TS4902 300mW at 3.3V SUPPLY AUDIO POWER AMPLIFIER WITH STANDBY MODE ACTIVE LOW ■ OPERATING FROM VCC = 2.2V to 5.5V ■ 0.7W OUTPUT POWER @ Vcc=5V, THD=1%, PIN CONNECTIONS (top view) f=1kHz, with an 8Ω load ■ 0.3W OUTPUT POWER @ Vcc=3.3V, THD=1%, f=1kHz, with an 8Ω load ■ ULTRA LOW CONSUMPTION IN STANDBY MODE (10nA) ■ 77dB PSRR @ 217Hz from 5V to 2.2V ■ ULTRA LOW POP & CLICK ■ ULTRA LOW DISTORTION (0.1%) ■ UNITY GAIN STABLE ■ AVAILABLE IN MiniSO8 & SO8 TS4902IS-TS4902IST - MiniSO8 Standby 1 8 VOUT2 Bypass 2 7 GND VIN+ 3 6 VCC VIN- 4 5 VOUT1 DESCRIPTION The TS4902 is an audio power amplifier designed to provide the best price to power ratio while preserving high audio quality. Available in MiniSO8 & SO8 package, it is capable of delivering up to 0.7W of continuous RMS ouput power into an 8Ω load @ 5V. TS4902 is also exhibiting an outstanding 0.1% distortion level (THD) from a 5V supply for a Pout of 200mW RMS. TS4902ID-TS4902IDT - SO8 Standby 1 8 VOUT2 Bypass 2 7 GND VIN+ 3 6 VCC VIN- 4 5 VOUT1 An externally controlled standby mode reduces the supply current to less than 10nA. It also includes an internal thermal shutdown protection. The unity-gain stable amplifier can be configured by external gain setting resistors. APPLICATIONS ■ Mobile Phones (Cellular / Cordless) ■ PDAs ■ Portable Audio Devices TYPICAL APPLICATION SCHEMATIC Cfeed Rfeed VCC Cs 6 Audio Input Rin 4 Vin- - 3 Vin+ + V C Vout 1 Cin ORDER CODE Part Number TS4902IST TS4902ID Temperature Range -40, +85°C 5 RL 8 Ohms Package - VCC ST AV = -1 D • • 2 Bypass 1 Standby Bias Rstb GND Cb Vout 2 8 + TS4902 7 S = MiniSO Package (MiniSO) is only available in Tape & Reel (ST) D = Small Outline Package (SO) - also available in Tape & Reel (DT) January 2002 1/19 TS4902 ABSOLUTE MAXIMUM RATINGS Symbol VCC Vi Parameter Supply voltage Input Voltage 1) 2) Unit 6 V GND to VCC V °C Toper Operating Free Air Temperature Range -40 to + 85 Tstg Storage Temperature Tj Rthja Pd -65 to +150 °C Maximum Junction Temperature 150 °C Thermal Resistance Junction to Ambient 3) SO8 MiniSO8 175 215 Power Dissipation 4) ESD Human Body Model ESD Machine Model Latch-up Latch-up Immunity Lead Temperature (soldering, 10sec) 1. 2. 3. 4. Value °C/W See the power derating curves Fig 20. 2 200 Class A 250 kV V Value Unit 2.2 to 5.5 V °C All voltages values are measured with respect to the ground pin. The magnitude of input signal must never exceed VCC + 0.3V / G ND - 0.3V Device is protected in case of over temperature by a thermal shutdown active @ 150°C. Exceeding the power derating curves during a long period, will cause abnormal operation. OPERATING CONDITIONS Symbol Parameter VCC Supply Voltage VICM Common Mode Input Voltage Range GND to VCC - 1.5V V VSTB Standby Voltage Input : Device ON Device OFF 1.5 ≤ VSTB ≤ VCC GND ≤ VSTB ≤ 0.5 V 4 - 32 Ω RL Rthja Load Resistor Thermal Resistance Junction to Ambient 1) SO8 MiniSO8 1. This thermal resistance can be reduced with a suitable PCB layout (see Power Derating Curves) 2/19 °C/W 150 190 TS4902 ELECTRICAL CHARACTERISTICS VCC = +5V, GND = 0V, Tamb = 25°C (unless otherwise specified) Symbol Typ. Max. Unit Supply Current No input signal, no load 6 8 mA Standby Current 1) No input signal, Vstdby = GND, RL = 8Ω 10 1000 nA Voo Output Offset Voltage No input signal, RL = 8Ω 5 20 mV Po Output Power THD = 1% Max, f = 1kHz, RL = 8Ω 0.7 W Total Harmonic Distortion + Noise Po = 250mW rms, Gv = 2, 20Hz < f < 20kHz, RL = 8Ω 0.15 % Power Supply Rejection Ratio 2) f = 217Hz, RL = 8Ω, RFeed = 22KΩ, Vripple = 200mV rms 77 dB ΦM Phase Margin at Unity Gain RL = 8Ω, CL = 500pF 70 Degrees GM Gain Margin RL = 8Ω, CL = 500pF 20 dB GBP Gain Bandwidth Product RL = 8Ω 2 MHz ICC ISTANDBY THD + N PSRR Parameter Min. 1. Standby mode is actived when Vstdby is tied to GND 2. Dynamic measurements - 20*log(rms(Vout)/rms(Vripple)). Vripple is the surimposed sinus signal to Vcc @ f = 217Hz VCC = +3.3V, GND = 0V, Tamb = 25°C (unless otherwise specified)3) Symbol Typ. Max. Unit Supply Current No input signal, no load 5.5 8 mA Standby Current 1) No input signal, Vstdby = GND, RL = 8Ω 10 1000 nA Voo Output Offset Voltage No input signal, RL = 8Ω 5 20 mV Po Output Power THD = 1% Max, f = 1kHz, RL = 8Ω 300 mW Total Harmonic Distortion + Noise Po = 250mW rms, Gv = 2, 20Hz < f < 20kHz, RL = 8Ω 0.15 % Power Supply Rejection Ratio 2) f = 217Hz, RL = 8Ω, RFeed = 22KΩ, Vripple = 200mV rms 77 dB ΦM Phase Margin at Unity Gain RL = 8Ω, CL = 500pF 70 Degrees GM Gain Margin RL = 8Ω, CL = 500pF 20 dB GBP Gain Bandwidth Product RL = 8Ω 2 MHz ICC ISTANDBY THD + N PSRR Parameter Min. 1. Standby mode is actived when Vstdby is tied to GND 2. Dynamic measurements - 20*log(rms(Vout)/rms(Vripple)). Vripple is the surimposed sinus signal to Vcc @ f = 217Hz 3. All electrical values are made by correlation between 2.6V and 5V measurements 3/19 TS4902 ELECTRICAL CHARACTERISTICS VCC = 2.6V, GND = 0V, Tamb = 25°C (unless otherwise specified) Symbol Typ. Max. Unit Supply Current No input signal, no load 5.5 8 mA Standby Current 1) No input signal, Vstdby = GND, RL = 8Ω 10 1000 nA Voo Output Offset Voltage No input signal, RL = 8Ω 5 20 mV Po Output Power THD = 1% Max, f = 1kHz, RL = 8Ω 180 mW Total Harmonic Distortion + Noise Po = 200mW rms, Gv = 2, 20Hz < f < 20kHz, RL = 8Ω 0.15 % Power Supply Rejection Ratio 2) f = 217Hz, RL = 8Ω, RFeed = 22KΩ, Vripple = 200mV rms 77 dB ΦM Phase Margin at Unity Gain RL = 8Ω, CL = 500pF 70 Degrees GM Gain Margin RL = 8Ω, CL = 500pF 20 dB GBP Gain Bandwidth Product RL = 8Ω 2 MHz ICC ISTANDBY THD + N PSRR Parameter Min. 1. Standby mode is actived when Vstdby is tied to GND 2. Dynamic measurements - 20*log(rms(Vout)/rms(Vripple)). Vripple is the surimposed sinus signal to Vcc @ f = 217Hz Components Functional Description Rin Inverting input resistor which sets the closed loop gain in conjunction with Rfeed. This resistor also forms a high pass filter with Cin (fc = 1 / (2 x Pi x Rin x Cin)) Cin Input coupling capacitor which blocks the DC voltage at the amplifier input terminal Rfeed Feed back resistor which sets the closed loop gain in conjunction with Rin Cs Supply Bypass capacitor which provides power supply filtering Cb Bypass pin capacitor which provides half supply filtering Cfeed Rstb Gv Low pass filter capacitor allowing to cut the high frequency (low pass filter cut-off frequency 1 / (2 x Pi x Rfeed x Cfeed)) Pull-up resistor which fixes the right supply level on the standby pin Closed loop gain in BTL configuration = 2 x (Rfeed / Rin) REMARKS 1. All measurements, except PSRR measurements, are made with a supply bypass capacitor Cs = 100µF. 2. The standby response time is about 1µs. 4/19 TS4902 Fig. 1 : Open Loop Frequency Response Fig. 2 : Open Loop Frequency Response 0 40 Phase (Deg) -120 -140 0 Vcc = 5V ZL = 8Ω + 560pF Tamb = 25°C -100 20 -120 -140 0 -180 -20 -200 