Analog, Active Crossover Circuit For Two-Way Loudspeakers: TI Precision Designs: Verified Design

John Caldwell
TI Precision Designs: Verified Design

Analog, Active Crossover Circuit for Two-Way
Loudspeakers
TI Precision Designs Circuit Description

TI Precision Designs are analog solutions created by This is an analog active crossover solution for two-
TI’s analog experts. Verified Designs offer the theory, way loudspeakers. The woofer signal pathway
component selection, simulation, complete PCB includes a low-pass shelving circuit for baffle step
th
schematic & layout, bill of materials, and measured compensation and a 4 -order Linkwitz-Riley low-pass
th
performance of useful circuits. Circuit modifications filter. The tweeter signal pathway includes a 4 -order
rd
that help to meet alternate design goals are also Linkwitz-Riley high-pass filter, 3 -order all-pass filter
discussed. for time alignment, output attenuation and buffering.
Design Resources
Ask The Analog Experts
Design Archive All Design files WEBENCH® Design Center
TINA-TI™ SPICE Simulator TI Precision Designs Library
FilterPro™ SPICE Simulator
OPA1604 Product Folder
OPA1602 Product Filter
3rd-Order All-
4th-Order High- Attenuation
Pass Time
Pass Filter and Buffering
Delay
Tweeter
Output
Audio
Input
Input
Buffer Woofer
Output
Baffle Step 4th-Order Low-

Compensation Pass Filter
An IMPORTANT NOTICE at the end of this TI reference design addresses authorized use, intellectual property matters and
other important disclaimers and information.
TINA-TI is a trademark of Texas Instruments
WEBENCH is a registered trademark of Texas Instruments
TIDU035-December 2013-Revised December 2013 Analog, Active Crossover Circuit for Two-Way Loudspeakers 1
Copyright © 2013, Texas Instruments Incorporated
www.ti.com
1 Design Summary
The design requirements are as follows:
 Supply Voltage: +/- 15 V
 Input and Output Levels: +4dBu / 1.228Vrms
 Crossover Frequency (Acoustic): 1.8kHz
 Loudspeakers and Enclosure Dimensions: See Appendix

The design goals and performance are summarized in Table 1.
Figure 1 depicts the measured transfer function of the design.

Table 1. Comparison of Design Goals, Simulation, and Measured Performance
Specification Design Goal Simulation Measured
Baffle Step Compensation (at 650 Hz) 6 dB 5.96 dB 6.137 dB
Low-pass Filter Corner Frequency 2.145 kHz 2.1096 kHz 2.113 kHz
Tweeter Attenuation 8 dB 7.91 dB 8.04 dB
High-pass Filter Corner Frequency 1.8 kHz 1.746 kHz 1.737 kHz
Nominal Tweeter Delay 155 us 155.28 us 155.261 us
10% Degradation in Tweeter Delay >3.6 kHz 3.814 kHz 4.083 kHz
THD+N (100 Hz) <.01% 0.000146% 0.0003%
THD+N (10 kHz) <.01% 0.000484% 0.00128%
Transfer Function of Individual Outputs

0
-10
-20
Amplitude (dB)
-30
-40
-50
Woofer Output
Tweeter Output
-60
10 100 1000 10000
Frequency (Hz)
Figure 1: Measured Transfer Function of the crossover network.
2 Analog, Active Crossover Circuit for Two-Way Loudspeakers TIDU035-December 2013-Revised December 2013
www.ti.com
2 Theory of Operation
The primary purpose of the crossover circuit in a loudspeaker is to split an incoming audio signal into
frequency bands that are passed to the speaker or “driver” best suited. For example, in a two-way system
a crossover circuit will pass low frequencies to the woofer and high frequencies to the tweeter. This is
accomplished using passive or active filters to remove frequencies outside of the desired band for a driver.
A secondary purpose of crossover circuits is to correct the frequency or phase response of the system for
errors introduced by the loudspeaker enclosure and listening environment.
Active crossover networks are commonly used in recording studios and some home high-fidelity systems.
In an active crossover system, the crossover network is placed before the power amplifiers in the audio
signal chain. The voltages at this point in the signal chain are much lower than those applied directly to the
speaker, allowing the use of active filters which employ op-amps, capacitors, and resistors. Expensive
passive components which must maintain linearity at high voltage levels can be eliminated from active
crossover circuits. Furthermore, the filter circuits in an active crossover do not directly interact with the
loudspeaker impedance, allowing them to more closely follow the desired transfer function without
complicated analysis.
R1
R2
590
VCC 590
C4 C7 VCC
U1A C9 C10
++
100n 100n U2A
- ++
R3 ¼ OPA1604 100n 100n
-
1.30k R4 ¼ OPA1604
VEE 1.30k
VEE
High-Pass Filter
C12
C11
R5 R6
10p
100n R7
1k 1k R8
VEE 422
VEE 1k VEE
VEE
- U2B R9 C13 VEE
U2C
- R10 - Tweeter
- ++ 221 100n U2D
R11
¼ OPA1604
++ - ++
475
++ ¼ OPA1604 U3B
U3A
VCC
++ 1.74k ½ OPA1602
½ OPA1602 ¼ OPA1604
R22 VCC VCC
+
C14 R14 R12

VCC R13 VCC
VG1 10k 1.15k
1k
Tweeter Level
332
100n All-Pass Filter Correction
C15 R15
C16
100n 11.8k C17
R16
100n
11.8k 100n
VEE
R17 VCC
U1B R18 R19 ¼ OPA1604 VCC
- U1C
11.8k R20 R21 ¼ OPA1604
++ ++ U1D
¼ OPA1604
845 1.4k ++ Woofer
- 845 1.4k
C18
VCC -
47n C19
Baffle Step VEE 47n
VEE
Compensation
Low-Pass Filter
Figure 2: Complete crossover network schematic with functional portions highlighted
This design can be broken into 5 parts which are shown in Figure 2. The signal path for the woofer output
includes a baffle step compensation circuit and a low-pass filter. The tweeter portion includes a high-pass
filter, an all-pass filter for time alignment, and output level correction. The design parameters are chosen to
represent a typical two-way monitor speaker which may be employed in a small room such as a recording
studio for near-field listening and sound mixing work.
Proper loudspeaker design requires the use of acoustic measurements during the design process. Each
portion of the design theory section will incorporate raw acoustic measurements of the drivers installed in
an enclosure and placed in the desired listening location. All measurements were made at a distance of
th
1m, on-axis with the tweeter. The plots presented use 1/8 octave smoothing and the measurements were
un-gated unless otherwise noted. This was done to show the effects produced by the loudspeaker’s
interaction with the listening environment.
www.ti.com
2.1 Baffle Step Compensation

