The TDA8950 is a high efficiency class-D audio poweramplifier. The typical output power
is 2 × 150 W with a speaker load impedance of 4 Ω.
The TDA8950 is available in the HSOP24 power package and DBS23P power package.
The amplifier operates over a wide supply voltage range from ±12.5 V to ±40 V and
consumes a low quiescent current.
2.Features
n Pin compatible with TDA8920B for both HSOP24 and DBS23P packages
n Symmetrical high operating supply voltage range from ±12.5 V to ±40 V
n Stereo full differential inputs, usableas stereo Single-Ended (SE) or mono Bridge-Tied
Load (BTL) amplifier
n High output power at typical applications:
u SE 2 × 150 W, RL=4Ω (VP = ±37 V)
u SE 2 × 170 W, RL=4Ω (VP = ±39 V)
u SE 2 × 100 W, RL=6Ω (VP = ±37 V)
u BTL 1 × 300 W, RL=8Ω (VP = ±37 V)
n Low noise in BTL due to BD-modulation
n Smooth pop noise-free start-up and switch down
n Zero dead time Pulse Width Modulation (PWM) output switching
n Fixed frequency
n Internal or external clock switching frequency
n High efficiency
n Low quiescent current
n Advanced protection strategy: voltage protection and output current limiting
n Thermal foldback
n Fixed gain of 30 dB in SE and 36 dB in BTL
n Full short-circuit proof across load
3.Applications
n DVD
n Mini and micro receiver
n Home Theater In A Box (HTIAB) system
n High power speaker system
Preliminary data sheetRev. 01 — 9 September 20084 of 39
NXP Semiconductors
7.2 Pin description
Table 3.Pin description
SymbolPinDescription
VSSA118negative analog supply voltage
SGND219signal ground
VDDA320positive analog supply voltage
IN2M421negative audio input for channel 2
IN2P522positive audio input for channel 2
MODE623mode selection input: Standby,Mute or Operating
OSC71oscillator frequency adjustment or tracking input
IN1P82positive audio input for channel 1
IN1M93negative audio input for channel 1
n.c.104not connected
n.c.115not connected
n.c.126not connected
PROT137decoupling capacitor for protection (OCP)
VDDP1148positive power supply voltage for channel 1
BOOT1159bootstrap capacitor for channel 1
OUT11610PWM output from channel 1
VSSP11711negative power supply voltage for channel 1
STABI1812decoupling of internal stabilizer for logic supply
n.c.19-not connected
VSSP22013negative power supply voltage for channel 2
OUT22114PWM output from channel 2
BOOT22215bootstrap capacitor for channel 2
VDDP22316positive power supply voltage for channel 2
VSSD2417negative digital supply voltage
TDA8950
2 × 150 W class-D power amplifier
TDA8950TH TDA8950J
mode
8.Functional description
8.1 General
The TDA8950 is a two-channel audio power amplifier using class-D technology.
The audio input signal is converted into a digital pulse width modulated signal via an
analog input stage and PWM modulator, see Figure 1. To enable the output power
transistors to be driven, this digital PWM signal is applied to a control and handshake
block and driver circuits for both the high side and low side. In this way a level shift is
performed from the low power digital PWM signal (at logic levels) to a high power PWM
signal that switches between the main supply lines.
A 2nd-order low-pass filter converts the PWM signal to an analog audio signal across the
loudspeakers.
Preliminary data sheetRev. 01 — 9 September 20085 of 39
NXP Semiconductors
The TDA8950 one-chip class-D amplifier contains high power switches, drivers, timing
and handshaking between the power switches and some control logic. Also an advanced
protection strategy is implemented which contains several voltage protections,
temperature protections and a maximum current protection to secure maximum system
robustness.
The two audio channels of the TDA8950 each contain a PWM modulator, an analog
feedback loop and a differential input stage. It also contains circuits common to both
channels such as the oscillator, all reference sources, the mode interface and a digital
timing manager.
The TDA8950 contains two independent amplifier channels with high output power, high
efficiency, low distortion and low quiescent current. The amplifier channels can be
connected in the following configurations:
• Mono Bridge-Tied Load (BTL) amplifier
• Stereo Single-Ended (SE) amplifiers
The amplifier system can be switched to one of three operating modes by pin MODE:
TDA8950
2 × 150 W class-D power amplifier
• Standby mode; with a very low supply current
• Mute mode; the amplifiers are operational, but the audio signal at the output is
suppressed by disabling the VI-converter input stages
• Operating mode; the amplifiers are fully operational
To ensure pop noise-free start-up, the DC output offset voltage is applied gradually to the
output at a level between Mute mode and Operating mode levels. The bias current setting
of the VI converters is related to the voltage on pin MODE; in Mute mode the bias current
setting of the VI converters is zero (VI converters disabled) and in Operating mode the
bias current is at maximum. The time constant required to apply the DC output offset
voltage gradually between Mute and Operating mode levels can be generated via an
RC-network on pin MODE. An example of a switching circuit for driving pin MODE is
illustrated in Figure 4. If the capacitor C is left out of the application, the voltage on pin
MODE will be applied with a much smaller time-constant, which might result in audible
pop noises during start-up (depending on DC output offset voltage and loudspeaker
used).