1 10 100 1000 10000 -200 -220 -40 0.3 1 10 Frequency (kHz) Fig. 3 : Open Loop Frequency Response Gain 60 Vcc = 33V RL = 8Ω Tamb = 25°C -140 -160 Gain (dB) -120 0 Gain -60 -100 20 -220 0 60 -40 Phase (Deg) Gain (dB) Phase 10000 80 -20 -80 40 100 1000 Frequency (kHz) Fig. 4 : Open Loop Frequency Response 0 80 -60 -160 -180 -40 0.3 -40 -80 Phase -160 -20 -20 Vcc = 3.3V ZL = 8Ω + 560pF Tamb = 25°C Phase 10 100 1000 Frequency (kHz) 10000 -140 -160 -180 -200 -20 -220 -240 -40 0.3 Fig. 5 : Open Loop Frequency Response Gain 60 Vcc = 2.6V RL = 8Ω Tamb = 25°C 60 -40 -60 -120 -140 -160 0 10000 Vcc = 2.6V ZL = 8Ω + 560pF Tamb = 25°C Phase -200 1 10 100 1000 Frequency (kHz) 10000 -240 -40 -60 -120 -140 -160 0 -180 -200 -20 -220 -220 -40 0.3 -20 -100 20 -180 -20 -240 -80 40 Gain (dB) -100 20 100 1000 Frequency (kHz) 0 Gain Phase (Deg) Gain (dB) Phase 10 80 -20 -80 40 1 Fig. 6 : Open Loop Frequency Response 0 80 -60 -120 -220 1 -40 -100 20 0 -200 -40 0.3 -20 -80 40 -180 -20 Phase (Deg) -60 -100 20 Gain -40 -80 Phase Gain (dB) 60 Phase (Deg) 40 0 -20 Phase (Deg) Vcc = 5V RL = 8Ω Tamb = 25°C Gain Gain (dB) 60 -40 0.3 1 10 100 1000 Frequency (kHz) 10000 -240 5/19 TS4902 Phase 60 100 -100 80 -120 60 Gain (dB) Gain -140 40 -160 20 0 -20 -40 0.3 -180 1 10 100 -40 0.3 -80 80 -100 Phase Gain (dB) Gain -140 40 -160 20 -180 0 -40 0.3 6/19 -200 Vcc = 2.6V CL = 560pF Tamb = 25°C 1 10 -220 100 1000 Frequency (kHz) 10000 -240 Phase (Deg) -120 60 -20 -180 -220 Fig. 9 : Open Loop Frequency Response -140 -160 -20 10000 -120 20 -200 100 1000 Frequency (kHz) -100 Phase 40 0 Vcc = 5V CL = 560pF Tamb = 25°C -80 Gain Gain (dB) 80 -80 Phase (Deg) 100 Fig. 8 : Open Loop Frequency Response -200 Vcc = 3.3V CL = 560pF Tamb = 25°C 1 10 -220 100 1000 Frequency (kHz) 10000 -240 Phase (Deg) Fig. 7 : Open Loop Frequency Response TS4902 Fig. 10 : Power Supply Rejection Ratio (PSRR) vs Power supply Fig. 11 : Power Supply Rejection Ratio (PSRR) vs Feedback Capacitor -30 -50 -60 -20 -30 PSRR (dB) PSRR (dB) -40 -10 Vripple = 200mVrms Rfeed = 22kΩ Input = floating RL = 8Ω Tamb = 25°C Vcc = 5V to 2.2V Cb = 1µF & 0.1µF -40 Vcc = 5 to 2.2V Cb = 1µF & 0.1µF Rfeed = 22kΩ Vripple = 200mVrms Input = floating RL = 8Ω Tamb = 25°C Cfeed=0 Cfeed=150pF Cfeed=330pF -50 -60 -70 -70 Cfeed=680pF -80 10 100 1000 10000 Frequency (Hz) 100000 Fig. 12 : Power Supply Rejection Ratio (PSRR) vs Bypass Capacitor -80 10 -10 Cb=10µF PSRR (dB) -30 Vcc = 5 to 2.2V Rfeed = 22k Rin = 22k, Cin = 1µF Rg = 100Ω, RL = 8Ω Tamb = 25°C -40 Cin=1µF Cb=47µF -50 100000 Vcc = 5 to 2.2V Rfeed = 22k, Rin = 22k Cb = 1µF Rg = 100Ω, RL = 8Ω Tamb = 25°C Cin=330nF -20 PSRR (dB) -20 1000 10000 Frequency (Hz) Fig. 13 : Power Supply Rejection Ratio (PSRR) vs Input Capacitor -10 Cb=1µF 100 Cin=220nF -30 -40 Cin=100nF -60 -50 Cin=22nF -70 Cb=100µF -80 10 100 1000 10000 100000 -60 10 100 1000 10000 100000 Frequency (Hz) Frequency (Hz) Fig. 14 : Power Supply Rejection Ratio (PSRR) vs Feedback Resistor -10 -20 PSRR (dB) -30 -40 Vcc = 5 to 2.2V Cb = 1µF & 0.1µF Vripple = 200mVrms Input = floating RL = 8Ω Tamb = 25°C Rfeed=110kΩ Rfeed=47kΩ -50 -60 Rfeed=22kΩ -70 -80 10 Rfeed=10kΩ 100 1000 10000 Frequency (Hz) 100000 7/19 TS4902 Fig. 15 : Pout @ THD + N = 1% vs Supply Voltage vs RL Fig. 16 : Pout @ THD + N = 10% vs Supply Voltage vs RL 1.2 0.8 Gv = 2 & 10 Cb = 1µF F = 1kHz BW < 125kHz Tamb = 25°C Output power @ 10% THD + N (W) Output power @ 1% THD + N (W) 1.0 8Ω 