Conventional speaker systems which use a “box” type enclosure will exhibit variations in the frequency
response due to diffraction effects of the enclosure. A major contributor to these variations in the frequency
response is diffraction from the front wall or “baffle” of the enclosure. At low frequencies, the wavelength of
the sound being radiated is very large compared to the width of the enclosure; at these frequencies the
speaker’s radiation pattern is essentially spherical. At higher frequencies, the wavelength of the sound
being produced approaches the width of the front baffle, and the speaker begins to radiate in a
hemispherical pattern [1]. Because the speaker is radiating the same amount of energy into half of the
space, the sound intensity will increase by 6dB as the speaker transitions from radiating in a spherical
pattern to a hemispherical pattern. This effect is commonly referred to as “baffle step” and visible in the
rising frequency response of the woofer shown in Figure 3. This measurement was taken with the woofer
in the enclosure and includes the effects of the listening room as well.
Woofer Frequency Response

10
5
0
Amplitude (dB)
-5
-10
-15
-20
-25
-30
-35
-40
10 100 1000 10000
Frequency (Hz)
Figure 3: Woofer frequency response measured at a 1m distance on the tweeter axis (designed
th
listening position). 1/8 Octave smoothing is applied and no gating is used (includes room effects)
Software packages are available which can predict the diffraction effects of the enclosure. Figure 4
illustrates the predicted diffraction effects from mounting a 178 mm woofer, on a 216 mm x 356 mm baffle,
127 mm from the bottom edge. The 3dB point of the step is at 313Hz with an additional peak near 1000
Hz, which is a diffraction effect of the corners on the enclosure. Rounding these edges helps to reduce the
amplitude of this peak.
The blue curve in Figure 4 depicts the diffraction effects in an anechoic environment. Most listening
environments are not anechoic and have acoustically reflective surfaces which will cause further variations
in the frequency response. The red curve shows an example of the effects of placing the loudspeaker in a
small room with adjacent reflective surfaces.
The term “baffle step compensation” refers to techniques used to compensate for the rise in a speaker’s
frequency response due to the changing radiation pattern. This can be accomplished by applying a
specific amount of attenuation to input frequencies above the transition frequency.
www.ti.com
Baffle Diffraction and Room Interaction Effects

20
Amplitude Response (dB) 18
16
14
12
10
8
6
4
2
0
10 100 1000 10000
Frequency (Hz)
Baffle Diffraction Combined Baffle Diffraction and Room Response
Figure 4: Predicted baffle diffraction effects (blue curve) of a 178 mm woofer mounted on an 216
mm x 356 mm baffle (13 mm radius edge rounding). The center point of the woofer is 127 mm from
the bottom edge. The red curve includes room acoustic effects as well. Curves generated by Baffle
Diffraction and Boundary Simulator v1.2 © Jeff Bagby, used with permission.
The low-pass shelving circuit shown in Figure 5 can be used to compensate for this rise in amplitude
response with increasing frequency. At low frequencies, the gain of the circuit is:
(1)
At high frequencies, when the impedance of the capacitor becomes very small, the gain of the circuit has
been decreased:
(2)
This shelving behavior is illustrated by the red curve in Figure 6. A more generic form of the magnitude
response for this circuit is:
| | (3)
In order to properly design the baffle step compensation circuit, the woofer frequency response from
Figure 3, was compared to a flat target response. A numerical solver algorithm compared the woofer
frequency response to the target response for frequencies between 100Hz and 1kHz and determined the
passive component values which provided the flattest response in this region. The final values are shown
in Figure 5.
www.ti.com
C15 R15
100n 11.8k
R16
11.8k
VEE
From Input Buffer
R17
U1B
-
11.8k To Low-Pass Filter
+ +
¼ OPA1604
VCC
Figure 5: A shelving circuit used to compensate for the baffle diffraction step
The frequency response of the baffle step compensation circuit is shown in Figure 6 in red. The raw
woofer response is shown in blue, and the corrected woofer response is shown in green.
Woofer Frequency Response with Baffle Step Compensation

10
5
0
-5
Amplitude (dB)
-10
-15
-20
-25
-30 Raw Woofer Response
-35 Baffle Step Compensation
Corrected Woofer Response
-40
10 100 1000 10000
Frequency (Hz)
Figure 6: Corrected woofer response including the effects of both the low-pass filter and shelving
circuit
As can be seen in Figure 6, the gain at low-frequencies is 1 (0dB) because R16 and R17 are equal. At
high frequencies the gain of the circuit is -6.0206dB:
( ) (4)
Using the component values shown in Figure 5, the -3dB point will be 95 Hz. The negative sign in the gain
equation indicates that the polarity of the output signal has been reversed. In order to correct for this phase
inversion, the woofer must be connected with reversed polarity to maintain proper phase orientation.
2.2 Low Pass Filter

For this design, the corner frequency for the woofer is 1.8kHz; a very common value for two-way
loudspeakers of this size. The selection of this frequency is beyond the scope of this document and is a
function of the woofer size, desired on-axis and off-axis frequency response, distortion characteristics, and
proximity to the tweeter.
www.ti.com
The Linkwitz-Riley filter characteristic was selected because these filters sum acoustically flat in the
th
crossover region [2]. A 4 -order Linkwitz-Riley filter has a steep roll-off (48 dB/octave, 80 dB/decade)
which limits high-frequency distortion from the woofer, and protects the tweeter from low-frequency content
which may damage it.
Although the desired corner frequency for the woofer transfer function is 1.8kHz, this does not necessarily
mean the corner frequency for the low-pass filter will be 1.8kHz. Examining the woofer frequency response
with baffle step compensation in Figure 7 (blue curve), shows that the driver itself begins to attenuate
signals above 2kHz. Therefore, the 1.8kHz corner frequency specifies the acoustic transfer function, which
is the combination of the low-pass filter transfer function with the driver’s frequency response.
th
Again, using a numerical solver, it was found that a 4 -order low-pass filter with a corner frequency of
2.145kHz provided the desired acoustic corner frequency of 1.8kHz. Figure 8 shows the predicted acoustic
transfer function of the woofer produced by multiplying the woofer frequency response (with baffle step
compensation) by the transfer function of the filter.
Target Woofer Response

10
-10
Amplitude (dB)
-20
-30
-40
-50
-60 Woofer Response w/BSC

Target Response
-70
10 100 1000 10000
Frequency (Hz)
th
Figure 7: Woofer frequency response including baffle step compensation (blue) and the target 4 -
order Linkwitz-Riley low-pass transfer function with 1.8 kHz corner frequency (black)
www.ti.com
Filtered Woofer Frequency Response