Preliminary data sheetRev. 01 — 9 September 20086 of 39
NXP Semiconductors
In order to fully charge the coupling capacitors at the inputs, the amplifier will remain
automatically in the Mute mode before switching to the Operating mode. A complete
overview of the start-up timing is given in Figure 5.
V
MODE
TDA8950
2 × 150 W class-D power amplifier
audio output
(1)
modulated PWM
50 %
duty cycle
2.2 V < V
2.2 V < V
> 4.2 V
< 3 V
MODE
0 V (SGND)
V
MODE
> 4.2 V
< 3 V
MODE
standby
mute
100 ms
50 %
duty cycle
mute
(1)
50 ms
operating
> 350 ms
audio output
modulated PWM
operating
time
0 V (SGND)
standby
100 ms
50 ms
> 350 ms
time
001aah657
(1) First1⁄4 pulse down.
Upper diagram: When switching from standby to mute, there is a delay of approximately 100 ms
before the output starts switching. The audio signal is available after V
has been set to
MODE
operating, but not earlier than 150 ms after switching to mute. For pop noise-free start-up it is
recommended that the time constant applied to pin MODE is at least 350 ms for the transition
between mute and operating.
Lower diagram: When switchingdirectlyfromstandbyto operating, there is a first delay of 100 ms
before the outputs starts switching. The audio signal is available after a second delay of 50 ms. For
pop noise-free start-up it is recommended that the time constant applied to pin MODE is at least
500 ms for the transition between standby and operating.
Preliminary data sheetRev. 01 — 9 September 20087 of 39
NXP Semiconductors
8.2 Pulse width modulation frequency
The output signal of the amplifier is a PWM signal with a carrier frequency that typically
lies between 300 kHz and 400 kHz. Using a 2nd-order LC demodulation filter in the
application results in an analog audio signal across the loudspeaker. The carrier
frequency is determined by an external resistor R
pin VSSA. An optimal setting for the carrier frequency is between 300 kHz and 400 kHz.
Using an external resistor of 30 kΩ on pin OSC, the carrier frequency is set to 345 kHz.
For more details see Table 8.
If two or more class-D amplifiers are used in the same audio application, it is
recommended that all devices operate at the same switching frequency by using an
external clock circuit.
Due to an internal clock divider:
• The external applied clock frequency must have the double frequency of the output
• The duty cycle of the external clock is not critical for product performance.
PWM frequency.
TDA8950
2 × 150 W class-D power amplifier
, connected between pin OSC and
OSC
8.3 Protections
The following protections are included in TDA8950:
The reaction of the device to the different fault conditions differs per protection.
8.3.1 Thermal protection
In the TDA8950 an advanced thermal protection strategy is implemented. It consists of a
TFB function that gradually reduces the out put power within a certain temperature range.
When temperature is still rising an OTP is implemented which shuts down the device
completely.
8.3.1.1 Thermal FoldBack (TFB)
If the junction temperature Tj exceeds a defined threshold value, the gain is gradually
reduced. This will result in a smaller output signal and less dissipation. Eventually the
temperature will stabilize.
Preliminary data sheetRev. 01 — 9 September 20088 of 39
NXP Semiconductors
TDA8950
2 × 150 W class-D power amplifier
TFB is specified at the temperature value T
reduced with 6 dB. The range of the TFB is:
T
act(th_fold)
For the TDA8950 the value of T
− 5 °C < T
act(th_fold)
< T
act(th_prot)
act(th_fold)
8.3.1.2 OverTemperature Protection (OTP)
If, despite the TFB function, the junction temperature Tj of the TDA8950 continues rising
and exceeds the threshold T
act(th_prot)
amplifier resumes switching approximately 100 ms after the temperature drops below
T
act(th_prot)
.
In Figure 6 the thermal behavior is visualized.
Gain
(dB)
30 dB
24 dB
0 dB
(T
act(th_fold)
12 3
act(th_fold)
where the closed loop voltage gain is
.
is about +153 °C. For more details see: Table 7.
the amplifier will shutdown immediately. The
(°C) T
− 5°C)
T
act(th_fold)
act(th_prot)
T
j
001aah656
(1) Duty cycle of PWM output modulated according audio input signal.
(2) Duty cycle of PWM output reduced due to TFB.
(3) Amplifier is switched off due to OTP.
Fig 6.Behavior of TFB and OTP
8.3.2 OverCurrent Protection (OCP)
If a short-circuit is applied to one of the demodulated outputs of the amplifier, the OCP will
detect this. If the output current exceeds the maximum of 9.2 A, it is automatically limited
to its maximum value by the OCP protection circuit. The amplifier outputs remain
switching (the amplifier is NOT shut-down completely). If the active current limiting
continues longer than time τ, the TDA8950 shuts down. Activation of current limiting and
the triggering of the OCP are observed at pin PROT.
