AN144: Audio Xcellence: XDCP Signal Integrity Report (Audio Characterization Report)

Audio Xcellence: XDCP Signal Integrity Report
(Audio Characterization Report)
®
Application Note
Introduction
Most audio equipment testing is done on a stimulusresponse basis. A signal of known characteristics is fed to
the input of the Device Under Test (DUT) and the output of
the DUT is measured. The performance of the DUT is
determined by degradation of the output signal from the
known input signal. Often, sets of measurements are made
as the stimulus is swept or stepped across the audio
frequency spectrum or across an amplitude range, and the
desired performance information is determined by the
relationship between the corresponding set of output
measurements.
The most common stimulus for audio testing is a sine wave.
The sine wave is unique since it is the only signal to have all
its energy concentrated at a single point in the frequency
spectrum, referred to as the fundamental frequency. It is
therefore relatively simple to analyze test results from single
sine wave testing. Multiple sine waves (usually two) are used
for intermodulation distortion testing, and large numbers of
multiple sine waves are used for some new testing
July 8, 2005
AN144.0
techniques. White noise, pink noise, square waves, and
impulses may also be used as stimulus for certain types of
audio testing. It is also possible to make certain
measurements using program material such as music or
voice as stimulus.
All three of the Intersil XDCPs used in this noise report were
subjected to the testing criteria outlined in the above two
paragraphs.
Table of Contents
Signal to Noise Ratio
2
Total Harmonic Distortion- THD+N,& Dual Tone Tests
4
Frequency Response
8
Mutli-Channel Crosstalk & Level Matching
10
Measurement Results Summary
11
Additional Test Results
12
Appendix
14
Test Setup Using the Audio Precision System 2
Benchtop
Power Supply
PC for AP2 Control and
Data Display/Storage
+5VDC
AP2 Interface Adapter Card
–5VDC
GND
Analog Test Signal Out
-60dB at 1kHz
BNC
OUT
AP2Audio Precision
System 2
BNC
IN
BNC
GND
Test Analog Signal In
VIN (1-4)
VW (1-4)
BNC Signal Ground
VL (1-4)
Device
Under
Testx9408
INTERSIL
SDCP
DEVELOPMENT
& EVALUATION
BOARD
FIGURE 1. AUDIO MEASUREMENT TEST SETUP
1
CAUTION: These devices are sensitive to electrostatic discharge; follow proper IC Handling Procedures.
1-888-INTERSIL or 1-888-468-3774 | Intersil (and design) is a registered trademark of Intersil Americas Inc.
Copyright Intersil Americas Inc. 2005. All Rights Reserved
All other trademarks mentioned are the property of their respective owners.
Application Note 144
Signal to Noise Ratio (SNR)
Applications for these parts include filters, analog signal
processing, linear level control, and in any application where
audio frequency analog signals are present and signal
integrity and low noise is important.
Signal to Noise Ratio (SNR) is the ratio of the normal
operating level of the device compared with the device’s
noise floor. SNR can be thought of as the effective dynamic
signal range of most types of analog audio devices.
SNR is determined by injecting a 0dB, 1kHz sine wave
(generated by the AP2*) into the input of the Device Under
Test (DUT). The signal is processed by the DUT and it’s
output signal is then fed back into the analyzer section of the
AP2. The AP2 is then adjusted to set up a 0dB reference for
this signal. The output of the AP2 is enabled again but this
time the 1kHz sine wave is removed. The output of the DUT
is analyzed by the AP2 again and the ratio between this
baseline noise level with no input signal and the original 0dB
level set in the first measurement is the SNR.
The Signal to Noise Ratios (SNR) of the X9408, X9241A,
X9C102 parts are all exceptional. All three parts exceeded
119dB SNR in the audio frequency spectrum (20Hz to
20kHz). (See graphs for all three parts and all
measurements at the end of this document). The higher the
SNR values the quieter the system will be, consequently,
low-level detail will become audible. Dropping the noise floor
by 10dB on the low end has the same effect as if the level
control were turned up by the same amount. The SNR plots
are found on the following graphs.
