LM4781 Overture™ Audio Power Amplifier Series 3 Channel 35W Audio Power Amplifier with Mute General Description Key Specifications The LM4781 is a three channel audio amplifier capable of typically delivering 35W per channel of continuous average output power into an 8Ω load with less than 0.5% THD+N from 20Hz - 20kHz. The LM4781 is fully protected utilizing National’s Self Peak Instantaneous Temperature (˚Ke) (SPiKeTM) protection circuitry. SPiKe provides a dynamically optimized Safe Operating Area (SOA). SPiKe protection completely safeguards the LM4781’s outputs against over-voltage, under-voltage, overloads, shorts to the supplies or GND, thermal runaway and instantaneous temperature peaks. The advanced protection features of the LM4781 places it in a class above discrete and hybrid amplifiers. Each amplifier of the LM4781 has an independent smooth transition fade-in/out mute. The LM4781 can easily be configured for bridge or parallel operation for higher power and bi-amp solutions. j Output Power/Channel with 0.5% THD+N, 1kHz into 8Ω 35W (typ) j THD+N at 3 x 20W into 8Ω (20Hz - 20kHz) 0.2% (typ) j THD+N at 3 x 20W into 6Ω (20Hz - 20kHz) 0.3% (typ) j THD+N at 3 x 20W into 4Ω (20Hz - 20kHz) 0.5% (typ) j Mute Attenuation 110dB (typ) j PSRR 85dB (min) j Slew Rate 9V/µs (typ) Features n n n n n SPiKe Protection Low external component count Quiet fade-in/out mute mode Wide supply range: 20V - 70V Signal-to-Noise Ratio ≥ 93dB (ref. to PO = 1W) Applications n n n n Audio Audio Audio Audio amplifier amplifier amplifier amplifier for for for for component stereo compact stereo self-powered speakers high-end and HD TVs Typical Application 200673B5 FIGURE 1. Typical Audio Amplifier Application Circuit SPiKe™ Protection and Overture™ are trademarks of National Semiconductor Corporation. © 2003 National Semiconductor Corporation DS200673 www.national.com LM4781 Overture™ Audio Power Amplifier Series 3 Channel 35W Audio Power Amplifier with Mute August 2003 LM4781 Connection Diagrams Plastic Package (Note 14) 200673B2 Top View Order Number LM4781TA See NS Package Number TA27A TO-220 Top Marking 200673A6 Top View U - Wafer Fab Code Z - Assembly Plant Code XY - Date Code TT - Die Run Traceability L4781TA - LM4781TA www.national.com 2 Soldering Information (Notes 1, TA Package (10 seconds) 2) + -40˚C to +150˚C Thermal Resistance - Supply Voltage |V | + |V | (No Signal) 260˚C Storage Temperature If Military/Aerospace specified devices are required, please contact the National Semiconductor Sales Office/ Distributors for availability and specifications. θJA 30˚C/W θJC 0.9˚C/W 72V Supply Voltage |V+| + |V-| Operating Ratings (Notes 1, 2) (Input Signal) 70V Temperature Range (V+ or V-) and Common Mode Input Voltage TMIN ≤ TA ≤ TMAX |V+| + |V-| ≤ 60V Differential Input Voltage (Note 13) Output Current −20˚C ≤ TA ≤ +85˚C 20V ≤ VTOTAL ≤ 70V Supply Voltage |V+| + |V-| 60V Internally Limited Power Dissipation (Note 3) 125W ESD Susceptability (Note 4) 2.0kV ESD Susceptability (Note 5) Note: Operation is guaranteed up to 70V; however, distortion may be introduced from SPiKe protection circuitry if proper thermal considerations are not taken into account. Refer to the Thermal Considerations section for more information. 200V Junction Temperature (TJMAX) (Note 9) 150˚C Electrical Characteristics (Notes 1, 2) The following specifications apply for V+ = +28V, V- = −28V, IMUTE = -1mA/channel and RL = 8Ω unless otherwise specified. Limits apply for TA = 25˚C. Symbol Parameter Conditions LM4781 Typical Limit Units (Limits) (Note 6) (Notes 7, 8) |V | + |V | Power Supply Voltage (Note 10) GND − V ≥ 9V 18 20 70 V (min) V (max) AM Mute Attenuation IMUTE = 0mA 110 80 dB (min) Output Power (RMS) THD+N = 0.5% (max) f = 1kHz |V+| = |V-| = 20V, RL = 4Ω |V+| = |V-| = 24V, RL = 6Ω |V+| = |V-| = 28V, RL = 8Ω 25 30 35 20 25 30 W (min) W (min) W (min) THD+N Total Harmonic Distortion + Noise PO = 20W, f = 20Hz - 20kHz AV = 26dB |V+| = |V-| = 20V, RL = 4Ω |V+| = |V-| = 24V, RL = 6Ω |V+| = |V-| = 28V, RL = 8Ω 0.5 0.3 0.2 Xtalk Channel Separation (Note 11) SR + PO IDD - - % % % PO = 10W, f = 1kHz 70 dB PO = 10W, f = 10kHz 66 dB Slew Rate VIN = 1.2VRMS, f = 10kHz square Wave.RL = 2kΩ 9 5 V/µs (min) 75 150 mA (max) Total Quiescent Power VCM = 0V Supply Current VO = 0V, IO = 0A VOS Input Offset Voltage VCM = 0V, IO = 0mA 1 10 mV (max) IB Input Bias Current VCM = 0V, IO = 0mA 0.2 1 µA (max) + - IO Output Current Limit |V | = |V | = 20V, tON = 10ms 4 3 A (min) VOD Output Dropout Voltage (Note 12) |V+ - VO|, V+ = 20V, IO = +100mA |VO - V-|, V- = -20V, IO = -100mA 1.5 2.5 4 4 V (max) V (max) V+ = 30V to 10V, V- = -30V VCM = 0V, IO = 0mA 125 85 dB (min) V+ = 30V, V- = -30V to -10V VCM = 0V, IO = 0mA 110 85 dB (min) PSRR Power Supply Rejection Ratio (Note 15) 3 www.national.com LM4781 Absolute Maximum Ratings LM4781 Electrical Characteristics (Notes 1, 2) (Continued) The following specifications apply for V+ = +28V, V- = −28V, IMUTE = -1mA/channel and RL = 8Ω unless otherwise specified. Limits apply for TA = 25˚C. Symbol Parameter Conditions LM4781 Typical Limit Units (Limits) (Note 6) (Notes 7, 8) CMRR Common Mode Rejection Ratio V+ = 30V to 10V, V- = -10V to -30V VCM = 20V to -20V, IO = 0mA (Note 15) AVOL Open Loop Voltage Gain |V+| = |V-| = 28V, RL = 2kΩ ∆VO = 20V GBWP Gain Bandwidth Product |V+| = |V-| = 28V, fIN = 100kHz VIN = 50mVRMS eIN Input Noise IHF-A-Weighting Filter, RIN = 600Ω (Input Referred) SNR IMD Signal-to-Noise Ratio Intermodulation Distortion 110 75 dB (min) 115 80 dB (min) 8 2 MHz (min) 3.0 8 µV (max) PO = 1WRMS; A-Weighted Filter fIN = 1kHz, RS = 25Ω 93 dB PO = 25WRMS; A-Weighted Filter fIN = 1kHz, RS = 25Ω 107 dB 0.004 0.006 % % 60Hz, 7kHz, 4:1 (SMPTE) 60Hz, 7kHz, 1:1 (SMPTE) Note 1: All voltages are measured with respect to the ground pins, unless otherwise specified. Note 2: Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. Operating Ratings indicate conditions for which the device is functional, but do not guarantee specific performance limits. Electrical Characteristics state DC and AC electrical specifications under particular test conditions which guarantee specific performance limits. This assumes that the device is within the Operating Ratings. Specifications are not guaranteed for parameters where no limit is given; however, the typical value is a good indication of device performance. Note 3: The maximum power dissipation must be de-rated