AD AD8117-EVAL

500 MHz, 32 x 32 Buffered
Video Crosspoint Switch
AD8117/AD8118
Preliminary Technical Data
FUNCTIONAL BLOCK DIAGRAM
FEATURES
Large, 32 x 32 High Speed, Nonblocking Switch Array
G = 1 (AD8117) or G = 2 (AD8118) Operation
Differential or Single-Ended Operation
Single +5 V supply, or dual ± 2.5 V supply
Serial or Parallel Programming of Switch Array
High impedance output disable allows connection of
multiple devices with minimal output bus load
Excellent Video Performance
100 MHz 0.1 dB Gain Flatness
0.1% Differential Gain Error (RL = 150 Ω)
0.1° Differential Phase Error (RL = 150 Ω)
Excellent AC Performance
Bandwidth: >500 MHz
Slew rate: 1,800 V/µs
Low power of 2.5 W
Low all hostile crosstalk:
-75 dB @ 5 MHz
-40 dB @ 500 MHz
Reset pin allows disabling of all outputs
(Connected through a capacitor to ground provides
power-on reset capability)
304 ball SBGA package (31 mm × 31 mm)
SER/PAR
D0 D1 D2 D3 D4 D5
VDD
DGND
WE
A0
A1
A2
A3
A4
1
192-BIT SHIFT REGISTER
WITH 6-BIT
PARALLEL LOADING
0
CLK
DATA IN
DATA
OUT
192
UPDATE
RESET
PARALLEL LATCH
AD8117
(AD8118)
192
32
x
DECODE
6:32 DECODERS
INPUT
RECEIVER
G = +1
(G = +2)
1024
32
OUTPUT
BUFFER
G = +1
2
2
SWITCH
MATRIX
32
OUTPUT
PAIRS
32 INPUT
PAIRS
APPLICATIONS
Routing of high speed signals including:
RGB and component video routing
Compressed video (MPEG, Wavelet)
Data communications
PRODUCT DESCRIPTION
The AD8117/AD8118 is a high speed 32 × 32 video crosspoint
switch matrix. It offers a 500 MHz bandwidth and slew rate of
1800 V/µs for high resolution computer graphics (RGB) signal
switching. With −75 dB of crosstalk and −100 dB isolation (@
5 MHz), the AD8117 is useful in many high-speed applications.
The 0.1 dB flatness out to 100 MHz makes the AD8117 ideal for
composite video switching.
The AD8117's 32 independent output buffers can be placed into
a high impedance state for paralleling crosspoint outputs so that
off-channels present minimal loading to an output bus. The
AD8117 is available in gain of 1 or 2 (AD8118) for ease of use in
VPOS VNEG
VOCM
Figure 1. AD8117 G = +1
back-terminated load applications. It operates as a fully
differential device or can be configured for single-ended
operation. Either a single +5 V supply, or dual ± 2.5 V supplies
can be used while consuming only 500 mA of idle current with
all outputs enabled. The channel switching is performed via a
double-buffered, serial digital control (which can accommodate
daisy chaining of several devices) or via a parallel control
allowing updating of an individual output without
reprogramming the entire array.
The AD8117/AD8118 is packaged in a 304 Ball BGA package
and is available over the extended industrial temperature range
of −40°C to +85°C.
Rev. PrA
Information furnished by Analog Devices is believed to be accurate and reliable.
However, no responsibility is assumed by Analog Devices for its use, nor for any
infringements of patents or other rights of third parties that may result from its use.
Specifications subject to change without notice. No license is granted by implication
or otherwise under any patent or patent rights of Analog Devices. Trademarks and
registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
www.analog.com
Fax: 781.461.3113
© 2006 Analog Devices, Inc. All rights reserved.