*AP2-Audio Precision System 2
A 1kHz Sine Wave at –60 dB Signal Plotted From 20Hz to 20kHz
+0
-20
-40
-60
d
B
r
-80
-100
A
-120
-140
-160
-180
20
50
100
200
500
1K
2K
Color
Line Style
Thick
Data
Axis
Blue
Solid
1
Fft. Ch. 1 Ampl
Left
5K
10K
20K
SNR Reads 116 dB on the AP2 Meter
XDCP9408 set at step 63 on chan 1
Global Specialties power supply #1310
2
AN144.0
July 8, 2005
Application Note 144
CHAN 0
Audio Precision Intersil X9408 SNR CHAN 0
+0
d
B
r
-100
A
-150
-50
20
50
100
200
500
1k
2k
5k
10k
20k
Hz
Color
Line Style
Thick
Data
Axis
Cyan
Blue
Solid
Solid
3
3
Fft.Ch.1 Ampl
Fft.Ch.1 Ampl
Left
Left
X9408w(10K ohms) SNR=120dB on Chan 0
1kHz@0dB Set Ref (from 1.4vp@0dB)
Step 63(Full On)
Audio Precision
XICOR X9241 SNR
CHANNEL 0
05/20/99 19:24:51
AUDIO PRECISION INTERSIL X9241A SNR CHANNEL 0
+0
-20
-40
-60
d
B
r
-80
-100
A
-120
-140
-160
-180
20
50
100
200
500
1k
2k
5k
10k 20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Cyan
Solid
Solid
3
3
Fft.Ch.1 Ampl
Fft.Ch.1 Ampl
Left
Left
X9241y(2k ohms) SNR=119 dB on Chan 0
1kHz@0dB Set Ref (from 1.4vp@0dB)
Step 63(Full On)
MBPWR
3
AN144.0
July 8, 2005
Application Note 144
AUDIO PRECISION INTERSIL X9C102 SNR
+0
-20
-40
d
B
r
A
-60
-80
-100
-120
-140
-160
-180
20
50
100
200
500
1k
2k
5k
10k 20k
Hz
Color
Line Style
Thick
Data
Axis
Cyan
Blue
Solid
Solid
3
3
Fft.Ch.1 Ampl
Fft.Ch.1 Ampl
Left
Left
X9C102(1K ohms) SNR= 119dB
1kHz@0dB Set Ref (from 1.4vp@0dB)
Step 99(Full On)
MBPWR
Total Harmonic Distortion- THD+N, & Dual
Tone Tests
The THD+N (Total Harmonic Distortion + Noise) level for all
three parts is below -100dB which is exceptional for this
class of parts. (See graphs for all three parts and all
measurements at the end of this document). The dual tone
type of tests are less common than the THD+N tests but are
valuable in revealing problems with intermodulation
distortion. In the parts that were measured with two lone
tests at various frequencies generated no harmonics greater
that -105dB.
In particular these two specifications define a parts audio
frequency performance quality and assure a circuit designer
that all three will work well as analog signal attenuators
without adding extra harmonics. The results of these tests
clearly show that all three of the Intersil parts exhibit
excellent low level harmonic tendencies.Applications for
these parts include filters, analog signal processing, linear
level control, and in any application where audio frequency
analog signals are present and need to be easily controlled
and where signal integrity is important.
4
Total Harmonic Distortion + Noise (THD+N): Measured by
attenuating the fundamental signal (lkHz @ 0dB) with a
narrow-band notch filter, then measuring the remaining
signals which consist of harmonics of various orders, wideband noise, and possibly interfering signals. THD causes
audio signals to become smeared and less clear/focused. It
can also raise baseline noise levels which masks out low
level audio information. Creating extra signals and adding
them back into the final output is always a BAD Thing.
Two Tone Harmonic Test: Where two 0dB signals at two
different frequencies are fed into the DUT and the output is
then measured by the AP2 across the entire audio frequency
range (from 20Hz to 20kHz). Any harmonics (extra signals at
even or odd multiples of the original two signals and
multiples of their differences as well) created from the
injection of these two signals into the DUT are displayed on
a log frequency vs. amplitude graph. This test is useful in
revealing any higher audio frequency anomalies which can
be created by bad layout, non-linear responses, stray
capacitance, and other circuit abnormalities. The audibility of
this type of distortion is another form of THD. Plots of
THD+N arc are found on this and the following 3 pages.