at elevated temperatures and is dictated by TJMAX, θJC, and the ambient temperature TA. The maximum allowable power dissipation is PDMAX = (TJMAX -TA)/θJC or the number given in the Absolute Maximum Ratings, whichever is lower. For the LM4781, TJMAX = 150˚C and the typical θJC is 0.9˚C/W. Refer to the DETERMINING THE CORRECT HEAT SINK section for more information. Note 4: Human body model, 100pF discharged through a 1.5kΩ resistor. Note 5: Machine Model: a 220pF - 240pF discharged through all pins. Note 6: Typical specifications are measured at 25˚C and represent the parametric norm. Note 7: Tested limits are guaranteed to National’s AOQL (Average Outgoing Quality Level). Note 8: Datasheet min/max specification limits are guaranteed by design, test, or statistical analysis. Note 9: The maximum operating junction temperature is 150˚C. However, the instantaneous Safe Operating Area temperature is 250˚C. Note 10: V- must have at least - 9V at its pin with reference to GND in order for the under-voltage protection circuitry to be disabled. In addition, the voltage differential between V+ and V- must be greater than 14V. Note 11: Cross talk performance was measured using the demo board shown in the datasheet. PCB layout will affect cross talk. It is recommended that the input and output traces be separated by as much distance as possible. Return ground traces from outputs should also be independent back to sinlge ground point and use as wide of traces as possible. Note 12: The output dropout voltage is the supply voltage minus the clipping voltage. Refer to the Clipping Voltage vs. Supply Voltage graph in the Typical Performance Characteristics section. Note 13: The Differential Input Voltage Absolute Maximum Rating is based on supply voltages V+ = 30V and V- = - 30V. Note 14: The TA27A is a non-isolated package. The package’s metal back and any heat sink to which it is mounted are connected to the V- potential when using only thermal compound. If a mica washer is used in addition to thermal compound, θCS (case to sink) is increased, but the heat sink will be electrically isolated from V-. Note 15: DC electrical test. www.national.com 4 LM4781 Bridged Amplifier Application Circuit 200673B4 FIGURE 2. Bridged Amplifier Application Circuit 5 www.national.com LM4781 Parallel Amplifier Application Circuit 200673B3 FIGURE 3. Parallel Amplifier Application Circuit www.national.com 6 LM4781 Single Supply Application Circuit 20067306 FIGURE 4. Single Supply Amplifier Application Circuit Note: *Optional components dependent upon specific design requirements. Auxiliary Amplifier Application Circuit 200673B6 FIGURE 5. Special Audio Amplifier Application Circuit 7 www.national.com LM4781 External Components Description (Figures 1-5) Components Functional Description 1 RB Prevents current from entering the amplifier’s non-inverting input. This current may pass through to the load during system power down, because of the amplifier’s low input impedance when the undervoltage circuitry is off. This phenomenon occurs when the V+ and V- supply voltages are below 1.5V. 2 Ri Inverting input resistance. Along with Rf, sets AC gain. 3 Rf Feedback resistance. Along with Ri, sets AC gain. 4 Rf2 (Note 16) Feedback resistance. Works with Cf and Rf creating a lowpass filter that lowers AC gain at high frequencies. The -3dB point of the pole occurs when: (Rf - Ri)/2 = Rf // [1/(2πfcCf) + Rf2] for the Non-Inverting configuration shown in Figure 5. 5 Cf (Note 16) Compensation capacitor. Works with Rf and Rf2 to reduce AC gain at higher frequencies. 6 CC (Note 16) Compensation capacitor. Reduces the gain at higher frequencies to avoid quasi-saturation oscillations of the output transistor. Also suppresses external electromagnetic switching noise created from fluorescent lamps. 7 Ci (Note 16) Feedback capacitor which ensures unity gain at DC. Along with Ri also creates a highpass filter at fc = 1/(2πRiCi). 8 CS Provides power supply filtering and bypassing. Refer to the Supply Bypassing application section for proper placement and selection of bypass capacitors. 9 RV (Note 16) Acts as a volume control by setting the input voltage level. 10 RIN (Note 16) Sets the amplifier’s input terminals DC bias point when CIN is present in the circuit. Also works with CIN to create a highpass filter at fC = 1/(2πRINCIN). If the value of RIN is too large, oscillations may be observed on the outputs when the inputs are floating. Recommended values are 10kΩ to 47kΩ. Refer to Figure 5. 11 CIN (Note 16) Input capacitor. Prevents the input signal’s DC offsets from being passed onto the amplifier’s inputs. 12 RSN (Note 16) Works with CSN to stabilize the output stage by creating a pole that reduces high frequency instabilities. 13 CSN (Note 16) Works with RSN to stabilize the output stage by creating a pole that reduces high frequency instabilities. The pole is set at fC = 1/(2πRSNCSN). Refer to Figure 5. 14 L (Note 16) 15 R (Note 16) Provides high impedance at high frequencies so that R may decouple a highly capacitive load and reduce the Q of the series resonant circuit. Also provides a low impedance at low frequencies to short out R and pass audio signals to the load. Refer to Figure 5. 16 RA Provides DC voltage biasing for the transistor Q1 in single supply operation. 17 CA Provides bias filtering for single supply operation. 18 RINP (Note 16) Limits the voltage difference between the amplifier’s inputs for single supply operation. Refer to the Clicks and Pops application section for a more detailed explanation of the function of RINP. 19 RBI Provides input bias current for single supply operation. Refer to the Clicks and Pops application section for a more detailed explanation of the function of RBI. 20 RE Establishes a fixed DC current for the transistor Q1 in single supply operation. This resistor stabilizes the half-supply point along with CA. 21 RM Mute resistance set up to allow 0.5mA to be drawn from each MUTE pin to turn the muting function off. → RM is calculated using: RM ≤ (|VEE| − 2.6V)/l where l ≥ 0.5mA. Refer to the Mute Attenuation vs Mute Current curves in the Typical Performance Characteristics section. 22 CM Mute capacitance set up to create a large time constant for turn-on and turn-off muting. 