10
-10
Amplitude (dB)
-20
-30
-40
-50 Woofer Response w/BSC

Filter Response
-60 Filtered Woofer Response
Target Response
-70
10 100 1000 10000
Frequency (Hz)
Figure 8: The woofer's frequency response including the filter response (green). The filter
frequency response is shown in red, target response in black, and woofer response with baffle
step compensation (BSC) shown in blue
th nd
A 4 -order Linkwitz-Riley filter may be constructed by cascading two 2 -order Butterworth low-pass filters
as shown in Figure 9[2]. The Sallen-Key (SK) filter topology is used because the noise gain and signal
gain for this topology are equal since the op amp is configured as a non-inverting amplifier. In the multiple
feedback (MFB) topology, the op amp is configured as an inverting amplifier and its noise gain would be
twice the signal gain for a unity gain filter, degrading the system signal to noise ratio.
C16
C17
VCC
R18 R19 VCC
U1C
++ R20 R21
U1D
++
-
C18
-
C19
VEE
VEE
2nd-Order Butterworth
Low-Pass Filter 2nd-Order Butterworth
Low-Pass Filter
nd
Figure 9: A 4th-Order Linkwitz-Riley low-pass filter made by cascading two 2 -order Butterworth
low-pass filters
www.ti.com
Texas Instruments FilterPro™ was used for all filter designs in this document. In order to design the filter in
nd
Figure 9, a 2 -order Sallen-Key Butterworth low-pass filter with an 2.145 kHz corner frequency was first
designed. The filter gain, order, and corner frequency are set in the filter specifications window of the New
Design Wizard as shown in Figure 10.
Figure 10: The Filter Specifications window of the FilterPro™ New Design Wizard
A Butterworth filter response is selected in the Filter Response dialogue (Figure 11) and the Sallen-Key
topology is selected on the final Filter Topology window (Figure 12).
www.ti.com
Figure 11: The Filter Response window of the FilterPro™ New Design Wizard
Figure 12: The Filter Topology window of the FilterPro™ New Design Wizard
www.ti.com
The resulting passive component values will be “exact” values which do not conform to standard resistor
and capacitor values. Figure 13 shows the design after changing the resistor tolerance to 1% and the
capacitor tolerance to 10%. The value of C2 in Figure 13 was forced to 100nF and FilterPro™ calculated
the resulting values of other circuit components. The 100nF value for C2 was selected after researching
multiple suppliers revealed that this was the largest practical value for a C0G ceramic capacitor in a 1206
surface mount package. Choosing the largest available capacitor value allows for lower resistor values,
reducing the thermal noise of the system and minimizing the input current noise contribution of the op
amp.
nd
Figure 13: Final 2 -order Butterworth low-pass filter design using 1% resistor tolerances and 10%
capacitor tolerances. The value for C2 was forced to 100nF
nd th
Once the 2 -order Butterworth low-pass filter has been designed, the final 4 -order Linkwitz-Riley filter is
nd
simply two cascaded 2 -order filters as shown in Figure 14. Finally, because this is a low-pass filter, it
should be placed last in the woofer signal path. This will allow the filter to attenuate the high-frequency
noise of the circuits before it, improving the system signal to noise ratio.
C16
C17
100n
100n
VCC
R18 R19 ¼ OPA1604 VCC
From Baffle Step U1C ¼ OPA1604
Compensation ++ R20 R21
845 1.4k U1D
++ To Woofer
- 845 1.4k
C18 Output
-
47n C19
VEE 47n
VEE
th
Figure 14: Final 1.8 kHz, 4 -order Linkwitz-Riley low-pass filter
www.ti.com
2.3 High Pass Filter

As stated in the previous section, the crossover frequency for this design is 1.8 kHz. Therefore the tweeter
transfer function should intersect with the woofer’s at 1.8 kHz. The raw response of the tweeter is shown in
th
Figure 15, and is fairly flat through the desired passband. A 4 -order Linkwitz-Riley high-pass filter with a
corner frequency at 1.8 kHz provides the desired acoustic transfer function (shown in green in Figure 15).
Filtered Tweeter Frequency Response (Predicted)

20
10
0
Amplitude (dB)
-10
-20
-30
Tweeter Raw Response
Filter Response
-40
Filtered Tweeter Response
Target Response
-50
100 1000 10000
Frequency (Hz)
th
Figure 15: Tweeter frequency response without filtering (red) and with a 1.8 kHz 4 -order Linkwitz-
Riley high-pass filter (green). The filter response and target responses are shown in blue and black
respectively
The high-pass filter portion of the tweeter signal path was designed using FilterPro™ in the same manner
as is described in section 2.1. All of the capacitor values in the filter are 100nF to allow for the lowest
resistor values possible and to reduce the number of Bill-of-Material (BOM) line items, theoretically
lowering system cost.
R1
R2
590
VCC 590
C4 C7 ¼ OPA1604 VCC
From Input C9 C10
++ U1A ¼ OPA1604
Buffer
100n 100n U2A
- ++ To All-Pass
R3 100n 100n Filter
-
1.30k R4
VEE 1.30k
VEE
th
Figure 16: 4 -order Linkwitz-Riley high-pass filter in the tweeter signal path
2.4 All Pass Filter

In a two-way loudspeaker, a time delay may be added to the tweeter signal in order to compensate for the
path length difference between the tweeter and woofer [2]. The path length difference depends upon the
size and mounting orientation of the drivers as well as the desired listening position.
www.ti.com
A
P
B
C
D
Figure 17: The path length geometry of a typical two-way loudspeaker
Figure 17 shows a cross section of a typical two-way loudspeaker and the relevant dimensions labeled.
The distance from each driver to the listening position (labeled point P) is measured from its “acoustic
center” which we will consider to be the top of the voice coil for both drivers. In order to calculate the
tweeter time delay necessary to ensure proper phase alignment, the woofer path length C must first be
calculated
√( ) (5)
The dimensions A, B, and D are the listening distance, driver center-to-center spacing, and the distance
the woofer acoustic center is behind the front of the enclosure. These dimensions are summarized in
Table 2.
Table 2: Dimensions for the diagram in Figure 17
Dimension Value (m)
A 1
B 0.1524
D 0.05
√( ) ( ) (6)
The difference in path length between the woofer and the tweeter is then:
(7)
Using 346.1 m/s as the speed of sound at 25°C, gives a delay value of:
(8)
This value may seem insignificant. However, consider that at the crossover frequency of 1.8 kHz, a 176 μs
time delay contributes a 114.21° phase shift.
www.ti.com
The 176 μs value is only the delay introduced by the physical arrangements of the drivers, it does not
include any additional delay introduced by the filter networks in the crossover. Figure 18 shows the group
delay of the woofer and tweeter signal pathways through the crossover region.
Group Delay of Woofer and Tweeter Networks

350
300
Group Delay (μs)
250
200
150
100
50 Woofer Group Delay

Tweeter Group Delay
0
1000 Frequency (Hz)
Figure 18: Group delay of woofer and tweeter circuits through the crossover region
Through the crossover region, the tweeter circuit has an average of 24.3 μs more delay than the woofer
circuit. Therefore, the additional delay that must be added to the tweeter signal to compensate for the
mounting orientation is:
(9)
Many aspects of these calculations can change drastically in a real listening environment. Listening
position, driver acoustic center variations, air temperature and humidity are all variables which will change
the necessary delay amount [2]. For this reason, a time delay value of 155 μs is selected as a compromise
value across multiple listening positions.
Adding delay in an analog active crossover is accomplished using an all-pass filter. While the amplitude
response of an all-pass filter is flat, the phase response varies with frequency. This behavior allows the
filter to add a known delay to the signal without affecting the amplitude response.
The order of an all-pass filter specifies the rate at which the phase change occurs at the corner frequency.
The advantage of a higher order all-pass filter is that it is able to maintain a constant group delay at higher
frequencies. According to FilterPro™, the lowest-order all-pass filter capable of maintaining a 155 μs delay
to 1.8 kHz is a third-order filter as shown in Figure 19.
C12
C11
R5 R6
10p
100n R7
1k 1k R8
VEE 422
VEE 1k
- U2B R9 C13 VEE