The amplifier can distinguish between an impedance drop of the loudspeaker and a
low-ohmic short-circuit across the load. In the TDA8950 the impedance threshold (Zth)
depends on the supply voltage used.
When a short-circuit is made across the load, causing the impedance to drop below the
threshold level (<Zth), the amplifier is switched off completely and, after a time of 100 ms,
it will try to restart. If the short-circuit condition is still present after this time, the cycle will
be repeated. The average dissipation will be low because of this low duty cycle.
Preliminary data sheetRev. 01 — 9 September 20089 of 39
NXP Semiconductors
Should there be an impedance drop (e.g. due to dynamic behavior of the loudspeaker) the
same protection will be activated. The maximum output current will again be limited to
9.2 A, but the amplifier will not switch-off completely (thus preventing audio holes from
occurring).
The result will be a clipping output signal.
See Section 13.7 for more information on this maximum output current limiting feature.
8.3.3 Window Protection (WP)
The WP checks the conditions at the output terminals of the power stage and is activated:
• During the start-up sequence, when pin MODE is switched from standby to mute. In
• When the amplifier is completely shut-down due to activation of the OCP because a
TDA8950
2 × 150 W class-D power amplifier
the event of a short-circuit at one of the output terminals to pin VDDPn or pin VSSPn
the start-up procedure is interrupted and the TDA8950 waits until the short-circuit to
the supply lines has been removed. Because the test is done before enabling the
power stages, no large currents will flow in an event of short-circuit.
short-circuit to one of the supply lines is made, then during restart (after 100 ms) the
WP will be activated. As a result the amplifier will not start-up until the short-circuit to
the supply lines is removed.
8.3.4 Supply voltage protections
If the supply voltage drops below minimum supply voltage, the UVP circuit is activated and
the system will shutdown correctly. If the internal clock is used, this switch-off will be silent
and without pop noise. When the supply voltage rises above the threshold level, the
system is restarted again after 100 ms.
If the supply voltage exceeds maximum supply voltage, the OVP circuit is activated and
the power stages will shutdown. When the supply voltage drops below the threshold level,
the system is restarted again after 100 ms.
An additional UBP circuit compares the positive analog (voltage on pin VDDA) and the
negative analog (voltage on pin VSSA) supply voltage and is triggered if the voltage
difference exceeds a factor of two.
When the supply voltage difference drops below the threshold level, the system is
restarted again after 100 ms.
Example: With a symmetrical supply of ±30 V, the protection circuit will be triggered if the
unbalance exceeds approximately 15 V. See Section 13.7.
In Table 4 an overview is given of all protections and their respective effects on the output
signal.
Preliminary data sheetRev. 01 — 9 September 200810 of 39
NXP Semiconductors
Table 4.Overview of TDA8950 protections
Protection name Complete
TFB
OTPYNYN
OCPY
WPN
UVPYNYN
OVPYNYN
UBPYNYN
[1] Amplifier gain will depend on junction temperature and heatsink size.
[2] Only complete shutdown of amplifier if short-circuit impedance is below threshold of 1 Ω. In all other cases
[3] Fault condition detected during (every) transition between standby-to-mute and during restart after
8.4 Differential audio inputs
Restart directlyRestart after
shutdown
[1]
current limiting results in clipping of the output signal.
activation of OCP (short-circuit to one of the supply lines).
NNNN
[2]
[3]
[2]
N
YNN
TDA8950
2 × 150 W class-D power amplifier
Pin PROT
100 ms
[2]
Y
detection
Y
Fora high common mode rejection ratio and a maximum of flexibility in the application, the
audio inputs are fully differential.
There are two possibilities:
• For stereo operation it is advised to use the inputs in anti phase and also to connect
the speakers in anti phase (to avoid acoustical phase differences). This construction
has several advantages:
– The peak current in the power supply is minimized
– The supply pumping effect is minimized, especially at low audio frequencies
• For mono BTL operation it is required that the inputs are connected in anti parallel.
The output of one of the channels is inverted and the speaker load is now connected
between the two outputs of the TDA8950. In principle the output power to the speaker
can be significantly boosted to two times the output power in single ended stereo.
The input configuration for a mono BTL application is illustrated in Figure 7.
IN1P
IN1M
V
in
IN2P
IN2M
OUT1
SGND
OUT2
power stage
mbl466
Fig 7. Input configuration for mono BTL application
Preliminary data sheetRev. 01 — 9 September 200812 of 39
thermal resistance from junction to ambientin free air40K/W
thermal resistance from junction to case1.1K/W
NXP Semiconductors
TDA8950
2 × 150 W class-D power amplifier
11. Static characteristics
Table 7.Static characteristics
VP =±35 V; f
SymbolParameterConditionsMinTypMaxUnit
Supply
V
P
V
P(ovp)
V
P(uvp)
I
q(tot)
I
stb
Mode select input; pin MODE
V
MODE
I
I
Audio inputs; pins IN1M, IN1P, IN2P and IN2M
V
I
Amplifier outputs; pins OUT1 and OUT2
V
O(offset)
Stabilizer output; pin STABI
V
O(STABI)
Temperature protection
T
act(th_prot)
T
act(th_fold)
= 345 kHz; T
osc
supply voltage
overvoltage protection supply voltagenon-operating; VDD− V
undervoltage protection supply voltage VDD− V
total quiescent currentOperating mode; no load; no
= 25°C; unless otherwise specified.