4Ω 0.6 16 Ω 0.4 0.2 32 Ω 0.0 2.5 3.0 3.5 4.0 4.5 Gv = 2 & 10 Cb = 1µF F = 1kHz BW < 125kHz Tamb = 25°C 1.0 0.8 4Ω 8Ω 16 Ω 0.6 0.4 0.2 32 Ω 0.0 2.5 5.0 3.0 3.5 Vcc (V) Fig. 17 : Power Dissipation vs Pout Vcc=3.3V f=1kHz THD+N<1% 0.5 RL=4Ω Power Dissipation (W) Power Dissipation (W) 5.0 0.6 Vcc=5V f=1kHz THD+N<1% 1.0 0.8 0.6 RL=8Ω 0.4 0.2 0.4 RL=4Ω 0.4 0.3 0.2 RL=8Ω 0.1 RL=16Ω 0.2 0.0 0.0 4.5 Fig. 18 : Power Dissipation vs Pout 1.4 1.2 4.0 Vcc (V) RL=16Ω 0.6 0.8 0.0 0.0 1.0 0.2 Output Power (W) 0.4 0.6 Output Power (W) Fig. 19 : Power Dissipation vs Pout Fig. 20 : Power Derating Curves 0.40 0.30 RL=4Ω 0.25 0.20 0.15 RL=8Ω 0.10 MiniSO8 on demoboard 0.8 0.6 0.4 MiniSO8 SO8 RL=16Ω 0.0 0.1 0.2 Output Power (W) 8/19 1.0 0.2 0.05 0.00 0.0 SO8 on demoboard 1.2 Vcc=2.6V f=1kHz THD+N<1% Power Dissipation (W) Power Dissipation (W) 0.35 0.3 0 25 50 75 100 Ambiant Temperature (°C) 125 150 TS4902 Fig. 21 : Output Power vs Load Resistance Fig. 22 : Output Power vs Load Resistance 1.0 Output power (W) 0.8 Vcc=5V 0.6 Vcc=4.5V Vcc=4V 0.4 1.2 1.0 Output Power (W) THD+N=1% Gv = 2 & 10 Cb = 1µF F = 1kHz BW < 125kHz Tamb = 25°C THD+N=10% Gv = 2 & 10 Cb = 1µF F = 1kHz BW < 125kHz Tamb = 25°C Vcc=5V 0.8 Vcc=4.5V 0.6 Vcc=4V 0.4 0.2 0.2 Vcc=3.5V Vcc=3.5V Vcc=3V 0.0 8 Vcc=2.5V 0.0 16 24 32 Fig. 23 : Clipping Voltage vs Supply Voltage Vcc=2.5V 16 24 32 Fig. 24 : Frequency response vs Cin & Cfeed 10 Tamb = 25°C 0.9 5 0.8 4Ω High Side 0 4Ω Low Side 0.7 Gain (dB) Dropout Voltage (V) 8 Load Resistance (ohm) Load Resistance (ohm) 1.0 Vcc=3V 0.6 0.5 8Ω High Side 8Ω Low Side 0.4 -10 -20 0.2 2.5 3.0 3.5 4.0 4.5 5.0 Supply Voltage (V) Cfeed = 680pF -5 -15 0.3 Cfeed = 330pF -25 10 Cin = 470nF Cfeed = 2.2nF Cin = 22nF Cin = 82nF Rin = Rfeed = 22kΩ Tamb = 25°C 100 1000 Frequency (Hz) 10000 Fig. 25 : Noise Floor Output Noise Voltage ( V) 100 80 60 Vcc = 2.5V to 5V Rin = Rfeed = 22kΩ Cb = Cin = 1µF Input Grounded BW < 22kHz Tamb = 25°C 40 VOUT1 + VOUT2 Standby = ON 20 0 20 100 1000 Frequency (Hz) 10000 9/19 TS4902 Fig. 26 : THD + N vs Output Power Fig. 27 : THD + N vs Output Power 10 RL = 4Ω, Vcc = 5V Gv = 10 Cb = Cin = 1µF BW < 125kHz, Tamb = 25°C Rl = 4Ω Vcc = 5V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C THD + N (%) THD + N (%) 10 1 20kHz 20kHz 1 20Hz 20Hz, 1kHz 0.1 1E-3 1kHz 0.01 0.1 Output Power (W) 0.1 1E-3 1 Fig. 28 : THD + N vs Output Power 1 Fig. 29 : THD + N vs Output Power 10 10 RL = 4Ω, Vcc = 3.3V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C THD + N (%) THD + N (%) 0.01 0.1 Output Power (W) 1 RL = 4Ω, Vcc = 3.3V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 20kHz 1 20kHz 0.1 20Hz 1kHz 20Hz, 1kHz 0.1 1E-3 0.01 0.1 Output Power (W) 1 Fig. 30 : THD + N vs Output Power 0.01 0.1 Output Power (W) 1 Fig. 31 : THD + N vs Output Power 10 RL = 4Ω, Vcc = 2.6V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C THD + N (%) THD + N (%) 10 1E-3 1 RL = 4Ω, Vcc = 2.6V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 1 20kHz 20kHz 20Hz 0.1 1kHz 20Hz, 1kHz 0.1 1E-3 10/19 0.01 Output Power (W) 0.1 1E-3 0.01 Output Power (W) 0.1 TS4902 Fig. 32 : THD + N vs Output Power Fig. 33 : THD + N vs Output Power 10 10 20Hz, 1kHz THD + N (%) THD + N (%) RL = 8Ω Vcc = 5V Gv = 2 Cb = Cin = 1µF BW < 125kHz 1 Tamb = 25°C 20kHz 0.1 RL = 8Ω Vcc = 5V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 1 20Hz 20kHz 0.1 1kHz 1E-3 0.01 0.1 Output Power (W) 1 1E-3 Fig. 34 : THD + N vs Output