U2C
From High- - R10
Pass Filter ++ 221 100n U2D
¼ OPA1604
++ - To
475 Attenuation
¼ OPA1604
VCC
++ Network
¼ OPA1604
C14 VCC
R13 R14
VCC
332 1k
100n
Figure 19: A 3rd-order all-pass filter providing 155 us delay to the tweeter signal path
www.ti.com
All capacitor values were forced to 100nF and the resistor values were made as low as practical (1%
tolerances are used). C12 is included to attenuate the noise contribution of amplifier U2D at frequencies
above the audible range.
2.5 Level Control and Buffering

In general, speaker drivers which reproduce high frequencies are more efficient than those intended to
reproduce low frequencies. “Sensitivity” is defined as the sound pressure level produced by the driver for a
1 W input signal (2.83 Vrms for an 8 Ohm driver) when measured at a 1 meter distance [1]. However,
because loudspeakers are typically driven by a voltage source, it is more appropriate to compare their
sound pressure levels for a certain input voltage. Table 3 compares the sound pressure level produced by
the woofer and tweeter for a 2.83 Vrms input signal.
Table 3: Comparison of sound output
Driver Output Level (2.83V/1m)
Woofer 87 dB
Tweeter 91 dB
The result of higher tweeter output is that the audio signal to the tweeter must be attenuated for the system
to have a flat frequency response. The listening environment and associated baffled step compensation
circuitry also must be considered when calculating the amount of attenuation.
Relative Output of Individual Drivers

20
10
0
-10
Amplitude (dB)
-20
-30
-40
-50
-60
100 1000 10000
Frequency (Hz)
Woofer Response Tweeter Response Tweeter Response (8dB attenuation)
Figure 20: Woofer and tweeter individual frequency responses including the filter effects. The
measurements are normalized to the woofer raw response.
As shown in Figure 20, the woofer output is nominally 8dB less than that of the tweeter when the effects of
the baffle step compensation circuit and listening environment are included. The green curve in Figure 20
shows the predicted tweeter output with 8dB of attenuation. Resistors R11 and R12 form a voltage divider
which attenuates the signal going to the tweeter. The ratio of the two resistor values necessary to achieve
8dB of attenuation is calculated:
(10)
For 1% resistor denominations the values closest to this ratio are shown in Figure 21.
( ) (11)
www.ti.com
VEE
- To Tweeter
R11 Output
From All- ++ U3B
Pass Filter
1.74k ½ OPA1602
R12 VCC
1.15k
Figure 21: Tweeter level correction and output buffer circuit.
A buffer is placed after the resistor network (U3B) in order to prevent the input impedance of any
subsequent circuit in the signal path from affecting the attenuation factor. Likewise, the other half of dual
op amp U3A is used as an input buffer to prevent the impedance of the source from affecting the individual
filter transfer functions. The predicted loudspeaker frequency response is shown in Figure 22 along with
the individual driver transfer functions. The predicted crossover point is 1.708 kHz.
Predicted System Frequency Response

10
-10
Amplitude (dB)
-20
-30
-40
-50
-60
10 100 1000 10000
Frequency (Hz)
Woofer Response Tweeter Response (8dB attenuation) System Response
Figure 22: Predicted loudspeaker frequency response (in room) with individual driver responses
shown
www.ti.com
3 Component Selection
3.1 Amplifiers
The basic amplifier requirements for this system are summarized in Table 4. The power supply
requirement is determined in the design summary. The gain bandwidth requirement is determined by
FilterPro™ for proper filter functionality. Because the second stage of the all-pass filter requires the highest
op-amp gain bandwidth of any filter stage, this value is determined to be the minimum for all amplifiers
used in the system.
Table 4: Basic requirements for amplifiers used in the crossover network
Requirement Value
Max Power Supply >30V
Gain Bandwidth Product 180.3kHz
Slew Rate 2.2 V/μs
The slew rate requirement is calculated from the maximum slew rate of a sinusoid at 20kHz, and
1.228Vrms:
V/μs (12)
An amplifier with this slew rate limitation would contribute significant distortion at 20 kHz because the
sinusoid would exhibit a triangular shape. A conservative design decision is to employ amplifiers with slew
rates 10 times greater than the levels calculated in equation 12 in order to minimize distortion from
slewing.
Table 5 shows a comparison of the cost, input voltage noise, power supply current, and measured THD+N
of several quad op-amps which meet the above requirements and are intended for audio applications. The
NE5532A and NE5534A are included in the table because these parts are ubiquitous in audio applications.
It is often very difficult to justify the use of newer amplifiers given the high level of performance and low
cost of the NE5532A and NE5534A.
Large capacitor values and low resistor values were selected in each portion of the circuit to minimize the
thermal noise of the filter components and the impedance presented to the op amp inputs, reducing the
contribution of amplifier input current noise to the total output noise. Therefore, bipolar input amplifiers are
the preferred solution because the benefits of lower input voltage noise outweigh the increased current
noise.
The OPA160x family of amplifiers was selected for this design because it is the lowest noise bipolar audio
op amp that is available in a quad package (4-amplifiers in one package). Furthermore, the OPA160x
family consumes significantly lower power supply current than the next lowest noise option, the
LME49740. High-fidelity audio systems may involve a significant number of operational amplifiers and any
additional supply current consumption may increase the cost of the power supply solution.
Table 5: A comparison of low noise bipolar op-amps commonly used in audio applications
Cost per Vn Power Supply THD+N (1kHz, G=+1,
Amplifiers
Amplifier Amplifier (1ku (nV/√Hz, Current (typ mA / 3Vrms, 600 Ohm
per Package
prices) 1kHz) amplifier) load)
NE5534A 1 $0.50 3.5 4 0.002%
NE5532A 2 $0.25 5 4 N/A
OPA1604 4 $0.49 2.5 2.8 0.00003%
OPA1664 4 $0.36 3.3 1.5 0.00006%
LME49740 4 $0.43 2.7 4.625 0.00003%
LME49743 4 $0.29 3.5 2.5 0.0001%
www.ti.com
The THD+N performance of the OPA160x family is shown in Figure 23. It is important to notice that there
is little difference between the curves for inverting and non-inverting configurations, indicating that the part
resists common-mode distortion. This is an important consideration in Sallen-Key filters where the part is
configured as a non-inverting amplifier [2].
Figure 23: Measured THD+N performance of the OPA160x family of amplifiers for multiple gains
and loading conditions.
Low impedance loads will increase the distortion produced by any amplifier because the output stage
becomes less-linear as it is forced to deliver greater amounts of current. At high frequencies, this distortion
becomes more apparent where there is less loop gain available to correct for output stage distortion
through negative feedback. For this reason, the resistor values cannot be reduced to extremely low values
in the pursuit of increased fidelity.
3.2 Passive Components