23 S1 Mute switch. When open or switched to GND, the amplifier will be in mute mode. 24 ROUT Reduces current flow between outputs that are caused by Gain or DC offset differences between the amplifiers. Note 16: Optional components dependent upon specific design requirements. www.national.com 8 Although the optional external components have specific desired functions that are designed to reduce the bandwidth and eliminate unwanted high frequency oscillations they may cause certain undesirable effects when they interact. Interaction may occur for components whose reactances are in close proximity to one another. One example would be the Typical Performance Characteristics THD+N vs Frequency THD+N vs Frequency ± 20V, POUT = 1W & 20W/Channel ± 24V, POUT = 1W & 20W/Channel RL = 4Ω, 80kHz BW RL = 6Ω, 80kHz BW 20067325 20067343 THD+N vs Output Power/Channel ± 20V, RL = 4Ω, 80kHz BW THD+N vs Frequency ± 28V, POUT = 1W & 20W/Channel RL = 8Ω, 80kHz BW 20067326 20067344 9 www.national.com LM4781 coupling capacitor, CC, and the compensation capacitor, Cf. These two components act as low impedances to certain frequencies which will couple signals from the input to the output. Please take careful note of basic amplifier component functionality when designing in these components. The optional external components shown in Figure 4 and Figure 5 and described above are applicable in both single and split voltage supply configurations. Optional External Component Interaction LM4781 Typical Performance Characteristics (Continued) THD+N vs Output Power/Channel ± 24V, RL = 6Ω, 80kHz BW THD+N vs Output Power/Channel ± 28V, RL = 8Ω, 80kHz BW 20067327 20067336 Output Power/Channel vs Supply Voltage f = 1kHz, RL = 6Ω, 80kHz BW Output Power/Channel vs Supply Voltage f = 1kHz, RL = 4Ω, 80kHz BW 20067331 20067340 Total Power Dissipation vs Output Power/Channel 1% THD (max), RL = 4Ω, 80kHz BW Output Power/Channel vs Supply Voltage f = 1kHz, RL = 8Ω, 80kHz BW 20067323 www.national.com 20067330 10 LM4781 Typical Performance Characteristics (Continued) Total Power Dissipation vs Output Power/Channel 1% THD (max), RL = 8Ω, 80kHz BW Total Power Dissipation vs Output Power/Channel 1% THD (max), RL = 6Ω, 80kHz BW 20067339 20067322 Crosstalk vs Frequency ± 28V, POUT = 10W, RL = 8Ω, 80kHz BW Crosstalk vs Frequency ± 20V, POUT = 10W, RL = 4Ω, 80kHz BW 20067338 200673C0 Two Channels On Crosstalk vs Frequency ± 20V, POUT = 10W, RL = 4Ω, 80kHz BW Two Channels On Crosstalk vs Frequency ± 28V, POUT = 10W, RL = 8Ω, 80kHz BW 20067348 20067329 11 www.national.com LM4781 Typical Performance Characteristics (Continued) Supply Current vs Supply Voltage Mute Attenuation vs Mute Pin Current POUT = 10W/Channel 20067321 20067333 Power Supply Rejection Ratio Large Signal Response 200673C1 200673C2 Common-Mode Rejection Ratio Open Loop Frequency Response 200673C4 200673C3 www.national.com 12 LM4781 Typical Performance Characteristics (Continued) THD+N vs Frequency ± 20V, POUT = 1W & 25W, Clipping Voltage vs Supply Voltage Chs 1&3 in Bridge Mode (Note17), RL = 8Ω, 80kHz BW 200673C5 20067335 THD+N vs Frequency THD+N vs Frequency ± 28V, POUT = 1W & 35W, ± 24V, POUT = 1W & 45W, Chs 1&3 in Parallel Mode (Note18), RL = 4Ω, 80kHz BW All Chs in Parallel Mode (Note18), RL = 2Ω, 80kHz BW 20067334 20067342 THD+N vs Output Power THD+N vs Output Power ± 20V, Chs 1&3 in Bridge Mode (Note17), ± 28V, Chs 1&3 in Parallel Mode (Note18), RL = 8Ω, 80kHz BW RL = 4Ω, 80kHz BW 20067345 20067337 13 www.national.com LM4781 Typical Performance Characteristics (Continued) Output Power vs Supply Voltage Chs 1&3 in Bridge Mode (Note17), f = 1kHz, RL = 8Ω, 80kHz BW THD+N vs Output Power ± 24V, All Chs in Parallel Mode (Note18), RL = 2Ω, 80kHz BW 20067328 20067332 Output Power vs Supply Voltage All Chs in Parallel Mode (Note18), f = 1kHz, RL = 2Ω, 80kHz BW Output Power vs Supply Voltage Chs 1&3 in Parallel Mode (Note18), f = 1kHz, RL = 4Ω, 80kHz BW 20067324 20067341 Safe Area SpiKe Protection Response 200673C7 200673C6 www.national.com 14 LM4781 Typical Performance Characteristics (Continued) Frequency Response of Demo Board POUT = 10W/Channel = 0dB, RIN = 47kΩ RL = 8Ω, No Filters 200673C9 Note 17: Bridge mode graphs were taken using the demo board and inverting the signal to the channel B input. Note 18: Parallel mode graphs were taken using the demo board connecting each output through a 0.1Ω/3W resistor to the load. be in mute mode. The series combination of RM1 and RM2 needs to satisfy the mute equation above for all operating voltages or mute mode may be activated during normal operation. For a longer turn-on mute time, a larger time constant, τ = RC = RM1CM (sec), is needed. For the values show above and with the MUTE pins tied together, the LM4781 will enter play mode when the voltage at point A is -25.1V. The voltage at point A is found with Equation (1) below. (1) VA(t) = (Vf - VO)e-t/τ (Volts) where: t = time (sec) τ = RC (sec) Vo = Voltage on C at t = 0 (Volts) Vf = Final voltage, -VEE in this circuit (Volts) Application Information MUTE MODE The muting function allows the user to mute the amplifier. This can be accomplished as shown in the Typical Application Circuit. The resistor RM is chosen with reference to the negative supply voltage and is used in conjunction with a switch. The switch, when opened or switched to GND, cuts off the current flow from the MUTE pins to −VEE, thus placing the LM4781 into mute mode. Refer to the Mute Attenuation vs Mute Current curves in the Typical Performance Characteristics section for values of attenuation per current out of each MUTE pin. The resistance RM is calculated by the following equation: RM ≤ (|−VEE| − 2.6V) / IMUTE Where IMUTE ≥ 0.5mA for each MUTE pin. The MUTE pins can be tied together so that only one resistor is required for the mute function. The mute resistor value must be chosen so that a minimum of 1.5mA is pulled through the resistor RM. This ensures that each amplifier is fully operational. Taking into account supply line fluctuations, it is a good idea to pull out 1mA per MUTE pin or 3mA total if all pins are tied together. A turn-on MUTE or soft start circuit may also be used during power up. A simple circuit like the one shown below may be used. UNDER-VOLTAGE PROTECTION Upon system power-up, the under-voltage protection circuitry allows the power supplies and their corresponding capacitors to come up close to their full values before turning on the LM4781. Since the supplies have essentially settled to their final value, no DC output spikes occur. At power down, the outputs of the LM4781 are forced to ground before the power supply voltages fully decay preventing transients on the output. OVER-VOLTAGE PROTECTION The LM4781 contains over-voltage protection circuitry that limits the output current while also providing voltage clamping. The clamp does not, however, use internal clamping diodes. The clamping effect is quite the same because the output transistors are designed to work alternately by sinking large current spikes. 