AN144.0
July 8, 2005
Application Note 144
AUDIO PRECISION INTERSIL X9408 THD+N CHAN 0
+0
-20
-40
d
B
r
-60
-80
A
-100
-120
-140
20
50
100
200
500
1k
2k
5k
10k
20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Fft.Ch.1 Ampl
Left
X9408w(10K ohms) CHAN 0
THD+N = 106dB
1kHz@ 1.4vp-0dB
Step 63(last step) - AP2 Z @ 100K
MBPWR
CHAN 0
AUDIO PRECISION INTERSIL X9241A THD+N CHAN 0
+0
-20
-40
d
B
r
A
-60
-80
-100
-120
-140
-160
-180
20
50
100
200
500
1k
2k
5k
10k 20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Fft.Ch.1 Ampl
Left
X9241y(2K ohms) THD+N @ 1kHz @ 0dB=101dB
Channel 0 Set @ Step 63
AP2 100k Input Z, 5000 pts.
MBPWR
5
AN144.0
July 8, 2005
Application Note 144
AUDIO PRECISION INTERSIL X9C102 THD+N
+0
-20
-40
d
B
r
A
-60
-80
-100
-120
-140
-160
-180
20
50
100
200
500
1k
2k
5k
10k 20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Fft.Ch.1 Ampl
Left
X9C102(1K ohms) THD+N = -102dB
1kHz@OdB (from 1.4vp@0dB)
Step 99
MBPwr
IMD TEST 18-19kHz CHAN
0
AUDIO PRECISION INTERSIL X9408 DUAL TONE IMD TEST
18-19kHz CHAN 0
+0
-20
d
B
r
A
-40
-60
-80
-100
-120
-140
20
50
100
200
500
1k
2k
5k
10k
20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Fft.Ch.1 Ampl
Left
X9408w(10K ohms) DUAL TONE TEST
CHAN 0
1819kHz@OdB (from 1.4vp@0dB)
AP2 Input Z = 100k ohms
Step 63
6
AN144.0
July 8, 2005
Application Note 144
CHAN 0
AUDIO PRECISION INTERSIL X9241A DUAL TONE IMD TEST CHAN 0
+0
-20
d
B
r
A
-40
-60
-80
-100
-120
-140
20
50
100
200
500
1k
2k
5k
10k
20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Blue
Solid
Solid
3
3
Fft.Ch.1 Ampl
Fft.Ch.1 Ampl
Left
Left
X9241y(2K ohms) Two Tone Test
18 19kHz @ 0dB
channel 0, Step 63
MBPWR
Audio Precision
XIC0R X9C102 DUAL TONE IMD
TEST
06/18/99 11:02:00
AUDIO PRECISION INTERSIL X9C102 DUAL TONE IMD TEST
+0
-20
-40
d
B
r
A
-60
-80
-100
-120
-140
20
50
100
200
500
1k
2k
5k
10k
20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Blue
Solid
Solid
3
3
Fft.Ch.1 Ampl
Fft.Ch.1 Ampl
Left
Left
X9C102(1K ohms) Two Tone Test
18 19kHz @ 0dB
Step 99
MBPWR
7
AN144.0
July 8, 2005
Application Note 144
FREQUENCY RESPONSE
bandwidth of a standard 20kHz audio signal). The AP2 plots
of the frequency response, of the three parts being tested,
are found in the following 2 pages.
The upper frequency response limit of an analog system is
usually determined by the point at which the input signal is
reduced by the DUT (the circuit or part being tested) on it’s
output by a factor of –3dB. This standard of –3dB down
corresponds to 1/2 power or 0.707 times the voltage of the
input signal. All the Intersil parts tested (using the Audio
Precision System 2) were flat to 200kHz, which is the upper
measurement frequency limit of the AP2 (10 times the
These three parts therefore have no frequency based
limitations in the audio frequency range which make them
suitable for any application that requires signal level
attenuation such as low frequency square wave generators,
data loggers, vibration analyzers, noise cancellation
equipment, environmental controls, etc.