Proper selection of passive components is absolutely crucial to achieving low levels of distortion in active
filters. High-K ceramic and electrolytic capacitors are not suitable for active filter circuits because they
introduce significant amounts of odd-order harmonic distortion. This effect has been shown in numerous
publications [2, 3, 4, 5].
The only capacitor types suitable for active filters are C0G ceramic, polypropylene film, or silvered mica.
However, because silvered mica capacitors are not available in large capacitances they are not
considered here. When comparing C0G ceramic capacitors to polypropylene film types, it was found that
for the same tolerance, C0G ceramics were cheaper and occupied less board area than film types of the
same capacitance.
Table 6: A Comparison of C0G ceramic to polypropylene film capacitors
2
Capacitor Type Tolerance Board Area (mm ) Cost (Each, Qty: 100)
NP0/C0G (1206) 5% 5.12 $0.33
Polypropylene Film 5% 65 $0.44
1206 surface mount packages (120mil x 60 mil) were chosen for both resistors and capacitors. The 1206
package was the smallest package in which a 100nF C0G capacitor was available.
Distortion issues have also been reported for surface mount resistors in extremely small packages and
with certain resistive materials [2,5]. Testing for this design showed that thick film resistors in 1206
packages produced no measurable distortion at the signal levels commonly found in line-level audio
equipment.
www.ti.com
4 Simulation
4.1 Woofer Signal Path

th
The woofer signal path is composed of the baffle step compensation circuit and 4 -order low-pass filter
and the Tina-TI™ simulation schematic is shown in Figure 24.
Figure 24: Tina-TI™ simulation schematic of the woofer signal path
4.1.1 Transfer Function
Performing an AC Transfer Characteristic simulation in Tina-TI™ shows the transfer function of the woofer
signal path. Because Linkwitz-Riley filters are defined by the -6dB point, the corner frequency shown in
Figure 25 includes 6dB of attenuation from the low-pass filter as well as 6dB of attenuation from the baffle
step compensation circuit. The simulated attenuation was 5.96 dB at 650 Hz and the corner frequency was
2.1096 kHz.
Figure 25: Tina-TI™ Simulation of the woofer signal path transfer function.
www.ti.com
4.1.2 Noise
Most SPICE macro-models for op amps do not accurately simulate the distortion behavior of integrated
circuits but noise can be used as an indicator of audio quality, with lower noise indicating a higher quality
system.
Figure 26: The simulated output noise integration over an 80 kHz frequency range for the woofer
signal path.
A common figure-of-merit in audio systems is the total harmonic distortion and noise (THD+N) which is
calculated using this equation:
∑ ( )
( ) √ (13)
Where Vi is the RMS voltage of the “ith” harmonic (i=2,3,4…), Vn is the RMS noise voltage, and Vf is the
RMS voltage of the fundamental. At low frequencies, the THD+N measurement will be dominated by the
noise of the system. This allows for the simplification of the THD+N equation to estimate the noise
contribution to the measured THD+N.
( ) √ (14)
According to the Tina-TI™ total noise simulation, the RMS noise voltage of the woofer signal path in an 80
kHz bandwidth will be 1.248 μVrms. At 100Hz, the fundamental frequency will be attenuated by 3.15 dB,
reducing the amplitude of the 1.228 Vrms input signal to .8544 Vrms.
( )
( ) √ (15)
( )
www.ti.com
4.2 Tweeter Signal Path

th rd
The tweeter section is composed of the 4 -order high-pass filter, 3 -order all-pass filter, and the output
attenuator and buffer.
Figure 27: TINA-TI™ simulation schematic of the tweeter signal path.
4.2.1 Transfer Function
An AC Transfer Characteristic was used to simulate the transfer function of the tweeter signal path. The
simulation shows 7.91dB of attenuation in the passband and a corner frequency of 1.746 kHz.
Figure 28: TINA-TI simulation of the tweeter signal path transfer function.
www.ti.com
4.2.2 Noise
A total noise simulation was performed in TINA-TI™ to predict the noise contribution to the measured
THD+N value.
Figure 29: Tina-TI™ total noise simulation of the tweeter signal path.
According to the noise simulation, the RMS noise voltage of the tweeter signal path in an 80 kHz
bandwidth will be 2.369 μVrms. The significant increase in noise is due to the greater number of op amp
circuits in the tweeter signal path. A 10 kHz, 1.228Vrms input signal will be attenuated by 8dB, reducing
the fundamental to .489 Vrms, giving a predicted THD+N at 10 kHz of:
( )
( ) √ (16)
( )
This number is very optimistic because the measured THD+N at high frequencies will be dominated by
distortion harmonics rather than the noise voltage.
4.2.3 Tweeter Signal Delay

rd
The 3 -order all-pass filter circuit can be tested by selecting the “Group Delay” option in the AC Transfer
Characteristic options window.
www.ti.com
Figure 30: Tina-TI™ simulation schematic for testing all-pass filter delay.
rd
A 3 -order all-pass filter is not able to maintain the necessary delay for the entire audible bandwidth.
However, it is preferable that significant reduction in the delay occur at least an octave above the 1.8 kHz
crossover point of the system. A 10% reduction in delay has been used in other literature as a reference
point for significant delay reduction [2].
Figure 31: Tina-TI™ simulation of the all-pass filter group delay.
The simulated nominal group delay of the all-pass filter circuit is 155.28 μs and falls to 90% of this value at
3.814kHz.
www.ti.com
4.3 Simulated Results Summary

A comparison of the simulation results to the design goals is given in Table 7.
Table 7: Simulated Results Summary
Specification Design Goal Simulation
Baffle Step Compensation (at 650Hz) 6 dB 5.96 dB
Low-pass Filter Corner Frequency 2.145 kHz 2.1096 kHz
Tweeter Attenuation 8 dB 7.91 dB
High-pass Filter Corner Frequency 1.8 kHz 1.746 kHz
Nominal Tweeter Delay 155 us 155.28 us
10% Degradation in Tweeter Delay >3.6 kHz 3.814 kHz
THD+N (100 Hz) <.01% 0.000146%
THD+N (10 kHz) <.01% 0.000484%
www.ti.com
5 PCB Design
The PCB schematic and bill of materials can be found in the Appendix.
5.1 PCB Layout

Standard low-noise PCB layout practices are required to achieve a high level of performance in audio
designs. Specifically for this design, a primary concern is minimizing the signal path length and loop areas
of the filter circuits. This helps to prevent extrinsic interference from entering the audio signal path. To
accomplish this goal, careful consideration must be given to which op amp in each quad package is used
for every circuit. For example, the first op amp in both signal paths are in the same package (U1A, U1B) to
allow for more efficient routing of the input signal. As usual, bypass capacitors must be placed as close as
possible to the op amp power supply pins and are connected to a low-impedance pathway to ground.
Figure 32: Top layer (left) and bottom layer (right) views of the PCB.
www.ti.com
6 Verification & Measured Performance
6.1 Electrical Measurements
6.1.1 Woofer Signal Path Transfer Function
The woofer signal path transfer function was measured using an industry standard audio analyzer. The
measured baffle step compensation was 6.137 dB and the corner frequency was 2.113 kHz; very close to
the simulated values.
Woofer Signal Path Transfer Function (Measured)