200673B7 SPiKe PROTECTION The LM4781 is protected from instantaneous peaktemperature stressing of the power transistor array. The Safe Operating graph in the Typical Performance Characteris- The RC combination of CM and RM1 may cause the voltage at point A to change more slowly than the -VEE supply voltage. Until the voltage at point A is low enough to have 0.5mA of current per MUTE pin flow through RM2, the IC will 15 www.national.com LM4781 Application Information resistance is analogous to electrical resistance, and temperature drops are analogous to voltage drops, the power dissipation out of the LM4781 is equal to the following: (Continued) tics section shows the area of device operation where SPiKe Protection Circuitry is not enabled. The SPiKe Protection Response waveform graph shows the waveform distortion when SPiKe is enabled. Please refer to AN-898 for more detailed information. PDMAX = (TJMAX−TAMB) / θJA (3) where TJMAX = 150˚C, TAMB is the system ambient temperature and θJA = θJC + θCS + θSA. THERMAL PROTECTION The LM4781 has a sophisticated thermal protection scheme to prevent long-term thermal stress of the device. When the temperature on the die exceeds 150˚C, the LM4781 shuts down. It starts operating again when the die temperature drops to about 145˚C, but if the temperature again begins to rise, shutdown will occur again above 150˚C. Therefore, the device is allowed to heat up to a relatively high temperature if the fault condition is temporary, but a sustained fault will cause the device to cycle in a Schmitt Trigger fashion between the thermal shutdown temperature limits of 150˚C and 145˚C. This greatly reduces the stress imposed on the IC by thermal cycling, which in turn improves its reliability under sustained fault conditions. Since the die temperature is directly dependent upon the heat sink used, the heat sink should be chosen so that thermal shutdown is not activated during normal operation. Using the best heat sink possible within the cost and space constraints of the system will improve the long-term reliability of any power semiconductor device, as discussed in the Determining the Correct Heat Sink section. 200673B8 Once the maximum package power dissipation has been calculated using Equation 2, the maximum thermal resistance, θSA, (heat sink to ambient) in ˚C/W for a heat sink can be calculated. This calculation is made using Equation 4 which is derived by solving for θSA in Equation 3. θSA = [(TJMAX−TAMB)−PDMAX(θJC +θCS)] / PDMAX (4) Again it must be noted that the value of θSA is dependent upon the system designer’s amplifier requirements. If the ambient temperature that the audio amplifier is to be working under is higher than 25˚C, then the thermal resistance for the heat sink, given all other things are equal, will need to be smaller. DETERMlNlNG MAXIMUM POWER DISSIPATION Power dissipation within the integrated circuit package is a very important parameter requiring a thorough understanding if optimum power output is to be obtained. An incorrect maximum power dissipation calculation may result in inadequate heat sinking causing thermal shutdown and thus limiting the output power. Equation (2) shows the theoretical maximum power dissipation point for each amplifier in a single-ended configuration where VCC is the total supply voltage. PDMAX = (VCC)2 / 2π2RL (2) SUPPLY BYPASSING The LM4781 has excellent power supply rejection and does not require a regulated supply. However, to improve system performance as well as eliminate possible oscillations, the LM4781 should have its supply leads bypassed with lowinductance capacitors having short leads that are located close to the package terminals. Inadequate power supply bypassing will manifest itself by a low frequency oscillation known as “motorboating” or by high frequency instabilities. These instabilities can be eliminated through multiple bypassing utilizing a large tantalum or electrolytic capacitor (10µF or larger) which is used to absorb low frequency variations and a small ceramic capacitor (0.1µF) to prevent any high frequency feedback through the power supply lines. If adequate bypassing is not provided, the current in the supply leads which is a rectified component of the load current may be fed back into internal circuitry. This signal causes distortion at high frequencies requiring that the supplies be bypassed at the package terminals with an electrolytic capacitor of 470µF or more. Thus by knowing the total supply voltage and rated output load, the maximum power dissipation point can be calculated. The package dissipation is three times the number which results from Equation (2) since there are three amplifiers in each LM4781. Refer to the graphs of Power Dissipation versus Output Power in the Typical Performance Characteristics section which show the actual full range of power dissipation not just the maximum theoretical point that results from Equation (2). BRIDGED AMPLIFIER APPLICATION The LM4781 has three operational amplifiers internally, allowing for a few different amplifier configurations. One of these configurations is referred to as “bridged mode” and involves driving the load differentially through two of the LM4781’s outputs. This configuration is shown in Figure 2. Bridged mode operation is different from the classical singleended amplifier configuration where one side of its load is connected to ground. A bridge amplifier design has a distinct advantage over the single-ended configuration, as it provides differential drive to the load, thus doubling output swing for a specified supply voltage. Theoretically, four times the output power is pos- DETERMINING THE CORRECT HEAT SINK The choice of a heat sink for a high-power audio amplifier is made entirely to keep the die temperature at a level such that the thermal protection circuitry is not activated under normal circumstances. The thermal resistance from the die to the outside air, θJA (junction to ambient), is a combination of