amb
SS
SS
[1]
±12.5±35±40V
85-90V
20-25V
- 5075mA
filter; no RC-snubbernetwork
connected
standby current-480600µA
voltage on pin MODEreferenced to SGND
Standby mode
Mute mode
Operating mode
[2]
0-6.0V
[2][3]
0-0.8V
[2][3]
2.2-3.0V
[2][3]
4.2-6.0V
input currentVI = 5.5 V-110150µA
input voltageDC input
[2]
-0-V
output offset voltageSE; mute--±15mV
[4]
SE; operating
--±150mV
BTL; mute--±21mV
[4]
BTL; operating
output voltage on pin STABImute and operating; with
--±210mV
9.39.810.3V
respect to VSSP1
thermal protection activation
-154-°C
temperature
thermal foldback activation
temperature
closed loop SE voltage gain
reduced with 6 dB
[5]
-153-°C
[1] The circuit is DC adjusted at VP= ±12.5 V to ±42.5 V.
[2] With respect to SGND (0 V).
[3] The transition between Standby and Mute mode has hysteresis, while the slope of the transition between Mute and Operating mode is
determined by the time-constant of the RC-network on pin MODE; see Figure 8.
[4] DC output offset voltage is gradually applied to the output during the transition between the Mute and Operating modes. The slope
caused by any DC output offset is determined by the time-constant of the RC-network on pin MODE.
[1] RsL is the series resistance of inductor of low-pass LC filter in the application.
[2] Output power is measured indirectly; based on R
[3] THD is measured in a bandwidth of 22 Hz to 20 kHz, using AES17 20 kHz brickwall filter. Maximum limit is not guaranteed100 % tested.
[4] V
[5] B = 22 Hz to 20 kHz, using AES17 20 kHz brickwall filter.
[6] B = 22 Hz to 22 kHz, using AES17 20 kHz brickwall filter; independent of Rs.
[7] Po = 1 W; Rs = 0 Ω; fi = 1 kHz.
[8] Vi = V
[9] Leads and bond wires included.
Preliminary data sheetRev. 01 — 9 September 200815 of 39
ripple
= V
i(max)
ripple(max)
= 2 V (p-p); Rs = 0 Ω. Measured independently between VDDPn and SGND and between VSSPn and SGND.
= 1 V (RMS); fi = 1 kHz.
measurement. See also Section 13.3.
DSon
NXP Semiconductors
12.3 Mono BTL application characteristics
TDA8950
2 × 150 W class-D power amplifier
Table 10. Dynamic characteristics
VP =±35 V; RL = 8Ω; fi = 1 kHz; f
= 345 kHz; RsL < 0.1
osc
[1]
Ω
; T
= 25°C; unless otherwise
amb
specified.
Symbol ParameterConditionsMinTypMaxUnit
P
o
output powerL = 22 µH; C = 680 nF;
T
=85°C; RL=8Ω
j
[2]
THD = 10 %; VP= ±39 V-340-W
= ±37 V-300-W
P
= ±37 V-200-W
P
[3]
-0.05-%
[3]
-0.05-%
-36-dB
THDtotal harmonic
distortion
G
v(cl)
closed-loop voltage
THD = 10 %; V
THD = 0.5 %; V
= 1 W; fi= 1 kHz
P
o
= 1 W; fi= 6 kHz
P
o
gain
SVRRsupply voltage ripple
rejection
between pin VDDPn and
SGND
operating; f
operating; f
mute; f
standby; f
= 100 Hz
i
= 1 kHz
i
= 100 Hz
i
= 100 Hz
i
[4]
-80-dB
[4]
-80-dB
[4]
-95-dB
[4]
-120-dB
between pin VSSPn and
SGND
operating; f
operating; f
mute; f
standby; f
Z
i
input impedancemeasured between the input
= 100 Hz
i
= 1 kHz
i
= 100 Hz
i
= 100 Hz
i
[4]
-75-dB
[4]
-75-dB
[4]
-90-dB
[4]
-130-dB
4563-kΩ
pins and SGND
V
α
n(o)
mute
output noise voltageoperating; Rs=0Ω
mute attenuationfi= 1 kHz; Vi= 2 V (RMS)
CMRRcommon mode
mute
= 1 V (RMS)-75-dB
V
i(CM)
[5]
-190-µV
[6]
-45-µV
[7]
-82-dB
rejection ratio
[1] RsL is the series resistance of inductor of low-pass LC filter in the application.
[2] Output power is measured indirectly; based on R
[3] Total harmonic distortion is measured in a bandwidth of 22 Hz to 20 kHz, using an AES17 20 kHz brickwall
filter. Maximum limit is guaranteed but may not be 100 % tested.