Audio Precision
XICOR X9408 CHAN 0 Frequency
05/26/99 14:40:26
Response
AUDIO PRECISION INTERSIL X9408 CHAN 0 FREQUENCY RESPONSE
+2.5
+2
+1.5
+1
d
B
r
A
+0.5
+0
-0.5
-1
-1.5
-2
20
50
100
200
500
1k
2k
5k
10k
20k
50k
200k
Hz
Color Line Style
Thick Data
Blue
3
Solid
Axis
Anlr.Level A Left
X9408W(10K ohm) Frequency Response
1.4vp 20Hz to 200 kHz on chan 0
AP2 Input Z = 100k ohms
MBPWR
8
AN144.0
July 8, 2005
Application Note 144
Response
AUDIO PRECISION INTERSIL
X9241A CHAN 0 FREQUENCY RESPONSE
+2.5
+2
+1.5
+1
d
B
r
A
+0.5
+0
-0.5
-1
-1.5
-2
20
50
100
200
500
1k
2k
5k
10k
20k
50k
200k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Anlr.Level A
Left
X241y(2K ohms) Frequency Response
1.4vp 20Hz to 200 kHz on chan 0
AP2 Input Z = 100k ohms
MBPWR
Audio Precision
XICOR X9C102 FREQUENCY
RESPONSE
05/26/99 14:28:00
AUDIO PRECISION INTERSIL X9C102 FREQUENCY RESPONSE
+2.5
+2
+1.5
+1
d
B
r
A
+0.5
+0
-0.5
-1
-1.5
-2
20
50
100
200
500
1k
2k
5k
10k
20k
50k
200k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Anlr.Level A
Left
X9C102(1K ohms) FREQUENCY RESPONSE
0 dB -- 1.4VP 20Hz TO 200kHz
STEP 99
MBPWR
9
AN144.0
July 8, 2005
Application Note 144
Multi-Channel Crosstalk & Level Matching
equals 1 bit in the digital representation of an analog audio
signal so a number of -120dB corresponds to a 20 bit
resolution which is present limit of high-end audio signals.
On multi-channel parts like the X924lA and X9408 any
interaction between channels where the signal on one
channel leaks into the other is undesirable and referred
to as crosstalk. To measure crosstalk a 0dB 1kHz signal is
inserted into one channel and the other adjacent
channel/channels are left floating with no signal present. The
floating channels are then measured for any signal content
that leaked over from the channel being driven with a signal.
The level of the signals measured on the floating channels
referenced to the 0dB, 1kHz signal channel is the
“Separation” specification number expressed in negative dB.
The more negative the number the better the circuit.
Where as a high negative number for crosstalk is a very
good thing the opposite is true for level matching between
channels on multi-channel parts. In this test the same test
signal is fed into all the channels of the OUT at the same
time. Then each of the individual channel output signals are
measured to insure the output signal level of all of the
channels match to within a certain dB level (usually 0.5 to
0.05dB in consumer audio equipment). Both of the multichannel Intersil parts (X9408 & X9241A) were exceptional
for this measurement.
Both of the Intersil parts tested measured very well for
crosstalk. The crosstalk numbers expressed in negative dB
were over minus –116dB on channels that were at opposite
ends of the chip and over -85dB on directly adjacent
channels. In audio applications any noise or separation
number over -120dB is considered excellent. Every 6dB
The X9408 measured slightly better than the X9241A but the
difference was so small as to be rendered meaningless. The
X9408 worst case channel match between two of the four
channels was .09dB and for the X9241A it was 0.ldB. Most
of the channels actually matched within 0.01dB!
Separation
Test Chan 0 to
2
AUDIO PRECISION INTERSIL X9408 SEPARATION TEST CHAN 0 TO 2
d
B
r
+0
-100
A
20
50
100
200
500
1k
2k
5k
10k 20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Fft.Ch.1 Ampl
Left
X9408w(10K ohms) Chan 0 to 2 SEP = -110dB
1kHz @ 0dB on 1st chan measuring wiper of 2sd channel. RL gr
MBPWR
10
AN144.0
July 8, 2005
Application Note 144
TO 2X9241A CROSSTALK CHAN 0 TO 2
AUDIO PRECISION INTERSIL
+0
-20
-40
d
B
r
A
-60
-80
-100
-120
-140
-160
-180
20
50
100
200
500
1k
2k
5k
10k 20k
Hz
Color
Line Style
Thick
Data
Axis
Blue
Solid
3
Fft.Ch.1 Ampl
Left
X9241y(2k ohms) Chan 0 to 2 SEP = -112dB
1kHz @ 0dB on 1st chan measuring wiper of 2sd channel
RL grounded on all Channels.