three thermal resistances, θJC (junction to case), θCS (case to sink), and θSA (sink to ambient). The thermal resistance, θJC (junction to case), of the LM4781T is 0.9˚C/W. Using Thermalloy Thermacote thermal compound, the thermal resistance, θCS (case to sink), is about 0.2˚C/W. Since convection heat flow (power dissipation) is analogous to current flow, thermal www.national.com 16 within an IC package is found by adding up the power dissipation for each amplifier in the IC package. Using the calculated PDMAX the correct heat sink size can be determined. Refer to the section, Determining the Correct Heat Sink, for more information and detailed discussion of proper heat sinking. If only two amplifiers of the LM4781 are used in parallel mode then the third amplifier should have a load impedance equal to or higher than the equivalent impedance seen by each of the amplifiers in parallel mode. Having the same load impedance on all amplifiers means that the power dissipation in each amplifier will be equal. Using a lower load impedance on the third amplifier will result in higher power dissipation in the third amplifier than the other two amplifiers and may result in unwanted activation of thermal shut down on the third amplifier. Having a higher impedance on the third amplifier than the equivalent impedance on the two amplifiers in parallel will reduce total IC package power dissipation reducing the heat sink size requirement. (Continued) sible as compared to a single-ended amplifier under the same conditions. This increase in attainable output power assumes that the amplifier is not current limited or clipped. A direct consequence of the increased power delivered to the load by a bridge amplifier is an increase in internal power dissipation. For each operational amplifier in a bridge configuration, the internal power dissipation will increase by a factor of two over the single ended dissipation. Using Equation (2) the load impedance should be divided by a factor of two to find the maximum power dissipation point for each amplifier in a bridge configuration. In the case of an 8Ω load in a bridge configuration, the value used for RL in Equation (2) would be 4Ω for each amplifier in the bridge. When using two of the amplifiers of the LM4781 in bridge mode, the third amplifier should have a load impedance equal to or higher than the equivalent impedance seen by each of the bridged amplifiers. In the example above where the bridge load is 8Ω and each amplifier in the bridge sees a load value of 4Ω then the third amplifier should also have a 4Ω load impedance or higher. Using a lower load impedance on the third amplifier will result in higher power dissipation in the third amplifier than the other two amplifiers and may result in unwanted activation of thermal shut down on the third amplifier. Once the impedance seen by each amplifier is known then Equation (2) can be used to calculated the value of PDMAX for each amplifier. The PDMAX of the IC package is found by adding up the power dissipation for each amplifier within the IC package. This value of PDMAX can be used to calculate the correct size heat sink for a bridged amplifier application. Since the internal dissipation for a given power supply and load is increased by using bridged-mode, the heatsink’s θSA will have to decrease accordingly as shown by Equation 4. Refer to the section, Determining the Correct Heat Sink, for a more detailed discussion of proper heat sinking for a given application. BI-AMP AND TRI-AMP APPLICATIONS Bi-amping is the practice of using two different amplifiers to power the individual drivers in a speaker enclosure. For example, a two-way speaker enclosure might have a tweeter and a subwoofer. One amplifier would drive the tweeter and another would drive the subwoofer. One advantage is that the gain of each amplifier can be adjusted for the different driver sensitivities. Another advantage is the crossover can be designed before the amplifier stages with low cost op amps instead of large passive components. With the crossover before the amplifier stages no power is wasted in the passive crossover as each individual amplifier provides the correct frequencies for the driver. Tri-Amping is using three different amplifier stages in the same way bi-amping is done. Bi-amping can also be done on a three-way speaker design by using one amplifier for the subwoofer and another for the midrange and tweeter. The LM4781 is perfectly suited for bi-amp or tri-amp applications with it’s three amplifiers. Two of the amplifiers can be configured for bridge or parallel mode to drive a subwoofer with the third amplifier driving the tweeter or tweeter and midrange. An example would be to use a 4Ω subwoofer and 8Ω tweeter/midrange with the LM4781 in parallel and singleended modes. Each amplifier would see an 8Ω load but the subwoofer would have twice the output power as the tweeter/midrange. The gain of each amplifier may also be adjusted for the desired response. Using the LM4781 in a tri-amp configuration would allow the gain of each amplifier to be adjusted to achieve the desired speaker response. PARALLEL AMPLIFIER APPLICATION Parallel configuration is normally used when higher output current is needed for driving lower impedance loads (i.e. 4Ω or lower) to obtain higher output power levels. As shown in Figure 3 , the parallel amplifier configuration consist of designing the amplifiers in the IC to have identical gain, connecting the inputs in parallel and then connecting the outputs in parallel through a small external output resistor. Any number of amplifiers can be connected in parallel to obtain the needed output current or to divide the power dissipation across multiple IC packages. Ideally, each amplifier shares the output current equally. Due to slight differences in gain the current sharing will not be equal among all channels. If current is not shared equally among all channels then the power dissipation will also not be equal among all channels. It is recommended that 0.1% tolerance resistors be used to set the gain (Ri and Rf) for a minimal amount of difference in current sharing. When operating two or more amplifiers in parallel mode the impedance seen by each amplifier is equal to the total load impedance multiplied by the number of amplifiers driving the load in parallel as shown by Equation (5) below: (5) RL(parallel) = RL(total) * Number of amplifiers