[4] V
[5] B = 22 Hz to 20 kHz, using an AES17 20 kHz brickwall filter. Low noise due to BD modulation.
[6] B = 22 Hz to 20 kHz, using an AES17 20 kHz brickwall filter; independent of Rs.
[7] Vi = V
Preliminary data sheetRev. 01 — 9 September 200816 of 39
ripple
= V
i(max)
= 2 V (p-p); Rs = 0 Ω.
ripple(max)
= 1 V (RMS); fi = 1 kHz.
measurement. See also Section 13.3.
DSon
NXP Semiconductors
13. Application information
13.1 Mono BTL application
When using the power amplifier in a mono BTL application the inputs of both channels
must be connected in parallel and the phase of one of the inputs must be inverted (see
Figure 7). In principle the loudspeaker can be connected between the outputs of the two
single-ended demodulation filters.
13.2 Pin MODE
For pop noise-free start-up, an RC time-constant must be applied on pin MODE. The
bias-current setting of the VI-converter input is directly related to the voltage on pin
MODE. In turn the bias-current setting of the VI converters is directly related to the DC
output offset voltage. Thus a slow dV/dt on pin MODE results in a slow dV/dt for the
DC output offset voltage, resulting in pop noise-free start-up. A time-constant of 500 ms is
sufficient to guarantee pop noise-free start-up (see also Figure 4, 5 and 8).
: single-sided supply voltage (or 0.5 × (VDD + |VSS|))
P
: output power at the onset of clipping
o(0.5 %)
L
DSon hs()RsL
1t
DSon hs()RsL
min
0.5 f
×–()×
DSon
V
1t
–()××
osc
2R
P
L
min
of power stage output DMOS (temperature dependent)
0.5×f
2
osc
(1)
(2)
Note that I
should be below 9.2 A (Section 8.3.2). I
o(peak)
o(peak)M
is the sum of the current
through the load and the ripple current. The value of the ripple current is dependent on the
coil inductance and voltage drop over the coil.
: single-sided supply voltage (or 0.5 × (VDD + |VSS|))
: output power at the onset of clipping
o(peak)M
should be below 9.2 A (Section 8.3.2). I
o(peak)
is the sum of the current
through the load and the ripple current. The value of the ripple current is dependent on the
coil inductance and voltage drop over the coil.
13.4 External clock
For duty cycle independent operation of the device, the external clock input frequency is
internally divided by two. This implies that the frequency of the external clock is as twice
as high as the internal clock (typical 2 × 345 kHz = 690 kHz).
If two or more class-D amplifiers are used it is recommended that all devices run at the
same switching frequency. This can be realized by connecting all OCS pins together and
feeding them from an external oscillator. When applying an external oscillator, it is
necessary to force pin OSC to a DC level above SGND. The internal oscillator is disabled
and the PWM modulator will switch with half the externally applied frequency.
The internal oscillator requires an external resistor R
ext(OSC)
pin OSC and PIN VSSA.
The noise contribution of the internal oscillator is supply voltage dependent. An external
low noise oscillator is recommended for low noise applications running at high supply
voltage.
Preliminary data sheetRev. 01 — 9 September 200818 of 39
and capacitor C
OSC
between
NXP Semiconductors
13.5 Noise
Noise should be measured using a high order low-pass filter with a cut-off frequency of
20 kHz. The standard audio band pass filters, used in audio analyzers, do not suppress
the residue of the carrier frequency sufficiently to ensure a reliable measurement of the
audible noise. Noise measurements should preferably be carried out using AES 17
(‘brickwall’) filters or an audio precision AUX 0025 filter, which was designed specifically
for measuring class-D switching amplifiers.
13.6 Heatsink requirements
In many applications it may be necessary to connect an external heatsink to the
TDA8950.
Equation 5 shows the relationship between the maximum power dissipation, before
activation of the TFB, and the total thermal resistance from junction to ambient
R
th j a–()
TjT
=
----------------------- -
P
–
diss
amb
TDA8950
2 × 150 W class-D power amplifier
(5)
Power dissipation (P
) is determined by the efficiency of the TDA8950. The efficiency
diss
measured as a function of output power is given in Figure 21. The power dissipation can
be derived as a function of output power (Figure 20).
T
amb
mbl469
(°C)
30
P
(1) R
(2) R
(3) R
(4) R
(5) R
th(j-a)
th(j-a)
th(j-a)
th(j-a)
th(j-a)
= 5 K/W.
= 10 K/W.
= 15 K/W.
= 20 K/W.
= 35 K/W.
(W)
20
10
0
020100406080
(1)
(2)
(3)
(4)
(5)
Fig 9.De-rating curves for power dissipation as a function of maximum ambient
temperature
In the following example, a heatsink calculation is made for an 8 Ω BTL application with a
±35 V supply:
Preliminary data sheetRev. 01 — 9 September 200819 of 39
NXP Semiconductors
The audio signal has a crest factor of 10 (the ratio between peak power and average
power (20 dB)), this means that the average output power is1⁄10 of the peak power.