0
-10
-20
Amplitude (dBV)
-30
-40
-50
-60
-70
-80
-90
10 100 1000 10000
Frequency (Hz)
Figure 33: Measured transfer function of the woofer signal path
6.1.2 THD+N Performance – Woofer Output
The THD+N performance of the woofer signal path was measured using an audio analyzer with a
measurement bandwidth of 80 kHz and a 1.228 Vrms (0 dBu) input signal level. THD+N measurements
are only displayed within the passband of the specific filter. In the stop band of the filter, the attenuation of
the input signal has the effect of increasing the measured THD+N (see equation 13).
The instrument measurement floor is displayed as the green curve in Figure 34. Because the woofer signal
path includes a low-pass filter, it also attenuates the high-frequency noise of the analyzer. This is why the
measured THD+N of the woofer signal pathway is below the instrument floor for frequencies less than 800
Hz. At 100 Hz, the measured THD+N is .0003%. This is more likely the noise floor of the instrument and
not the actual distortion of the woofer signal pathway.
www.ti.com
Woofer Signal Path THD+N (1.228 Vrms Input, 80kHz)

0.001
THD+N (%)
Woofer Signal Path

Analyzer
0.0001
10 100 1000
Frequency (Hz)
Figure 34: Measured THD+N performance of the woofer signal path
6.1.3 Tweeter Signal Path Transfer Function
The measured transfer function of the tweeter signal path showed 8.04 dB of attenuation and a corner
frequency of 1.737 kHz.
Tweeter Signal Path Transfer Function (Measured)

0
-10
-20
Amplitude (dBV)
-30
-40
-50
-60
-70
-80
-90
100 1000 10000
Frequency (Hz)
Figure 35: Measured transfer function of the tweeter signal path.
www.ti.com
6.1.4 Tweeter Signal Path THD+N
As expected, the greater number of op-amps in the tweeter signal pathway does degrade the THD+N
performance slightly. The measured THD+N at 10 kHz is .00128% which is significantly below the design
goal.
Tweeter Signal Path THD+N (1.228 Vrms Input, 80 kHz)

0.01
THD+N (%)
0.001
Tweeter Signal Path

Analyzer
0.0001
1000 Frequency (Hz) 10000
Figure 36: Measured THD+N of the tweeter signal path.
6.1.5 All-pass Filter Group Delay
The group delay of the all-pass filter in the tweeter signal path can be calculated from phase
measurements using the relationship:
(17)
Where θ is the unwrapped phase in degrees and f is the frequency in Hz.
All-pass Filter Group Delay

1.8E-04
1.6E-04
1.4E-04
Group Delay (s)
1.2E-04
1.0E-04
8.0E-05
6.0E-05
4.0E-05
2.0E-05
0.0E+00
10 100 1000 10000
Frequency (Hz)
Figure 37: The group delay of the all-pass filter was calculated from a phase measurement
www.ti.com
6.2 Acoustic Measurements

The acoustic transfer functions of the drivers were measured using a calibrated measurement microphone,
at a distance of 1m, on-axis with the tweeter. The measurements are taken un-gated to show the effects of
the listening environment, except where noted. A diagram of the measurement setup is given in Appendix
C.
6.2.1 Woofer Acoustic Transfer Function
The negative -6dB point of the woofer acoustic transfer function is 1725Hz.
Predicted and Measured Woofer Frequency Response

10
-10
Amplitude (dB)
-20
-30
-40
-50 Predicted Woofer Response

Measured Woofer Response
-60
10 100 1000 10000
Frequency (Hz)
Figure 38: Predicted and measured acoustic transfer function of the woofer. 1/8th octave
smoothing
www.ti.com
6.2.2 Tweeter Acoustic Transfer Function
The measured -6dB point of the tweeter acoustic transfer function was 1.5kHz.
Predicted and Measured Tweeter Frequency Response

10
-10
Amplitude (dB)
-20
-30
-40
-50 Predicted Tweeter Response

Measured Tweeter Response
-60
100 1000 10000
Frequency (Hz)
Figure 39: Predicted and measured acoustic transfer function of the tweeter (1/8th octave
smoothing) Ambient noise in the measurement environment affects the results below -40dB
6.2.3 System Frequency Response
The measured system frequency response closely matches the predicted response over much of the
frequency range. A small hump is apparent near the crossover frequency. This is because the -6dB point
of the tweeter response is lower than anticipated, causing a region of overlapping response from the
woofer and tweeter.
Predicted and Measured System Frequency Response

10
-10
Amplitude (dB)
-20
-30
-40
Measured System Response
-50
Predicted System Response
-60
10 100 1000 10000
Frequency (Hz)
Figure 40: Predicted and measured system frequency response.
www.ti.com
6.2.4 Phase Alignment Testing
Inverting the phase of one of the drivers is a test commonly used to confirm proper phase alignment of the
drivers in a loudspeaker at the crossover frequency. By inverting the phase of one of the drivers, a 180
degree phase shift is introduced to that driver’s output, causing destructive interference at the crossover
frequency. This test verifies that when the drivers are connected with correct polarity, their outputs are in-
phase at the crossover frequency. Gating is used to remove the effects of reflected sound from the
listening room. For this reason, the displayed frequency response below 500 Hz is inaccurate. The
inverted polarity connection produced a 25dB notch at 1.6 kHz.
Inverted Woofer Polarity

10
-10
Amplitude (dB)
-20
-30
-40
-50 Correct Phase

Inverted Phase
-60
100 1000 10000
Frequency (Hz)
Figure 41: Acoustic measurement of the loudspeaker with drivers connected in-phase and inverted
phase. Connecting in inverted phase produces a 25 dB notch at 1.6 kHz.
6.3 Measured Results Summary

The measured results satisfy the design goals meet or exceed the design goals in most categories.
Table 8: A comparison of the measured results to the design goals.
Specification Design Goal Measured
Baffle Step Compensation (at 650Hz) 6 dB 6.137 dB
Low-pass Filter Corner Frequency 2.145 kHz 2.113 kHz
Tweeter Attenuation 8 dB 8.04 dB
High-pass Filter Corner Frequency 1.8 kHz 1.737 kHz
Nominal Tweeter Delay 155 us 155.261 us
10% Degradation in Tweeter Delay >3.6 kHz 4.083 kHz
THD+N (100 Hz) <.01% 0.0003%
THD+N (10 kHz) <.01% 0.00128%
www.ti.com
7 Modifications
Increased resistances and lower capacitances may be desired in the filter sections for reasons of cost and
board area. In this case, FET input amplifiers may offer improved performance due to their low input
current noise. However, FET input amplifiers are more susceptible to distortion when placed in a non-
inverting configuration such as a Sallen-Key filter. Therefore, a compromise must be made between cost,
noise, and harmonic distortion. Two FET-input amplifiers which are suitable for this design due to their
exceptional audio performance and availability in quad packages are given in Table 9.
Table 9: FET input amplifiers which offer lower input current noise and are suitable for use with
larger resistance values.
Amplifiers Cost per Vn Power Supply THD+N (1kHz,
Amplifier per Amplifier (nV/rtHz, Current (typ mA / G=+1, 3Vrms,
Package (1ku prices) 1kHz) amplifier) 600 Ohm load)
OPA1644 4 $0.49 5.1 2.3 0.00006%
OPA1652 4 $0.24 4.5 2 0.00008%
8 About the Author