AD8117/AD8118
Preliminary Technical Data
TABLE OF CONTENTS
AD8117 Specifications..................................................................... 3
Typical Performance Characteristics ........................................... 15
Timing Characteristics (Serial Mode) ........................................... 5
Theory of Operation ...................................................................... 18
Timing Characteristics (Parallel Mode) ........................................ 6
Applications..................................................................................... 19
Absolute Maximum Ratings............................................................ 7
Programming.............................................................................. 19
Thermal Resistance ...................................................................... 7
Operating Modes........................................................................ 20
Power Dissipation......................................................................... 7
Outline Dimensions ....................................................................... 31
ESD Caution.................................................................................. 7
Ordering Guide .......................................................................... 31
Pin Configurations and Function Descriptions ........................... 8
REVISION HISTORY
Revision PrA: Preliminary Datasheet
Rev. PrA | Page 2 of 32
Preliminary Technical Data
AD8117/AD8118
AD8117 SPECIFICATIONS
VS = ± 2.5 V at TA = 25°C, G = +1, RL = 100 Ω, Differential I/O mode, unless otherwise noted.
Table 1. AD8117ABPZ
Parameter
DYNAMIC PERFORMANCE
−3 dB bandwidth
Gain flatness
Propagation delay
Settling time
Slew rate
NOISE/DISTORTION
PERFORMANCE
Differential Gain Error
Differential Phase Error
Crosstalk, all hostile
Off isolation, input-output
Input Voltage Noise
DC PERFORMANCE
Gain Error
Gain Matching
OUTPUT CHARACTERISTICS
Output impedance
Output disable capacitance
Output leakage current
Output voltage range
INPUT CHARACTERISTICS
Input offset voltage
Input Voltage Range Common Mode
Input Voltage Range Differential Mode
Common-mode rejection
ratio
Input capacitance
Input resistance
Input bias current
SWITCHING CHARACTERISTICS
Enable on time
Switching time, 2 V step
Switching transient (glitch)
Conditions
Min
Typ
Max
Unit
200 mV p-p, RL = 100 Ω
2 V p-p, RL = 100 Ω
0.1 dB, 200 mV p-p, RL = 100 Ω
0.1 dB, 2 V p-p, RL = 100 Ω
2 V p-p, RL =100 Ω
1% , 2 V step, RL = 100 Ω
2 V Step, RL = 100 Ω, peak
2 V Step, RL = 100 Ω, 10-90%
>500
>420
100
70
1.3
2.5
1,800
1,500
MHz
MHz
MHz
MHz
ns
ns
V/µs
V/µs
NTSC or PAL, RL = 150 Ω or RL = 1kΩ
NTSC or PAL, RL = 150 Ω or RL = 1kΩ
ƒ = 5 MHz
ƒ = 10 MHz
ƒ = 100 MHz
ƒ = 500 MHz
ƒ = 10 MHz, RL = 100 Ω, one channel
0.01 MHz to 50 MHz
0.1
0.1
–75
–70
–50
–40
−100
45
%
Degrees
dB
dB
dB
dB
dB
nV/√Hz
RL = 100 Ω or 150 Ω
No Load, Channel-Channel
RL = 100 Ω, Channel-Channel
1
0.5
0.5
DC, Enabled
Disabled, differential
Disabled
Disabled
No Load
0.1
30
2
1
Differential
10
4
mV
V p-p
2
V p-p
ƒ = 10 MHz
–48
dB
Any switch configuration
Differential
2
5
3
pF
kΩ
µA
50% update to 1% settling
50% settling
Differential
200
20
40
ns
ns
mV p-p
Rev. PrA | Page 3 of 32
2
1
1
%
%
%
2
Ω
kΩ
pF
µA
V p-p
AD8117/AD8118
POWER SUPPLIES
Supply current
Supply voltage range
PSRR
OPERATING TEMPERATURE
RANGE
Temperature range
θJA
Preliminary Technical Data
VPOS, outputs enabled, no load
Outputs disabled
VNEG, outputs enabled, no load
Outputs disabled
DVDD, outputs enabled, no load
VNEG, VPOS, ƒ = 1 MHz
VOCM, ƒ = 1 MHz
4.5 to 5.5
–85
–75
mA
mA
mA
mA
mA
V
dB
dB
Operating (still air)
Operating (still air)
−40 to +85
15
°C
°C/W
Rev. PrA | Page 4 of 32
500
210
500
220
1
Preliminary Technical Data
AD8117/AD8118
TIMING CHARACTERISTICS (SERIAL MODE)
Parameter
Serial Data Setup Time
CLK Pulsewidth
Serial Data Hold Time
CLK Pulse Separation
CLK to UPDATE Delay
UPDATE Pulsewidth
CLK to DATA OUT Valid
Propagation Delay, UPDATE to Switch On or Off
Data Load Time, CLK = 5 MHz, Serial Mode
CLK, UPDATE Rise and Fall Times
RESET Time
Symbol
t1
t2
t3
t4
t5
t6
t7
Limit
Typ
Min
Max
Unit
ns
ns
ns
ns
ns
ns
ns
ns
µs
ns
ns
Specifications subject to change without notice.