In this case, the peak RMS output power level would be the 0.5 % THD level, i.e. 220 W.
The average power would then be1⁄10× 220W=22W.
The dissipated power at an output power of 22 W is approximately 5 W.
TDA8950
2 × 150 W class-D power amplifier
When the maximum expected ambient temperature is 85 °C, the total R
become
R
th(j-a)
R
R
140 85–()
-------------------------
5
= R
th(j-c)
th(c-h)
+ R
th(j-c)
= 1.1 K/W
= 0.5 K/W to 1 K/W (dependent on mounting)
th(c-h)
11 K/W=
+ R
So the thermal resistance between heatsink and ambient temperature is:
R
= 11 − (1.1 + 1) = 8.9 K/W
th(h-a)
The de-rating curves (given for several values of R
maximum junction temperature Tj= 150 °C is taken into account. From Figure 9 the
maximum allowable power dissipation for a given heatsink size can be derived or the
required heatsink size can be determined at a required dissipation level.
13.7 Output current limiting
To guarantee the robustness of the TDA8950, the maximum output current that can be
delivered by the output stage is limited. An advanced OverCurrent Protection (OCP) is
included for each output power switch.
When the current flowing through any of the power switches exceeds the defined internal
threshold current of 9.2 A (e.g. in case of a short-circuit to the supply lines or a
short-circuit across the load), the maximum output current of the amplifier will be
regulated to 9.2 A.
th(h-a)
th(j-a)
) are illustrated in Figure 9. A
th(j-a)
would then
The TDA8950 amplifier can distinguish between a low-ohmic short-circuit condition and
other over current conditions like dynamic impedance drops of the loudspeakers used.
The impedance threshold (Zth) depends on the supply voltage used.
Depending on the impedance of the short-circuit, the amplifier will react as follows:
• Short-circuit impedance > Z
: The maximum output current of the amplifier is
th
regulated to 9.2 A, but the amplifier will not shut-down its PWM outputs. Effectively
this results in a clipping output signal across the load (behavior is very similar to
voltage clipping).
• Short-circuit impedance < Z
: The amplifier will limit the maximum output current to
th
9.2 A and at the same time the capacitor on pin PROT is discharged. When the
voltage across this capacitor drops below an internal threshold voltage, the amplifier
will shutdown completely and an internal timer will be started.
A typical value for the capacitor on pin PROT is 220 pF. After a fixed time of 100 ms the
amplifier is switched on again. If the requested output current is still too high, the amplifier
will switch-off again. Thus the amplifier will try to switch to the Operating mode every
Preliminary data sheetRev. 01 — 9 September 200820 of 39
NXP Semiconductors
100 ms. The averagedissipation will be low in this situation because of this low duty cycle.
If the over current condition is removed the amplifier will remain in Operating mode after
being restarted. In this way the TDA8950 amplifier is fully protected against short-circuit
conditions while at the same time so-called audio holes, as a result of loudspeaker
impedance drops, are eliminated.
13.8 Pumping effects
In a typical stereo half-bridge SE application the TDA8950 is supplied by a symmetrical
voltage (e.g. VDD= +35 V and VSS= −35 V). When the amplifier is used in an SE
configuration, a ‘pumping effect’ can occur. During one switching interval, energy is taken
from one supply (e.g. VDD), while a part of that energy is returned to the other supply line
(e.g. VSS) and vice versa. When the voltage supply source cannot sink energy,the voltage
across the output capacitors of that voltage supply source will increase and the supply
voltage is pumped to higher levels. The voltage increase caused by the pumping effect
depends on:
• Speaker impedance
• Supply voltage
• Audio signal frequency
• Value of decoupling capacitors on supply lines
• Source and sink currents of other channels
TDA8950
2 × 150 W class-D power amplifier
When applying the TDA8950, measures must be taken within the application to minimize
the pumping effect and prevent malfunctions of either the audio amplifier and/or the
voltage supply source. Amplifier malfunction due to the pumping effect can cause
triggering of the UVP, OVP or UBP.
The best remedy against pumping effects is to use the TDA8950 in a mono full-bridge
application or, in the case of stereo half-bridge applications, adapt the power supply (e.g.
increase supply decoupling capacitors).
13.9 Application schematics
Notes for the application schematic:
• A solid ground plane connected to V
prevent emission.
• 100 nF capacitors must be placed as close as possible to the power supply pins of the
TDA8950.
• The internal heat spreader of the TDA8950 is internally connected to V
• The external heatsink must be connected to the ground plane.
• Use a thermally conductive, electrically non-conductive, Sil-Pad between the backside
of the TDA8950 and a small external heatsink.
• The differential inputs enable the best system level audio performance with
unbalanced signal sources. In case of hum, due to floating inputs, connect the
shielding or source ground to the amplifier ground. Jumpers J1 and J2 are open on
set level and are closed on the stand-alone demo board.
• Minimum total required capacitance per power supply line is 3300 µF.