John Caldwell is an applications engineer with Texas Instruments Precision Analog, supporting operational
amplifiers and industrial linear devices. He specializes in precision circuit design for sensors, low-noise
design and measurement, and electromagnetic interference issues. He received his MSEE and BSEE
from Virginia Tech with a research focus on biomedical electronics and instrumentation. Prior to joining TI
in 2010, John worked at Danaher Motion and Ball Aerospace.
www.ti.com
9 References
1. V. Dickason, The Loudspeaker Design Cookbook: Sixth Edition. Audio Amateur Press, 2000.
2. D. Self, The Design of Active Crossover. Elsevier/ Focal Press,2011.
3. D. Self, Small Signal Audio Design. Elsevier/ Focal Press,2010.
4. J. Caldwell. (2013, June 16). Signal Distortion from High-K Ceramic Capacitors. Available:
http://www.edn.com/design/analog/4416466/Signal-distortion-from-high-K-ceramic-
capacitors
5. B. Hofer. Designing for Ultra-Low THD+N. Audio Express, December 2013.
www.ti.com
Appendix A.
A.1 Electrical Schematic
Figure A-1: Electrical Schematic
www.ti.com
A.2 Bill of Materials
Quantity Value Designator Description Manufacturer Manufacturer Part # Supplier Part #

CAP, CERM, 4.7uF,
2 4.7uF C1, C2 16V, +/-10%, X5R, Kemet C1206C475K4PACTU 4520506
1206
C3, C5, C6, C8, CAP, CERM, 0.1uF,
6 0.1uF AVX 0603YC104JAT2A 478-3726-1-ND
C20, C21 16V, +/-5%, X7R, 0603
C4, C7, C9, C10, CAP, CERM, 0.1uF,
10 0.1uF C11, C13, C14, 25V, +/-5%, C0G/NP0, TDK C3216C0G1E104J 445-2691-1-ND
C15, C16, C17 1206
CAP, CERM, 10pF,
1 10pF C12 50V, +/-5%, C0G/NP0, AVX 06035A100JAT2A 478-1163-1-ND
0603
CAP, CERM, 0.047uF,
2 0.047uF C18, C19 50V, +/-5%, C0G/NP0, MuRata GRM31M5C1H473JA01L 490-1764-1-ND
1206
Standard Banana Jack,
3 J1, J2, J3 Keystone 575-4 575-4K-ND
Uninsulated, 5.5mm
RCA Jack, Metal, Right
3 J4, J5, J6 CUI Inc RCJ-011 CP-1400-ND
Angle
RES, 590 ohm, 1%,
2 590 R1, R2 Vishay-Dale CRCW1206590RFKEA 541-590FCT-ND
0.25W, 1206
RES, 1.30k ohm, 1%,
2 1.30k R3, R4 Vishay-Dale CRCW12061K30FKEA 541-1.30KFCT-ND
0.25W, 1206
RES, 1.00k ohm, 1%,
4 1.00k R5, R6, R8, R14 Vishay-Dale CRCW12061K00FKEA 541-1.00KFCT-ND
0.25W, 1206
RES, 422 ohm, 1%,
1 422 R7 Vishay-Dale CRCW1206422RFKEA 541-422FCT-ND
0.25W, 1206
RES, 221 ohm, 1%,
0.25W, 1206
RES, 475 ohm, 1%,
0.25W, 1206
RES, 1.74k ohm, 1%,
1 1.74k R11 Vishay-Dale CRCW12061K74FKEA 541-1.74KFCT-ND
0.25W, 1206
RES, 1.15k ohm, 1%,
0.25W, 1206
RES, 332 ohm, 1%,
0.25W, 1206
RES, 11.8k ohm, 1%,
3 11.8k R15, R16, R17 Vishay-Dale CRCW120611K8FKEA 541-11.8KFCT-ND
0.25W, 1206
RES, 845 ohm, 1%,
2 845 R18, R20 Vishay-Dale CRCW1206845RFKEA 541-845FCT-ND
0.25W, 1206
RES, 1.40k ohm, 1%,
2 1.40k R19, R21 Vishay-Dale CRCW12061K40FKEA 541-1.40KFCT-ND
0.25W, 1206
RES, 10.0k ohm, 1%,
0.25W, 1206
Quad, Low Noise,
Texas
2 U1, U2 Audio Operational OPA1604AIDR 296-30115-1-ND
Instruments
Amplifier
Dual, Low Noise, Audio Texas
1 U3 OPA1602AID 296-28930-5-ND
Operational Amplifier Instruments
Figure A-2: Bill of Materials
www.ti.com
Appendix B. Loudspeaker Dimensions and Drivers
B.1 Loudspeaker Front Panel Dimensions
216 mm
m
m
25
152.4 mm
356 mm
m m
1 78
127 mm
108 mm
B.2 Speaker Drivers Used
Driver Manufacturer Part Number

Woofer Dayton Audio RS180-8
Tweeter Dayton Audio RS28F-4
www.ti.com
Appendix C. Acoustic Measurement Setup
Speaker Calibrated Microphone

