfficiency is critical to the success of the overall powermanagement of the system. In a load Switch circuit, the loadcurrent flows directly through the pass transistor when it isturned on.

8 Therefore, the main power loss is the 4)The RDS(ON) of the pass transistor causes a voltage dropbetween the input voltage and the output voltage, as shownin Equation 5. For applications requiring high load currentsor low voltage rails, this voltage drop becomes critical. Thevoltage drop will increase as the load current increases, andthe voltage drop at maximum load must be taken intoconsideration when selecting the pass 5)As discussed in previous sections, the N-channelMOSFET has an RDS(on) advantage over the P-channelMOSFET for a given die size. The RDS(on) of an N-channeldevice can be two times lower than the RDS(on) ofa P-channel device of similar die area. This difference ismost prominent at higher currents, but the N-channelRDS(on) advantage becomes less prominent at lowercurrents.

9 For applications such as cell phones and otherportable low power devices, higher efficiency can beattained Using a P-channel pass transistor, with theadvantage of a simpler control illustrate this, let s assume that a 30 mW N-channeltransistor and a 50 mW P-channel transistor have similar diesize. The efficiency impact of the two devices will beexamined for a high current application and a low the first example, consider an application that requiresa maximum load current of 10 A. Using Equations 4 and 5,the power loss at the maximum load is calculated to be 3 Wfor the N-channel transistor, and the voltage drop across thetransistor is 300 mV. The power loss at the maximum loadis 5 W for the P-channel transistor, and the voltage dropacross the transistor is 500 consider an application in which the maximumcurrent is 2 A.

10 The power loss at maximum load is 120 mWfor the N-channel device and 200 mW for the p-channeldevice. The voltage drop for the N-channel transistor is60 mV and is 100 mV for the P-channel a final example, consider an application with an850 mA maximum load current. The 30 mW N-channeltransistor s power loss is mW compared to mW power loss of the 50 mW P-channel transistor ofsimilar die size. For low current applications, the N-channelAND9093/ (ON) advantage becomes negligible. P-channel passtransistors can be designed to have RDS(on) as low as 8 RDS(on) is critical for maximizing the efficiency of theload Switch circuit and minimizing the voltage drop acrossthe pass transistor. The specific conditions of the load switchapplication must be considered to make the final decision touse a PMOS or NMOS pass Voltage ConsiderationsThe applied gate-to-source voltage of the pass transistordirectly affects the efficiency of the circuit because RDS(on)is inversely proportional to the applied gate-to-sourcevoltage.