Preliminary data sheetRev. 01 — 9 September 200822 of 39
xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxx x x x xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xx xx
xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxx x x
xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx xxx
NXP Semiconductors
VDDP
GND
VSSP
IN1
IN2
Rvdda
Cvddp
470 µF
Cvssp
470 µF
Cin
+
470 nF
Cin
−
470 nF
Cin
−
470 nF
Cin
+
470 nF
IN1P
IN1N
SGND
IN2P
IN2N
Cvp
22 µF
Rvssa
8
9
19
22
21
2018
VDDA
Cvdda
100 nF
VDDAVSSA
10 Ω
10 Ω
n.c.
VDDA
VDDP
SINGLE ENDED
OUTPUT FILTER VALUES
LOADLlcClc
2 Ω - 3 Ω10 µH1000 nF
mode
Rosc
30 kΩ
OSC
1
PROT
Cprot
100 nF
V
control
MODE
23
SSA
177
VSSD
Cvddp
100 nF
Cvddp
100 nF
VDDP
VDDP1
8
16
VDDP2
VDDP
Cvp
100 nF
Cvp
100 nF
VSSP
VSSP1
11
13
VSSP2
VSSP
Cvssp
100 nF
Cvssp
100 nF
10
15
14
9
OUT1
BOOT1
BOOT2
OUT2
Rsn
10 Ω
Cbo
15 nF
Cbo
15 nF
Rsn
10 Ω
VSSA
VSSP
VSSA
n.c.
n.c.
4
6
5
TDA8950J
12
VSSA
Cvssa
100 nF
STABI
V
SSP
Cstab
470 nF
3 Ω - 6 Ω15 µH680 nF
4 Ω - 8 Ω22 µH470 nF
VDDP
Csn
220 pF
Csn
220 pF
VSSP
Llc
Llc
VDDP
Csn
220 pF
Csn
220 pF
VSSP
Clc
Clc
Rzo
22 Ω
Czo
100 nF
Rzo
22 Ω
Czo
100 nF
+
−
−
+
001aai420
2 × 150 W class-D power amplifier
TDA8950
Fig 10. Simplified application diagram
NXP Semiconductors
13.10 Layout and grounding
To obtain a high-level system performance, certain grounding techniques are essential.
The input reference grounds have to be tied with their respective source grounds and
must have separate tracks from the power ground tracks. This will prevent the large
(output) signal currents from interfering with the small AC input signals. The small-signal
ground tracks should be physically located as far as possible from the power ground
tracks. Supply and output tracks should be as wide as possible for delivering maximum
output power.
Preliminary data sheetRev. 01 — 9 September 200834 of 39
NXP Semiconductors
15. Soldering of SMD packages
This text provides a very brief insight into a complex technology. A more in-depth account
of soldering ICs can be found in Application Note
soldering description”
15.1 Introduction to soldering
Soldering is one of the most common methods through which packages are attached to
Printed Circuit Boards (PCBs), to form electrical circuits. The soldered joint provides both
the mechanical and the electrical connection. There is no single soldering method that is
ideal for all IC packages. Wave soldering is often preferred when through-hole and
Surface Mount Devices (SMDs) are mixed on one printed wiring board; however, it is not
suitable for fine pitch SMDs. Reflow soldering is ideal for the small pitches and high
densities that come with increased miniaturization.
15.2 Wave and reflow soldering
Wave soldering is a joining technology in which the joints are made by solder coming from
a standing wave of liquid solder. The wave soldering process is suitable for the following:
.
TDA8950
2 × 150 W class-D power amplifier
AN10365 “Surface mount reflow
• Through-hole components
• Leaded or leadless SMDs, which are glued to the surface of the printed circuit board
Not all SMDs can be wave soldered. Packages with solder balls, and some leadless
packages which have solder lands underneath the body, cannot be wave soldered. Also,
leaded SMDs with leads having a pitch smaller than ~0.6 mm cannot be wave soldered,
due to an increased probability of bridging.
The reflow soldering process involves applying solder paste to a board, followed by
component placement and exposure to a temperature profile. Leaded packages,
packages with solder balls, and leadless packages are all reflow solderable.
Key characteristics in both wave and reflow soldering are:
• Board specifications, including the board finish, solder masks and vias
• Package footprints, including solder thieves and orientation
• The moisture sensitivity level of the packages
• Package placement
• Inspection and repair
• Lead-free soldering versus SnPb soldering
15.3 Wave soldering
Key characteristics in wave soldering are:
• Process issues, such as application of adhesive and flux, clinching of leads, board
transport, the solder wave parameters, and the time during which components are
exposed to the wave
• Solder bath specifications, including temperature and impurities
Preliminary data sheetRev. 01 — 9 September 200835 of 39
NXP Semiconductors
15.4 Reflow soldering
Key characteristics in reflow soldering are:
• Lead-free versus SnPb soldering; note that a lead-free reflow process usually leads to
• Solder paste printing issues including smearing, release, and adjusting the process
• Reflow temperature profile; this profile includes preheat, reflow (in which the board is
Table 11. SnPb eutectic process (from J-STD-020C)
Package thickness (mm)Package reflow temperature (°C)
< 2.5235220
≥ 2.5220220
TDA8950
2 × 150 W class-D power amplifier
higher minimum peak temperatures (see Figure 31) than a SnPb process, thus
reducing the process window
window for a mix of large and small components on one board
heated to the peak temperature) and cooling down. It is imperative that the peak
temperature is high enough forthe solder to make reliable solder joints (a solder paste
characteristic). In addition, the peak temperature must be low enough that the
packages and/or boards are not damaged. The peak temperature of the package
depends on package thickness and volume and is classified in accordance with
Table 11 and 12
Volume (mm3)
< 350≥ 350
Table 12. Lead-free process (from J-STD-020C)
Package thickness (mm)Package reflow temperature (°C)
Volume (mm3)
< 350350 to 2000> 2000
< 1.6260260260
1.6 to 2.5260250245
> 2.5250245245
Moisture sensitivity precautions, as indicated on the packing, must be respected at all
times.