AP2@100k MBPWR
Measurement Results Summary
X9408
1. Signal to Noise Ratio (SNR):
-120dB (AP2 Calc.)
2. Total Harmonic Distortion + Noise (THD+N):
-106dB (AP2 Calc.)
3. Frequency Response 20Hz to 20kHz:
± .05dB
4. Two Tone Harmonic Test (18 + 19kHz):
-122/-119dB at 1kHz
5. Listen for Zipper noise (Must be below audibility):
PASSED
6. Channel Separation Range:
-110/-85dB
X9241A
1. Signal to Noise Ratio (SNR):
-119dB (AP2 Calc.)
2. Total Harmonic Distortion + Noise (THD+N):
-102dB (AP2 Calc.)
3. Frequency Response 20Hz to 20kHz:
± .05dB
4. Two Tone Harmonic Test (18 + 19kHz):
-108/-99 dB at 1kHz
5. Listen for Zipper noise (Must be below audibility):
PASSED
6. Channel Separation:
-116/-106dB
X9C102
1. Signal to Noise Ratio (SNR):
-119dB (AP2 Calc.)
2. Total Harmonic Distortion + Noise (THD+N):
-103dB (AP2 Calc.)
3. Frequency Response 20 H~ to 20kHz:
± .05dB
4. Two Tone harmonic Test (18 + 19kHz):
-103dB at 1kHz
5. Listen for Zipper noise (Must be below audibility):
PASSED
11
AN144.0
July 8, 2005
Application Note 144
X9408w Test Results & Observations
Channel Matching
The X9408 was the best Intersil part when evaluated for
Frequency Response, Noise (SNR and THD+N) and Channel
Separation. Below are comments on each of the
measurements taken and the effect of each on the audio
performance of the part. It should be noted that none of these
measurements were weighted.
A series of measurements were made with all four channels at
different steps (volume levels, see graphs). The worst case
measurement at the lowest level setting of –36dB was ± .09dB
for all four channels to each other and the other five level
settings were better than 0.02dB! This was better than the other
multi-channel part in the test, the X9241A, by a wide margin.
This is also as good as any multi-channel volume control on the
market today.
Frequency Response
The X9408 exhibited excellent frequency response all the way
to the limit of the AP2, 200kHz. The response to 20kHz (the
normal audio frequency measurement range limit) is
outstanding, ± .005dB. Ten times bandwidth to -3dB is a good
rule of thumb when designing high quality audio circuits and the
X9408 exceeded that requirement being only ± 0.05dB down at
200kHz!
THD+N
The THD+N for the four channels on the X9408 with a 1kHz 0
dB reference signal was between 106 and 107dB for all four
channels! As you look at the X9408w graph on page 3, notice
that there is only one harmonic at 3kHz above -120dB. This
level of performance makes this part suitable for all types of
consumer audio applications.
Conclusions
The X9408 exhibits excellent audio frequency performance. In
any application where a linear taper level control along with low
noise and low harmonic distortion are required, the X9408 is a
top notch performer.
X9241A Test Results & Observations
The X9241A was the second best Intersil part when evaluated
for Frequency Response, Noise (SNR and THD+N) and
Channel Separation. Below are comments on each of the
measurements taken and the effect of each on the audio
performance of the part. It should be noted that none of these
measurements were weighted.
SNR
Frequency Response
The SNR on the four channels of the X9408 ranged from 118dB
to 120dB. In the SNR graphs it can be seen that the dominant
noise in the test setup was at 60Hz. If you notch out that signal
then the SNR for the four channels falls below 150dB! In either
case this spec is more than sufficient for any consumer audio
application including high end equipment that uses up to 20 bits
of resolution.
The X9241A exhibited excellent frequency response all the way
to the limit of the AP2, 200kHz. The response to 20kHz (the
normal audio frequency measurement range limit) is
outstanding, ± .005dB. Ten times bandwidth to -3dB is a good
rule of thumb when designing high quality audio circuits and the
X9241A exceeded that requirement being only ± 0.05dB down
at 200kHz!
Two Tone Test
THD+N
Two 0 dB signals at different frequencies are fed into the DUT
and the output is measured by the AP2 from 20Hz to 20kHz.
Any harmonics created from the injection of these two signals
into the DUT are displayed on a log frequency vs. amplitude
graph. The graphs of the four channels show the harmonics
created in this test were very low level, on the order of –110dB!