Power 1 Fig. 35 : THD + N vs Output Power 10 10 RL = 8Ω, Vcc = 3.3V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C THD + N (%) THD + N (%) 0.01 0.1 Output Power (W) 1 RL = 8Ω, Vcc = 3.3V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 1 20kHz 20Hz 20kHz 20Hz, 1kHz 0.1 0.1 1kHz 1E-3 0.01 0.1 Output Power (W) 1 Fig. 36 : THD + N vs Output Power 0.01 0.1 Output Power (W) 1 Fig. 37 : THD + N vs Output Power 10 RL = 8Ω, Vcc = 2.6V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C THD + N (%) THD + N (%) 10 1E-3 1 20Hz, 1kHz 1 20Hz 20kHz 20kHz 0.1 1E-3 RL = 8Ω, Vcc = 2.6V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 0.1 0.01 Output Power (W) 0.1 1E-3 1kHz 0.01 Output Power (W) 0.1 11/19 TS4902 Fig. 38 : THD + N vs Output Power Fig. 39 : THD + N vs Output Power 10 10 THD + N (%) 1 20kHz RL = 16Ω, Vcc = 5V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 1 THD + N (%) RL = 16Ω, Vcc = 5V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 20kHz 0.1 0.1 1kHz 20Hz, 1kHz 0.01 1E-3 0.01 0.1 Output Power (W) 1 0.01 1E-3 0.01 0.1 Output Power (W) 10 10 THD + N (%) RL = 16Ω, Vcc = 3.3V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 1 20kHz RL = 16Ω Vcc = 3.3V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 1 20kHz 0.1 0.1 1kHz 20Hz 20Hz, 1kHz 0.01 1E-3 0.01 Output Power (W) 0.01 1E-3 0.1 RL = 16Ω Vcc = 2.6V Gv = 2 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C THD + N (%) THD + N (%) 0.1 10 10 20kHz 1 RL = 16Ω Vcc = 2.6V Gv = 10 Cb = Cin = 1µF BW < 125kHz Tamb = 25°C 20Hz 20Hz, 1kHz 0.01 1E-3 20kHz 0.1 0.1 12/19 0.01 Output Power (W) Fig. 43 : THD + N vs Output Power Fig. 42 : THD + N vs Output Power 1 1 Fig. 41 : THD + N vs Output Power Fig. 40 : THD + N vs Output Power THD + N (%) 20Hz 0.01 Output Power (W) 1kHz 0.1 0.01 1E-3 0.01 Output Power (W) 0.1 TS4902 Fig. 44 : Signal to Noise Ratio vs Power Supply with Unweighted Filter (20Hz to 20kHz) Fig. 45 : Signal to Noise Ratio Vs Power Supply with Unweighted Filter (20Hz to 20kHz) 100 90 90 RL=16Ω 80 RL=4Ω RL=8Ω SNR (dB) SNR (dB) 80 70 Gv = 2 Cb = Cin = 1µF THD+N < 0.4% Tamb = 25°C 60 50 2.5 3.0 3.5 4.0 4.5 RL=8Ω 70 RL=4Ω RL=16Ω Gv = 10 Cb = Cin = 1µF THD+N < 0.7% Tamb = 25°C 60 50 2.5 5.0 3.0 3.5 4.0 4.5 5.0 Vcc (V) Vcc (V) Fig. 46 : Signal to Noise Ratio vs Power Supply with Weighted Filter type A Fig. 47 : Signal to Noise Ratio vs Power Supply with Weighted Filter Type A 110 90 100 80 RL=4Ω RL=8Ω 90 SNR (dB) SNR (dB) RL=16Ω 80 Gv = 2 Cb = Cin = 1µF THD+N < 0.4% Tamb = 25°C 70 60 2.5 3.0 3.5 4.0 4.5 RL=8Ω 70 RL=4Ω RL=16Ω Gv = 10 Cb = Cin = 1µF THD+N < 0.7% Tamb = 25°C 60 50 2.5 5.0 3.0 3.5 Fig. 48 : Current Consumption vs Power Supply Voltage 4.5 5.0 Fig. 49 : Current Consumption vs Standby Voltage @ Vcc = 5V 7 7 Vstandby = Vcc Tamb = 25°C 6 6 5 5 Icc (mA) Icc (mA) 4.0 Vcc (V) Vcc (V) 4 3 4 3 2 2 1 1 0 0 0.0 0 1 2 3 Vcc (V) 4 5 Vcc = 5V Tamb = 25°C 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Vstandby (V) 4.0 4.5 5.0 13/19 TS4902 Fig. 50 : Current Consumption vs Standby Voltage @ Vcc = 3.3V Fig. 51 : Current Consumption vs Standby Voltage @ Vcc = 2.6V 6 6 5 5 4 Icc (mA) Icc (mA) 4 3 2 2 1 0 0.0 14/19 3 Vcc = 3.3V Tamb = 25°C 0.5 1.0 1.5 2.0 Vstandby (V) 2.5 3.0 1 0 0.0 Vcc = 2.6V Tamb = 25°C 0.5 1.0 1.5 Vstandby (V) 2.0 2.5 TS4902 ■ BTL Configuration Principle The TS4902 is a monolithic power amplifier with a BTL (Bridge Tied Load) output configuration. BTL means