t2
1
CLK
0
t1
t4
LOAD DATA INTO
SERIAL REGISTER
ON FALLING EDGE
t3
1
OUT7 (D4)
DATA IN
OUT7 (D3)
OUT00 (D0)
0
t5
1 = LATCHED
UPDATE
0 = TRANSPARENT
t6
TRANSFER DATA FROM SERIAL
REGISTER TO PARALLEL
LATCHES DURING LOW LEVEL
t7
DATA OUT
Figure 2. Timing Diagram, Serial Mode
Table 2. Logic Levels
VIH
RESET,
SERPAR, CLK,
DATA IN,
UPDATE
2.0 V min
VIL
RESET,
SERPAR, CLK,
DATA IN,
UPDATE
0.8 V max
VOH
DATA OUT
VOL
DATA OUT
2.7 V min
0.5 V max
IIH
RESET,
SERPAR, CLK,
DATA IN,
UPDATE
20 µA max
Rev. PrA | Page 5 of 32
IIL
RESET,
SERPAR, CLK,
DATA IN,
UPDATE
–400 µA max
IOH
DATA OUT
IOL
DATA OUT
–400 µA max
1 mA min
AD8117/AD8118
Preliminary Technical Data
TIMING CHARACTERISTICS (PARALLEL MODE)
Parameter
Parallel Data Setup Time
WE Pulsewidth
Parallel Data Hold Time
WE Pulse Separation
WE to UPDATE Delay
UPDATE Pulsewidth
Propagation Delay, UPDATE to Switch On or Off
WE, UPDATE Rise and Fall Times
RESET Time
Symbol
t1
t2
t3
t4
t5
t6
Min
Limit
Typ
Max
Unit
ns
ns
ns
ns
ns
ns
ns
ns
ns
Specifications subject to change without notice.
t2
t4
1
WE
0
t1
D0–D5
A0–A4
t3
1
0
t5
t6
1 = LATCHED
UPDATE
0 = TRANSPARENT
Figure 3. Timing Diagram, Parallel Mode
Table 3. Logic Levels
VIH
RESET,
SERPAR, WE,
D0, D1, D2, D3,
D4, D5, A0, A1,
A2, A3, A4,
UPDATE
2.0 V min
VIL
RESET,
SERPAR, WE,
D0, D1, D2, D3,
D4, D5, A0, A1,
A2, A3, A4,
UPDATE
0.8 V max
VOH
DATA OUT
VOL
DATA OUT
disabled
disabled
IIH
RESET,
SERPAR, WE,
D0, D1, D2, D3,
D4, D5, A0, A1,
A2, A3, A4,
UPDATE
20 µA max
Rev. PrA | Page 6 of 32
IIL
RESET,
SERPAR, WE,
D0, D1, D2, D3,
D4, D5, A0, A1,
A2, A3, A4,
UPDATE
–400 µA max
IOH
DATA OUT
IOL
DATA OUT
disabled
disabled
Preliminary Technical Data
AD8117/AD8118
ABSOLUTE MAXIMUM RATINGS
Table 4.