The THD+N for the four channels on the X9241A with a 1kHz
0dB reference signal was between 101 and 102dB for all four
channels! As you look at the X9241A graph on page 4, notice
that there is only one harmonic at 3kHz above -120dB. This
level of performance makes this part suitable for all types of
consumer audio applications.
Crosstalk Noise (i.e. Channel Separation)
SNR
The level of channel separation on the X9408 depends on
which channels are considered. Adjacent channels have less
separation (worse number) than channels at opposite ends of
the part. The worst measurement was between channels 2 and
3, -85dB, and the best was between channels 2 and 0, -110dB.
With proper spacing and layout –110dB is achievable now and
is acceptable for nearly all consumer audio.
The SNR measured –119dB on four channels of the X9241A.
In the SNR graphs it can be seen that the dominant noise in the
test setup was at 60Hz. If you notch out that signal then the
SNR for the four channels falls below 150dB! In either case this
spec is more than sufficient for any consumer audio application
including high end equipment that uses up to 20 bits of
resolution.
Level Switching Noise Listening Test
Two Tone Test
The X9408 was inserted as a volume control (attenuator to
ground) into the signal path of a very high end preamp that
directly fed a set of ultra sensitive headphones (Grado). No
level switching noise was detected by the listener at all levels.
Two 0dB signals at different frequencies are fed into the
DUT and the output is measured by the AP2 from 20Hz to
20kHz. Any harmonics created from the injection of these
two signals into the DUT are displayed on a log frequency
vs. amplitude graph. The graphs of the four channels show
12
AN144.0
July 8, 2005
Application Note 144
the harmonics created in this test were very low level, on the
order of –110dB!
Crosstalk Noise (i.e. Channel Separation)
The level of channel separation on the X9241A depends on
which channels are considered. Adjacent channels have
less separation (worse number) than channels at opposite
ends of the part. The worst measurement was between
channels 2 and 3, -106dB (-21dB better than the X9408),
and the best was between channels 2 and 0, -116dB (-6dB
better than X9408). With proper spacing and layout –116dB
is achievable now and is acceptable for nearly all consumer
audio.
Level Switching Noise Listening Test
The X9241A was inserted as a volume control (attenuator to
ground) into the signal path of a very high end preamp that
directly fed a set of ultra sensitive headphones (Grado). No
level switching noise was detected by the listener at all
levels.
Channel Matching
A series of measurements were made with all four channels
at different steps (volume levels, see graphs). The worst
case measurement was ± 0.1dB for all four channels to each
other and the best case was 0.02dB! This is also as good as
any multi-channel volume control on the market today.
Conclusions
The X9241A exhibits excellent audio frequency
performance. In any application where a linear taper level
control along with low noise and low harmonic distortion are
required, the X9241A is a top notch performer.
X9C102 Test Results & Observations
The X9C102 masurements are discussed below with
comments on each of the measurements taken and the
effect of each on the audio performance of the part. It should
be noted that none of these measurements were weighted.
Frequency Response
The X9C102 exhibited excellent frequency response all the
way to the limit of the AP2, 200kHz. The response to 20kHz
(the normal audio frequency measurement range limit) is
outstanding, ± 0.005dB. Ten times bandwidth to -3dB is a
good rule of thumb when designing high quality audio
circuits and the X9C102 exceeded that requirement being
only ± 0.05dB down at 200kHz!
THD+N
The THD+N for X9C102 with a 1kHz 0dB reference signal
was between 103dB. As you look at the X9C102 graph on
page 5, notice that there is only one harmonic at 3kHz above
-120dB. This level of performance makes this part suitable
for all types of consumer audio applications.
SNR
The SNR measured –119dB on the X9C102. In the SNR
graphs it can be seen that the dominant noise in the test
setup was at 60Hz. This spec is more than sufficient for any
consumer audio application including high end equipment
that uses up to 20 bits of resolution.
Two Tone Test
Two 0 dB signals at different frequencies are fed into the DUT
and the output is measured by the AP2 from 20Hz to 20kHz.
Any harmonics created from the injection of these two signals
into the DUT are displayed on a log frequency vs. amplitude
graph. The graphs show the harmonics created in this test
were very low level, on the order of –103dB!