that each end of the load is connected to two single ended output amplifiers. Thus, we have: Single ended output 1 = Vout1 = Vout (V) Single ended output 2 = Vout2 = -Vout (V) In high frequency region, you can limit the bandwidth by adding a capacitor (Cfeed) in parallel with Rfeed. Its form a low pass filter with a -3dB cut off frequency 1 F C H = ----------------------------------------------- ( Hz ) 2π Rfe ed Cfeed ■ Power dissipation and efficiency Hypothesis : And Vout1 - Vout2 = 2Vout (V) The output power is : Pout = ( 2 Vout RMS ) 2 (W ) RL For the same power supply voltage, the output power in BTL configuration is four times higher than the output power in single ended configuration. • Voltage and current in the load are sinusoidal (Vout and Iout) • Supply voltage is a pure DC source (Vcc) Regarding the load we have : V O UT = V PEAK sin ωt (V) and V OU T I OU T = ----------------- (A) RL ■ Gain In Typical Application Schematic (cf. page 1) In flat region (no effect of Cin), the output voltage of the first stage is : R fe ed Vout1 = – Vin -------------------- (V) Rin For the second stage : Vout2 = -Vout1 (V) The differential output voltage is Rfee d Vout2 – Vo ut1 = 2Vin -------------------- (V) Rin The differential gain named gain (Gv) for more convenient usage is : Vout2 – Vou t1 Rfee d G v = --------------------------------------- = 2 -------------------Vin Rin Remark : Vout2 is in phase with Vin and Vout1 is 180 phased with Vin. It means that the positive terminal of the loudspeaker should be connected to Vout2 and the negative to Vout1. ■ Low and high frequency response In low frequency region, the effect of Cin starts. Cin with Rin forms a high pass filter with a -3dB cut off frequency 1 F C L = -------------------------------- ( Hz ) 2 π R in Cin and VPEAK 2 P O U T = ---------------------- (W) 2 RL Then, the average current delivered by the supply voltage is: I CC AVG VPEAK = 2 -------------------- (A) πR L The power delivered by the supply voltage is Psupply = Vcc IccAVG (W) Then, the power dissipated by the amplifier is Pdiss = Psupply - Pout (W) 2 2 Vcc P di ss = ---------------------- P OU T – P O UT (W) π RL and the maximum value is obtained when ∂Pdiss ---------------------- = 0 ∂P OU T and its value is: Pdiss max = 2 Vcc 2 π2RL (W) Remark : This maximum value is only depending on power supply voltage and load values. 15/19 TS4902 The efficiency is the ratio between the output power and the power supply P O UT πV P E A K η = ------------------------ = ----------------------Psup ply 4V C C The maximum theoretical value is reached when Vpeak = Vcc, so π ----- = 78.5% 4 ■ Decoupling of the circuit Two capacitors are needed to bypass properly the TS4902, a power supply bypass capacitor Cs and a bias voltage bypass capacitor Cb. Cs has especially an influence on the THD+N in high frequency (above 7kHz) and indirectly on the power supply disturbances. With 100µF, you can expect similar THD+N performances like shown in the datasheet. If Cs is lower than 100µF, in high frequency increases, THD+N and disturbances on the power supply rail are less filtered. To the contrary, if Cs is higher than 100µF, those disturbances on the power supply rail are more filtered. Cb has an influence on THD+N in lower frequency, but its function is critical on the final result of PSRR with input grounded in lower frequency. If Cb is lower than 1µF, THD+N increase