Once the impedance seen by each amplifier in the parallel configuration is known then Equation (2) can be used with this calculated impedance to find the amount of power dissipation for each amplifier. Total power dissipation (PDMAX) SINGLE-SUPPLY AMPLIFIER APPLICATION The typical application of the LM4781 is a split supply amplifier. But as shown in Figure 4, the LM4781 can also be used in a single power supply configuration. This involves using some external components to create a half-supply bias which is used as the reference for the inputs and outputs. Thus, the signal will swing around half-supply much like it swings around ground in a split-supply application. Along with proper circuit biasing, a few other considerations must be accounted for to take advantage of all of the LM4781 functions, like the mute function. CLICKS AND POPS In the typical application of the LM4781 as a split-supply audio power amplifier, the IC exhibits excellent “click” and “pop” performance when utilizing the mute mode. In addition, 17 www.national.com LM4781 Application Information LM4781 Application Information oscillations. If the value of RIN is decreased then the value of CIN will need to increase in order to maintain the same -3dB frequency response. (Continued) the device employs Under-Voltage Protection, which eliminates unwanted power-up and power-down transients. The basis for these functions are a stable and constant halfsupply potential. In a split-supply application, ground is the stable half-supply potential. But in a single-supply application, the half-supply needs to charge up at the same rate as the supply rail, VCC. This makes the task of attaining a clickless and popless turn-on more challenging. Any uneven charging of the amplifier inputs will result in output clicks and pops due to the differential input topology of the LM4781. To achieve a transient free power-up and power-down, the voltage seen at the input terminals should be ideally the same. Such a signal will be common-mode in nature, and will be rejected by the LM4781. In Figure 4, the resistor RINP serves to keep the inputs at the same potential by limiting the voltage difference possible between the two nodes. This should significantly reduce any type of turn-on pop, due to an uneven charging of the amplifier inputs. This charging is based on a specific application loading and thus, the system designer may need to adjust these values for optimal performance. As shown in Figure 4, the resistors labeled RBI help bias up the LM4781 off the half-supply node at the emitter of the 2N3904. But due to the input and output coupling capacitors in the circuit, along with the negative feedback, there are two different values of RBI, namely 10kΩ and 200kΩ. These resistors bring up the inputs at the same rate resulting in a popless turn-on. Adjusting these resistors values slightly may reduce pops resulting from power supplies that ramp extremely quick or exhibit overshoot during system turn-on. HIGH PERFORMANCE CONSIDERATIONS Using low cost electrolytic capacitors in the signal path such as CIN and Ci (see Figures 1 - 5) will result in very good performance. However, electrolytic capacitors are less linear than other premium capacitors. Higher THD+N performance may be obtained by using high quality polypropylene capacitors in the signal path. A more cost effective solution may be the use of smaller value premium capacitors in parallel with the larger electrolytic capacitors. This will maintain signal quality in the upper audio band where any degradation is most noticeable while also coupling in the signals in the lower audio band for good bass response. Distortion is introduced as the audio signal approaches the lower -3dB point, determined as discussed in the section above. By using larger values of capacitors such that the -3dB point is well outside of the audio band will reduce this distortion and improve THD+N performance. Increasing the value of the large supply bypass capacitors will improve burst power output. The larger the supply bypass capacitors the higher the output pulse current without supply droop increasing the peak output power. This will also increase the headroom of the amplifier and reduce THD. SIGNAL-TO-NOISE RATIO In the measurement of the signal-to-noise ratio, misinterpretations of the numbers actually measured are common. One amplifier may sound much quieter than another, but due to improper testing techniques, they appear equal in measurements. This is often the case when comparing integrated circuit designs to discrete amplifier designs. Discrete transistor amps often “run out of gain” at high frequencies and therefore have small bandwidths to noise as indicated below. PROPER SELECTION OF EXTERNAL COMPONENTS Proper selection of external components is required to meet the design targets of an application. The choice of external component values that will affect gain and low frequency response are discussed below. The gain of each amplifier is set by resistors Rf and Ri for the non-inverting configuration shown in Figure 1. The gain is found by Equation (6) below: (6) AV = 1 + Rf / Ri (V/V) For best noise performance, lower values of resistors are used. A value of 1kΩ is commonly used for Ri and then setting the value of Rf for the desired gain. For the LM4781 the gain should be set no lower than 10V/V and no higher than 50V/V. Gain settings below 10V/V may experience instability and using the LM4781 for gains higher than 50V/V will see an increase in noise and THD. The combination of Ri with Ci (see Figure 1) creates a high pass filter. The low frequency response is determined by these two components. The -3dB point can be found from Equation (7) shown below: (7) fi = 1 / (2πRiCi) (Hz) If an input coupling capacitor is used to block DC from the inputs as shown in Figure 5, there will be another high pass filter created with the combination of CIN and RIN. When using a input coupling capacitor RIN is needed to set the DC bias point on the amplifier’s input terminal. The resulting -3dB frequency response due to the combination of CIN and RIN can be found from Equation (8) shown below: (8) fIN = 1 / (2πRINCIN) (Hz) With large values of RIN oscillations may be observed on the outputs when the inputs are left floating. Decreasing the value of RIN or not letting the inputs float will remove the www.national.com 20067399 Integrated circuits have additional open loop gain allowing additional feedback loop gain in order to lower harmonic distortion and improve frequency response. It is this additional bandwidth that can lead to erroneous signal-to-noise measurements if not considered during the measurement process. In the typical example above, the difference in bandwidth appears small on a log scale but the factor of 10in bandwidth, (200kHz to 2MHz) can result in a 10dB theoretical difference in the signal-to-noise ratio (white noise is proportional to the square root of the bandwidth in a system). In comparing audio amplifiers it is necessary to measure the magnitude of noise in the audible bandwidth by using a “weighting” filter (Note 19). A “weighting” filter alters the frequency response in order to compensate for the average human ear’s sensitivity to the frequency spectra. The weighting filters at the same time provide the bandwidth limiting as discussed in the previous paragraph. 