Under Test Microphone Preamplifier
1m
Line In
Woofer
Output Line Out Computer with Acoustic
Measurement Software
Full Duplex
Sound Card
Tweeter Crossover
Output Circuit
Audio Power
Amplifier
Figure 42: Conceptual diagram of the equipment used for the acoustic measurements in this document.
The crossover circuit is omitted for raw driver measurements.
Figure 43: The loudspeaker under test with a calibrated measurement microphone placed on the tweeter
axis at a 1 m distance.
IMPORTANT NOTICE FOR TI REFERENCE DESIGNS
Texas Instruments Incorporated ("TI") reference designs are solely intended to assist designers (“Buyers”) who are developing systems that
incorporate TI semiconductor products (also referred to herein as “components”). Buyer understands and agrees that Buyer remains
responsible for using its independent analysis, evaluation and judgment in designing Buyer’s systems and products.
TI reference designs have been created using standard laboratory conditions and engineering practices. TI has not conducted any
testing other than that specifically described in the published documentation for a particular reference design. TI may make
corrections, enhancements, improvements and other changes to its reference designs.
Buyers are authorized to use TI reference designs with the TI component(s) identified in each particular reference design and to modify the
reference design in the development of their end products. HOWEVER, NO OTHER LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL
OR OTHERWISE TO ANY OTHER TI INTELLECTUAL PROPERTY RIGHT, AND NO LICENSE TO ANY THIRD PARTY TECHNOLOGY
OR INTELLECTUAL PROPERTY RIGHT, IS GRANTED HEREIN, including but not limited to any patent right, copyright, mask work right,
or other intellectual property right relating to any combination, machine, or process in which TI components or services are used.
Information published by TI regarding third-party products or services does not constitute a license to use such products or services, or a
warranty or endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual
property of the third party, or a license from TI under the patents or other intellectual property of TI.
TI REFERENCE DESIGNS ARE PROVIDED "AS IS". TI MAKES NO WARRANTIES OR REPRESENTATIONS WITH REGARD TO THE
REFERENCE DESIGNS OR USE OF THE REFERENCE DESIGNS, EXPRESS, IMPLIED OR STATUTORY, INCLUDING ACCURACY OR
COMPLETENESS. TI DISCLAIMS ANY WARRANTY OF TITLE AND ANY IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS
FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT, QUIET POSSESSION, AND NON-INFRINGEMENT OF ANY THIRD PARTY
INTELLECTUAL PROPERTY RIGHTS WITH REGARD TO TI REFERENCE DESIGNS OR USE THEREOF. TI SHALL NOT BE LIABLE
FOR AND SHALL NOT DEFEND OR INDEMNIFY BUYERS AGAINST ANY THIRD PARTY INFRINGEMENT CLAIM THAT RELATES TO
OR IS BASED ON A COMBINATION OF COMPONENTS PROVIDED IN A TI REFERENCE DESIGN. IN NO EVENT SHALL TI BE
LIABLE FOR ANY ACTUAL, SPECIAL, INCIDENTAL, CONSEQUENTIAL OR INDIRECT DAMAGES, HOWEVER CAUSED, ON ANY
THEORY OF LIABILITY AND WHETHER OR NOT TI HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES, ARISING IN
ANY WAY OUT OF TI REFERENCE DESIGNS OR BUYER’S USE OF TI REFERENCE DESIGNS.
TI reserves the right to make corrections, enhancements, improvements and other changes to its semiconductor products and services per
JESD46, latest issue, and to discontinue any product or service per JESD48, latest issue. Buyers should obtain the latest relevant
information before placing orders and should verify that such information is current and complete. All semiconductor products are sold
subject to TI’s terms and conditions of sale supplied at the time of order acknowledgment.
TI warrants performance of its components to the specifications applicable at the time of sale, in accordance with the warranty in TI’s terms
and conditions of sale of semiconductor products. Testing and other quality control techniques for TI components are used to the extent TI
deems necessary to support this warranty. Except where mandated by applicable law, testing of all parameters of each component is not
necessarily performed.
TI assumes no liability for applications assistance or the design of Buyers’ products. Buyers are responsible for their products and
applications using TI components. To minimize the risks associated with Buyers’ products and applications, Buyers should provide
adequate design and operating safeguards.
Reproduction of significant portions of TI information in TI data books, data sheets or reference designs is permissible only if reproduction is
without alteration and is accompanied by all associated warranties, conditions, limitations, and notices. TI is not responsible or liable for
such altered documentation. Information of third parties may be subject to additional restrictions.
Buyer acknowledges and agrees that it is solely responsible for compliance with all legal, regulatory and safety-related requirements
concerning its products, and any use of TI components in its applications, notwithstanding any applications-related information or support
that may be provided by TI. Buyer represents and agrees that it has all the necessary expertise to create and implement safeguards that
anticipate dangerous failures, monitor failures and their consequences, lessen the likelihood of dangerous failures and take appropriate
remedial actions. Buyer will fully indemnify TI and its representatives against any damages arising out of the use of any TI components in
Buyer’s safety-critical applications.
In some cases, TI components may be promoted specifically to facilitate safety-related applications. With such components, TI’s goal is to
help enable customers to design and create their own end-product solutions that meet applicable functional safety standards and
requirements. Nonetheless, such components are subject to these terms.
No TI components are authorized for use in FDA Class III (or similar life-critical medical equipment) unless authorized officers of the parties
have executed an agreement specifically governing such use.
Only those TI components that TI has specifically designated as military grade or “enhanced plastic” are designed and intended for use in
military/aerospace applications or environments. Buyer acknowledges and agrees that any military or aerospace use of TI components that
have not been so designated is solely at Buyer's risk, and Buyer is solely responsible for compliance with all legal and regulatory
requirements in connection with such use.
TI has specifically designated certain components as meeting ISO/TS16949 requirements, mainly for automotive use. In any case of use of
non-designated products, TI will not be responsible for any failure to meet ISO/TS16949.
Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265

Analog, Active Crossover Circuit For Two-Way Loudspeakers: TI Precision Designs: Verified Design

Uploaded by

Copyright:

Available Formats

Analog, Active Crossover Circuit For Two-Way Loudspeakers: TI Precision Designs: Verified Design

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analog, Active Crossover Circuit For Two-Way Loudspeakers: TI Precision Designs: Verified Design

Uploaded by

Copyright:

Available Formats

John Caldwell

TI Precision Designs: Verified Design

TI Precision Designs Circuit Description

Baffle Step 4th-Order Low-

 Supply Voltage: +/- 15 V

 Input and Output Levels: +4dBu / 1.228Vrms

 Crossover Frequency (Acoustic): 1.8kHz

 Loudspeakers and Enclosure Dimensions: See Appendix

Figure 1 depicts the measured transfer function of the design.

Transfer Function of Individual Outputs

Figure 1: Measured Transfer Function of the crossover network.

C14 R14 R12

Figure 2: Complete crossover network schematic with functional portions highlighted

2.1 Baffle Step Compensation

Woofer Frequency Response

Baffle Diffraction and Room Interaction Effects

Woofer Frequency Response with Baffle Step Compensation

2.2 Low Pass Filter

Target Woofer Response

-60 Woofer Response w/BSC

Filtered Woofer Frequency Response

-50 Woofer Response w/BSC

2.3 High Pass Filter

Filtered Tweeter Frequency Response (Predicted)

2.4 All Pass Filter

Group Delay of Woofer and Tweeter Networks

50 Woofer Group Delay

- U2B R9 C13 VEE

2.5 Level Control and Buffering

Relative Output of Individual Drivers

Figure 21: Tweeter level correction and output buffer circuit.

Predicted System Frequency Response

Woofer Response Tweeter Response (8dB attenuation) System Response

3.2 Passive Components

4.1 Woofer Signal Path

Figure 24: Tina-TI™ simulation schematic of the woofer signal path

4.1.1 Transfer Function

4.2 Tweeter Signal Path

Figure 27: TINA-TI™ simulation schematic of the tweeter signal path.

4.2.1 Transfer Function

4.2.3 Tweeter Signal Delay

Figure 31: Tina-TI™ simulation of the all-pass filter group delay.

4.3 Simulated Results Summary

5.1 PCB Layout

6 Verification & Measured Performance

6.1 Electrical Measurements

6.1.1 Woofer Signal Path Transfer Function

Woofer Signal Path Transfer Function (Measured)

Figure 33: Measured transfer function of the woofer signal path

6.1.2 THD+N Performance – Woofer Output

Woofer Signal Path THD+N (1.228 Vrms Input, 80kHz)

Woofer Signal Path

Figure 34: Measured THD+N performance of the woofer signal path

6.1.3 Tweeter Signal Path Transfer Function

Tweeter Signal Path Transfer Function (Measured)

Figure 35: Measured transfer function of the tweeter signal path.

6.1.4 Tweeter Signal Path THD+N

Tweeter Signal Path THD+N (1.228 Vrms Input, 80 kHz)

Tweeter Signal Path