in lower frequency (see THD+N vs frequency curves) and the PSRR worsens up If Cb is higher than 1µF, the benefit on THD+N in lower frequency is small but the benefit on PSRR is substantial (see PSRR vs. Cb curve : fig.12). Note that Cin has a non-negligible effect on PSRR in lower frequency. Lower is its value, higher is the PSRR (see fig. 13). ■ Pop and Click performance Pop and Click performance is intimately linked with the size of the input capacitor Cin and the bias voltage bypass capacitor Cb. Size of Cin is due to the lower cut-off frequency and PSRR value requested. Size of Cb is due to THD+N and PSRR requested always in lower frequency. 16/19 Moreover, Cb determines the speed that the amplifier turns ON. The slower the speed is, the softer the turn ON noise is. The charge time of Cb is directly proportional to the internal generator resistance 50kΩ. Then, the charge time constant for Cb is τb = 50kΩxCb (s) As Cb is directly connected to the non-inverting input (pin 2 & 3) and if we want to minimize, in amplitude and duration, the output spike on Vout1 (pin 5), Cin must be charged faster than Cb. The charge time constant of Cin is τin = (Rin+Rfeed)xCin (s) Thus we have the relation τin << τb (s) The respect of this relation permits to minimize the pop and click noise. Remark : Minimize Cin and Cb has a benefit on pop and click phenomena but also on cost and size of the application. Example : your target for the -3dB cut off frequency is 100 Hz. With Rin=Rfeed=22 kΩ, Cin=72nF (in fact 82nF or 100nF). With Cb=1µF, if you choose the one of the latest two values of Cin, the pop and click phenomena at power supply ON or standby function ON/OFF will be very small 50 kΩx1µF >> 44kΩx100nF (50ms >> 4.4ms). Increasing Cin value increases the pop and click phenomena to an unpleasant sound at power supply ON and standby function ON/OFF. Why Cs is not important in pop and click consideration ? Hypothesis : • Cs = 100µF • Supply voltage = 5V • Supply voltage internal resistor = 0.1Ω • Supply current of the amplifier Icc = 6mA At power ON of the supply, the supply capacitor is charged through the internal power supply resistor. So, to reach 5V you need about five to ten times the charging time constant of Cs (τs = 0.1xCs (s)). Then, this time equal 50µs to 100µs << τb in the majority of application. TS4902 At power OFF of the supply, Cs is discharged by a constant current Icc. The discharge time from 5V to 0V of Cs is 5Cs t D i s ch C s = -------------- = 83 ms Icc ■ Remark on PSRR measurement conditions What is the PSRR ? The PSRR is the Power Supply Rejection Ratio. Now, we must consider the discharge time of Cb. At power OFF or standby ON, Cb is discharged by a 100kΩ resistor. So the discharge time is about τb Disch ≈ 3xCbx100kΩ (s). In the majority of application, Cb=1µF, then τbDisch≈300ms >> tdischCs. It's a kind of SVR in a determined frequency range. The PSRR of a device is the ratio between the power supply disturbance and the result on the output. We can say that the PSRR is the ability of a device to minimize the impact of power supply disturbances to the output. ■ How to use the PSRR curves (page 7) How do we measure the PSRR ? We have finished a design and we have chosen the components values : Fig. B : PSRR measurement schematic • Rin=Rfeed=22kΩ, Cin=100nF, Cb=1µF The process to obtain the final curve (Cb=100µF, Cin=100nF, Rin=Rfeed=22kΩ) is a