18 surges are important only if the lead length exceeds a couple feet ( > 1µH lead inductance). Twisting together the supply and ground leads minimizes the effect. (Continued) Note 19: CCIR/ARM: A Practical Noise Measurement Method; by Ray Dolby, David Robinson and Kenneth Gundry, AES Preprint No. 1353 (F-3). In addition to noise filtering, differing meter types give different noise readings. Meter responses include: 1. 2. RMS reading, average responding, 3. 4. peak reading, and quasi peak reading. PHYSICAL IC MOUNTING CONSIDERATIONS Mounting of the package to a heat sink must be done such that there is sufficient pressure from the mounting screws to insure good contact with the heat sink for efficient heat flow. Over tightening the mounting screws will cause the package to warp reducing contact area with the heat sink. Less contact with the heat sink will increase the thermal resistance from the package case to the heat sink (θCS) resulting in higher operating die temperatures and possible unwanted thermal shut down activation. Extreme over tightening of the mounting screws will cause severe physical stress resulting in cracked die and catastrophic IC failure. The recommended mounting screw size is M3 with a maximum torque of 50 N-cm. Additionally, it is best to use washers under the screws to distribute the force over a wider area or a screw with a wide flat head. To further distribute the mounting force a solid mounting bar in front of the package and secured in place with the two mounting screws may be used. Other mounting options include a spring clip. If the package is secured with pressure on the front of the package the maximum pressure on the molded plastic should not exceed 150N/mm2. Additionally, if the mounting screws are used to force the package into correct alignment with the heat sink, package stress will be increased. This increase in package stress will result in reduced contact area with the heat sink increasing die operating temperature and possible catastrophic IC failure. Although theoretical noise analysis is derived using true RMS based calculations, most actual measurements are taken with ARM (Average Responding Meter) test equipment. Typical signal-to-noise figures are listed for an A-weighted filter which is commonly used in the measurement of noise. The shape of all weighting filters is similar, with the peak of the curve usually occurring in the 3kHz–7kHz region. LEAD INDUCTANCE Power op amps are sensitive to inductance in the output leads, particularly with heavy capacitive loading. Feedback to the input should be taken directly from the output terminal, minimizing common inductance with the load. Lead inductance can also cause voltage surges on the supplies. With long leads to the power supply, energy is stored in the lead inductance when the output is shorted. This energy can be dumped back into the supply bypass capacitors when the short is removed. The magnitude of this transient is reduced by increasing the size of the bypass capacitor near the IC. With at least a 20µF local bypass, these voltage 19 www.national.com LM4781 Application Information LM4781 Application Information The load current IL will be much larger than input bias current II, thus V1 will follow the output voltage directly, i.e. in phase. Therefore the voltage appearing at the non-inverting input is effectively positive feedback and the circuit may oscillate. If there was only one device to worry about then the values of R1 and R2 would probably be small enough to be ignored; however, several devices normally comprise a total system. Any ground return of a separate device, whose output is in phase, can feedback in a similar manner and cause instabilities. Out of phase ground loops also are troublesome, causing unexpected gain and phase errors. (Continued) LAYOUT, GROUND LOOPS AND STABILITY The LM4781 is designed to be stable when operated at a closed-loop gain of 10 or greater, but as with any other high-current amplifier, the LM4781 can be made to oscillate under certain conditions. These oscillations usually involve printed circuit board layout or output/input coupling issues. When designing a layout, it is important to return the load ground, the output compensation ground, and the low level (feedback and input) grounds to the circuit board common ground point through separate paths. Otherwise, large currents flowing along a ground conductor will generate voltages on the conductor which can effectively act as signals at the input, resulting in high frequency oscillation or excessive distortion. It is advisable to keep the output compensation components and the 0.1µF supply decoupling capacitors as close as possible to the LM4781 to reduce the effects of PCB trace resistance and inductance. For the same reason, the ground return paths should be as short as possible. The solution to most ground loop problems is to always use a single-point ground system, although this is sometimes impractical. The third figure above is an example of a singlepoint ground system. The single-point ground concept should be applied rigorously to all components and all circuits when possible. Violations of single-point grounding are most common among printed circuit board designs, since the circuit is surrounded by large ground areas which invite the temptation to run a device to the closest ground spot. As a final rule, make all ground returns low resistance and low inductance by using large wire and