Studies have shown that small packages reach higher temperatures during reflow
soldering, see Figure 31.
Preliminary data sheetRev. 01 — 9 September 200837 of 39
NXP Semiconductors
17. Legal information
17.1Data sheet status
TDA8950
2 × 150 W class-D power amplifier
Document status
Objective [short] data sheetDevelopmentThis document contains data from the objective specification for product development.
Preliminary [short] data sheet QualificationThis document contains data from the preliminary specification.
Product [short] data sheetProductionThis document contains the product specification.
[1] Please consult the most recently issued document before initiating or completing a design.
[2] The term ‘short data sheet’ is explained in section “Definitions”.
[3] The product status of device(s)described in this document mayhave changedsince this document was publishedand maydiffer incase of multipledevices. Thelatest product status
information is available on the Internet at URL
[1][2]
Product status
17.2Definitions
Draft — The document is a draft version only. The content is still under
internal review and subject to formal approval, which may result in
modifications or additions. NXP Semiconductors does not give any
representations or warranties as to the accuracy or completeness of
information includedherein and shall have no liabilityfor the consequencesof
use of such information.
Short data sheet — A short data sheet is an extract from a full data sheet
with thesame product type number(s)and title. Ashort datasheet is intended
for quickreference only and shouldnot be relied upon to containdetailed and
full information. For detailed and full information see the relevant full data
sheet, which is available on request via the local NXP Semiconductors sales
office. In case of any inconsistency or conflict with the short data sheet, the
full data sheet shall prevail.
17.3Disclaimers
General — Information in this document is believed to be accurate and
reliable. However, NXPSemiconductors does not give anyrepresentations or
warranties, expressed or implied, as tothe accuracy or completeness ofsuch
information and shall have no liability for the consequences of use of such
information.
Right to make changes — NXP Semiconductors reserves the right tomake
changes to information published in this document, including without
limitation specifications and product descriptions, at any time and without
notice. This documentsupersedes and replaces all informationsuppliedprior
to the publication hereof.
Suitability for use — NXP Semiconductors products are not designed,
authorized or warranted to be suitable for use in medical, military, aircraft,
space or life support equipment, nor in applications where failure or
malfunction of an NXP Semiconductors product can reasonably be expected
to result in personal injury, death or severe property or environmental
[3]
http://www.nxp.com.
Definition
damage. NXP Semiconductors accepts no liability for inclusion and/or use of
NXP Semiconductors products in such equipment or applications and
therefore such inclusion and/or use is at the customer’s own risk.
Applications — Applications that are described herein for any of these
products are for illustrative purposes only. NXP Semiconductors makes no
representation or warranty that such applications will be suitable for the
specified use without further testing or modification.
Limiting values — Stress above one or more limiting values (as defined in
the Absolute MaximumRatings System of IEC 60134)may cause permanent
damage tothedevice. Limiting valuesare stress ratingsonly and operation of
the device at these or any other conditions above those given in the
Characteristics sections of this document is not implied. Exposure to limiting
values for extended periods may affect device reliability.
Terms and conditions of sale — NXP Semiconductors products are sold
subject to the general terms and conditions of commercialsale,as published
at
http://www.nxp.com/profile/terms, including those pertaining to warranty,
intellectual property rights infringement and limitation of liability, unless
explicitly otherwise agreed to in writing by NXP Semiconductors. In case of
any inconsistency or conflict between information in this document and such
terms and conditions, the latter will prevail.
No offer to sell or license — Nothing in this document may be interpreted
or construed as an offer to sell products that is open for acceptance or the
grant, conveyance or implication of anylicense under any copyrights, patents
or other industrial or intellectual property rights.
Quick reference data — The Quick reference data is an extract of the
product data given in the Limiting values and Characteristics sections of this
document, and as such is not complete, exhaustive or legally binding.
17.4Trademarks
Notice: Allreferenced brands,product names, service namesand trademarks
are the property of their respective owners.
18. Contact information
For more information, please visit: http://www.nxp.com
For sales office addresses, please send an email to: salesaddresses@nxp.com
Please be aware that important notices concerning this document and the product(s)
described herein, have been included in section ‘Legal information’.