Level Switching Noise Listening Test
The X9C102 was inserted as a volume control (attenuator to
ground) into the signal path of a very high end preamp that
directly fed a set of ultra sensitive headphones (Grado). No
level switching noise was detected by the listener at all
levels.
Conclusions
The X9C102 exhibits excellent audio frequency
performance. In any application where a linear taper level
control along with low noise and low harmonic distortion are
required, the X9C102 is a top notch performer.
13
AN144.0
July 8, 2005
Application Note 144
Appendix
AUDIO PRECISION INTERSIL X9408 4 CHANNEL LEVEL MATCHING
AT SIX DIFFERENT LEVEL SETTINGS
+0
d
B
r
-20
A
-40
2k
4k
6k
8k
10k
12k
14k
16k
18k
20k
Hz
Color
Line Style
Thick
Data
Axis
Cyan
Green
Yellow
Solid
Solid
Solid
1
1
1
Anlr.LevelA
Anlr.LevelA
Anlr.LevelA
Left
Left
Left
X9408 Frequency Response ALL 4 CHANS
STEP 31 = -6dB, STEP20=-10dB, STEP 10=--16dB, STEP5=-22dB, STEP 2=--30dB, STEP 1=-38dB.
LAST STEP PURPLE =3, BLUE=2, CYAN=1, GREEN=0
Spread @ -6dB = ±.02dB, @ -10dB = ±.015dB, @ -16dB = ±.01dB,
@-22dB = ±.01dB, @-30dB = ±.012dB, @-36dB = ±.09dB,
1.4vp20Hz to 200kHz on chan 0
AP2 Input Z = 100kΩ, MB PWR
AUDIO PRECISION INTERSIL X9241A ALL CHAN FREQUENCY RESPONSE
+5
+0
-5
-10
d
B
V
-15
-20
-25
-30
-35
-40
20
50
100
200
500
1k
2k
Hz
Color
Line Style
Thick
Data
Axis
Cyan
Green
Yellow
Solid
Solid
Solid
3
3
3
Anlr.LevelA
Anlr.LevelA
Anlr.LevelA
Left
Left
Left
5k
10k
20k
50k
200k
X241y(2kΩ) Frequency Response ALL 4 CHANNELS TOGETHER - STEP 10
0=CYAN, 1=GREEN, 2=YELLOW, 3=RED
1.4vp20Hz to 200kHz on chan 3
AP2 Input Z = 100kΩ,
MB PWR
14
AN144.0
July 8, 2005
Application Note 144
Appendix
Audio Precision
XICOR X9241 ALL CHANS Frequency
05/21/99 11:09:19
Response
AUDIO PRECISION INTERSIL
X9241A ALL CHANS FREQUENCY RESPONSE
AT Six Different Level Settings
+ 0.4
+ 0.2
d
B
r
A
-0
-0.2
-0.4
20
50
100
200
500
1k
2k
Hz
Color
Line Style
Thick
Data
Axis
Cyan
Green
Yellow
Solid
Solid
Solid
3
3
3
Anlr.LevelA
Anlr.LevelA
Anlr.LevelA
Left
Left
Left
5k
10k
20k
50k
200k
X241y(2kΩ) Frequency Response ALL 4 CHANNELS TOGETHER
0=CYAN, 1=GREEN, 2=YELLOW, 3=RED
1.4vp20Hz to 200kHz on chan 3 @ step 63
AP2 Input Z = 100kΩ,
MB PWR
X9408-Total Harm onic Distortion + Noise
vs. In put Voltage & DCP Position
0.0
-20.0
-40.0
1vPeak @ 1kH z
-60.0
2vPeak @ 1kH z
3vPeak @ 1kH z
-80.0
4vPeak @ 1kH z
5vPeak @ 1kH z
-100.0
-120.0
63
60
55
50
45 40
35
30
25
20
15
10
5
4
3
2
1
D C P P osition #
For more information please refer to datasheets X9408, X9421 and X9C102/103/104/503 at www.intersil.com
Intersil Corporation reserves the right to make changes in circuit design, software and/or specifications at any time without notice. Accordingly, the reader is cautioned to
verify that the Application Note or Technical Brief is current before proceeding.
For information regarding Intersil Corporation and its products, see www.intersil.com
15
AN144.0
July 8, 2005