wide traces. Occasionally, current in the output leads (which function as antennas) can be coupled through the air to the amplifier input, resulting in high-frequency oscillation. This normally happens when the source impedance is high or the input leads are long. The problem can be eliminated by placing a small capacitor, CC, (on the order of 50pF to 500pF) across the LM4781 input terminals. Refer to the External Components Description section relating to component interaction with Cf. In general, with fast, high-current circuitry, all sorts of problems can arise from improper grounding which again can be avoided by returning all grounds separately to a common point. Without isolating the ground signals and returning the grounds to a common point, ground loops may occur. “Ground Loop” is the term used to describe situations occurring in ground systems where a difference in potential exists between two ground points. Ideally a ground is a ground, but unfortunately, in order for this to be true, ground conductors with zero resistance are necessary. Since real world ground leads possess finite resistance, currents running through them will cause finite voltage drops to exist. If two ground return lines tie into the same path at different points there will be a voltage drop between them. The first figure below shows a common ground example where the positive input ground and the load ground are returned to the supply ground point via the same wire. The addition of the finite wire resistance, R2, results in a voltage difference between the two points as shown below. REACTIVE LOADING It is hard for most power amplifiers to drive highly capacitive loads very effectively and normally results in oscillations or ringing on the square wave response. If the output of the LM4781 is connected directly to a capacitor with no series resistance, the square wave response will exhibit ringing if the capacitance is greater than about 0.2µF. If highly capacitive loads are expected due to long speaker cables, a method commonly employed to protect amplifiers from low impedances at high frequencies is to couple to the load through a 10Ω resistor in parallel with a 0.7µH inductor. The inductor-resistor combination as shown in the Figure 5 isolates the feedback amplifier from the load by providing high output impedance at high frequencies thus allowing the 10Ω resistor to decouple the capacitive load and reduce the Q of the series resonant circuit. The LR combination also provides low output impedance at low frequencies thus shorting out the 10Ω resistor and allowing the amplifier to drive the series RC load (large capacitive load due to long speaker cables) directly. INVERTING AMPLIFIER APPLICATION The inverting amplifier configuration may be used instead of the more common non-inverting amplifier configuration shown in Figure 1. The inverting amplifier can have better THD+N performance and eliminates the need for a large capacitor (Ci) reducing cost and space requirements. The values show in Figure 6 are only one example of an amplifier with a gain of 20V/V (Gain = -Rf/Ri). For different resistor values, the value of RB should be eqaul to the parallel combination of Rf and Ri. 20067398 www.national.com 20 LM4781 Application Information (Continued) If the DC blocking input capacitor (CIN) is used as shown, the lower -3dB point is found using Equation (8) as discussed in the Proper Selection of External Components section. 200673B9 FIGURE 6. Inverting Amplifier Application Circuit 21 www.national.com LM4781 Application Information (Continued) 200673C8 FIGURE 7. Reference PCB Schematic www.national.com 22 LM4781 Application Information (Continued) LM4781 REFERENCE BOARD ARTWORK 200673B0 200673A9 Silk Layer Composite Layer 200673A8 200673B1 Top Layer Bottom Layer 23 www.national.com LM4781 Application Information (Continued) Bill Of Materials for Reference PCB Symbol Value Tolerance Type/Description RIN1, RIN2, RIN3 33kΩ 5% 1/4 Watt RB1, RB2, RB3 1kΩ 1% 1/4 Watt RF1, RF2, RF3 20kΩ 1% 1/4 Watt Ri1, Ri2, Ri3 1kΩ 1% 1/4 Watt RSN1, RSN2, RSN3 4.7Ω 5% 1/4 Watt RG 2.7Ω 5% 1/4 Watt RM 5.6kΩ 5% 1/4 Watt CIN1, CIN2, CIN3 1µF 10% Metallized Polyester Film Ci1, Ci2, Ci3 47µF 20% Electrolytic Radial / 50V CSN1, CSN2, CSN3 0.1µF 20% Monolithic Ceramic CS1, CS2 0.1µF 20% Monolithic Ceramic CS3, CS4 10µF 20% Electrolytic Radial / 50V CS5, CS6 2,200µF 20% Electrolytic Radial / 50V S1 SPDT (on-on) Switch J 1, J 2, J 3 Non-Switched PC Mount RCA Jack J5, J7, J9, J11 PCB Banana Jack - BLACK J4, J6, J8, J10, J12 PCB Banana Jack - RED U1 27 lead TO-220 Power Socket with push release lever or LM4781 IC www.national.com 24 Comment LM4781 Overture™ Audio Power Amplifier Series 3 Channel 35W Audio Power Amplifier with Mute Physical Dimensions inches (millimeters) unless otherwise noted Non-Isolated TO-220 27-Lead Package Order Number LM4781TA NS Package Number TA27A LIFE SUPPORT POLICY NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR SYSTEMS WITHOUT THE EXPRESS WRITTEN APPROVAL OF THE PRESIDENT AND GENERAL COUNSEL OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein: 1. Life support devices or systems are devices or systems which, (a) are intended for surgical implant into the body, or (b) support or sustain life, and whose failure to perform when properly used in accordance with instructions for use provided in the labeling, can be reasonably expected to result in a significant injury to the user. National Semiconductor Americas Customer Support Center Email: [email protected] Tel: 1-800-272-9959 www.national.com National Semiconductor Europe Customer Support Center Fax: +49 (0) 180-530 85 86 Email: [email protected] Deutsch Tel: +49 (0) 69 9508 6208 English Tel: +44 (0) 870 24 0 2171 Français Tel: +33 (0) 1 41 91 8790 2. A critical component is any component of a life support device or system whose failure to perform can be reasonably expected to cause the failure of the life support device or system, or to affect its safety or effectiveness. National Semiconductor Asia Pacific Customer Support Center Email: [email protected] National Semiconductor Japan Customer Support Center Fax: 81-3-5639-7507 Email: [email protected] Tel: 81-3-5639-7560 National does not assume any responsibility for use of any circuitry described, no circuit patent licenses are implied and National reserves the right at any time without notice to change said circuitry and specifications.