simple transfer point by point on each frequency of the curve on fig. 13 to the curve on fig. 12. The measurement result is shown on figure A. 6 Vcc Vripple Vcc Rin 4 Vin- 3 Vin+ - Vout1 5 Vs- + Cin RL Rg 100 Ohms 2 Bypass 1 Standby Av=-1 + Cb Vout2 8 Vs+ Bias GND Now, on fig. 13, we can see the PSRR (input grounded) vs frequency curves. At 217Hz we have a PSRR value of -36dB. In fact, we want a value of about -70dB. So, we need a gain of +34dB ! Now, on fig. 12 we can see the effect of Cb on the PSRR (input grounded) vs. frequency. With Cb=100µF, we can reach the -70dB value. Rfeed TS4902 7 ■ Measurement process: Fig. A : PSRR changes with Cb • Fix the DC voltage supply (Vcc) PSRR (dB) -40 • Fix the AC sinusoidal ripple voltage (Vripple) Vcc = 5, 3.3 & 2.6V Rfeed = 22k, Rin = 22k Rg = 100Ω, RL = 8Ω Tamb = 25°C -30 Cin=100nF Cb=1µF • No bypass capacitor Cs is used The PSRR value for each frequency is : -50 -60 PSRR ( d B ) = 20 x Log 10 Cin=100nF Cb=100µF Remark : The measurement of the RMS voltage is -70 10 R ms ( V r i p pl e ) --------------------------------------------Rms ( Vs + - Vs - ) 100 1000 Frequency (Hz) 10000 100000 not a selective RMS measurement but a full range (2 Hz to 125 kHz) RMS measurement. This means we have: the effective RMS signal + the noise. 17/19 TS4902 PACKAGE MECHANICAL DATA 8 PINS - PLASTIC MICROPACKAGE (SO) s b1 b a1 A a2 C c1 a3 L E e3 D M 5 F 8 1 4 Millimeters Inches Dim. Min. A a1 a2 a3 b b1 C c1 D E e e3 F L M S 18/19 Typ. Max. 0.65 0.35 0.19 0.25 1.75 0.25 1.65 0.85 0.48 0.25 0.5 4.8 5.8 5.0 6.2 0.1 Min. Typ. Max. 0.026 0.014 0.007 0.010 0.069 0.010 0.065 0.033 0.019 0.010 0.020 0.189 0.228 0.197 0.244 0.004 45° (typ.) 1.27 3.81 3.8 0.4 0.050 0.150 4.0 1.27 0.6 0.150 0.016 8° (max.) 0.157 0.050 0.024 TS4902 PACKAGE MECHANICAL DATA 8 PINS - PLASTIC MICROPACKAGE (miniSO) k 0,25mm .010inch GAGEPLANE C SEATING PLANE E1 L1 L c A E A2 A1 4 8 1 e C ccc b D 5 PIN1IDENTIFICA TION Dim. Millimeters Min. A A1 A2 b c D E E1 e L L1 k ccc 0.050 0.780 0.250 0.130 2.900 4.750 2.900 0.400 0d Typ. 0.100 0.860 0.330 0.180 3.000 4.900 3.000 0.650 0.550 0.950 3d Inches Max. Min. 1.100 0.150 0.940 0.400 0.230 3.100 5.050 3.100 0.002 0.031 0.010 0.005 0.114 0.187 0.114 0.700 0.016 6d 0.100 0d Typ. 0.004 0.034 0.013 0.007 0.118 0.193 0.118 0.026 0.022 0.037 3d Max. 0.043 0.006 0.037 0.016 0.009 0.122 0.199 0.122 0.028 6d 0.004 Information furnished is believed to be accurate and reliable. However, STMicroelectronics assumes no responsibility for the consequences of use of such information nor for any infringement of patents or other rights of third parties which may result from its use. No license is granted by implication or otherwise under any patent or patent rights of STMicroelectronics. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all information previously supplied. STMicroelectronics products are not authorized for use as critical components in life support devices or systems without express written approval of STMicroelectronics. © The ST logo is a registered trademark of STMicroelectronics © 2002 STMicroelectronics - Printed in Italy - All Rights Reserved STMicroelectronics GROUP OF COMPANIES Australia - Brazil - Canada - China - Finland - France - Germany - Hong Kong - India - Israel - Italy - Japan - Malaysia Malta - Morocco - Singapore - Spain - Sweden - Switzerland - United Kingdom - United States © http://www.st.com 19/19