Parameter
Analog Supply Voltage (VPOS – VNEG)
Digital Supply Voltage (VDD – DGND)
Ground potential difference (VNEG –
DGND)
Maximum potential difference
(VDD – VNEG)
Common-Mode Analog Input
Voltage
Differential Analog Input Voltage
Digital Input Voltage
Output Voltage (Disabled Analog
Output)
Output Short-Circuit Duration
Storage Temperature
Operating Temperature Range
Lead Temperature Range
(Soldering 10 sec)
Junction Temperature
THERMAL RESISTANCE
Rating
+6 V
+6 V
+0.5 V to –2.5 V
θJA is specified for the worst-case conditions, that is, a device
soldered in a circuit board for surface-mount packages.
Table 5. Thermal Resistance
θJA
15
Package Type
BGA
+6 V
(VNEG – 0.5 V) to (VPOS +
0.5 V)
±2V
VDD
(VPOS – 1 V) to (VNEG + 1 V)
Momentary
−65°C to +125°C
−40°C to +85°C
300°C
150°C
NOTE
Stresses above those listed under Absolute Maximum Ratings
may cause permanent damage to the device. This is a stress
rating only and functional operation of the device at these or
any other conditions above those indicated in the operational
section of this specification is not implied. Exposure to absolute
maximum rating conditions for extended periods may affect
device reliability.
θJC
Unit
°C/W
POWER DISSIPATION
The AD8117/AD8118 are operated with ±2.5 V or +5 V
supplies and can drive loads down to 100 Ω, resulting in a large
range of possible power dissipations. For this reason, extra
care must be taken derating the operating conditions based on
ambient temperature.
Packaged in a 308-lead BGA, the AD8117/AD8118 junctionto-ambient thermal impedance (θJA) is 15°C/W. For long-term
reliability, the maximum allowed junction temperature of the
die should not exceed 150°C. Temporarily exceeding this limit
may cause a shift in parametric performance due to a change in
stresses exerted on the die by the package. Exceeding a junction
temperature of 175°C for an extended period can result in
device failure. The following curve shows the range of allowed
internal die power dissipations that meet these conditions over
the −40°C to +85°C ambient temperature range. When using
the table, do not include external load power in the Maximum
Power calculation, but do include load current dropped on the
die output transistors.
8.0
MAXIMUM POWER – Watts
TJ = 150 C
7.0
6.0
5.0
4.0
15
25
35
45
55
65
AMBIENT TEMPERATURE – C
75
85
Figure 4. Maximum Die Power Dissipation vs. Ambient Temperature
ESD CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on
the human body and test equipment and can discharge without detection. Although this product features
proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy
electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance
degradation or loss of functionality.
Rev. PrA | Page 7 of 32
AD8117/AD8118
Preliminary Technical Data
PIN CONFIGURATIONS AND FUNCTION DESCRIPTIONS
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
vpos
vpos
vpos
vpos
op17
on17
op19
on19
op21
on21
op23
on23
op25
on25
op27
on27
op29
on29
op31
on31
vpos
vpos
vpos
vpos
vpos
vpos
op16
on16
op18
on18
op20
on20
op22
on22
op24
on24
op26
on26
op28
on28
op30
on30
vpos
vpos
vpos
vpos
vpos
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vpos
in16
vpos
vpos
vneg
vocm
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vocm
vneg
vpos
ip0
vpos
ip16
in17
vneg
vocm
vocm
vneg
in0
ip1
in18
ip17
vneg
vdd
vdd
vneg
ip2
in1
ip18
in19
vneg
dgnd
dgnd
vneg
in2
ip3
in20
ip19
vneg
resetb
data_out
vneg
ip4
in3
ip20
in21
vneg
updateb
clk
vneg
in4
ip5
in22
ip21
vneg
web
data_in
vneg
ip6
in5
ip22
in23
vpos
d5
PRELIMINARY
serbpar
vpos
in6
ip7
in24
ip23
vpos
d4
Bottom View
a4
vpos
ip8
in7
ip24
in25
vpos
d3
PRELIMINARY
a3
vpos
in8
ip9
in26
ip25
vneg
d2
a2
vneg
ip10
in9
ip26
in27
vneg
d1
a1
vneg
in10
ip11
in28
ip27
vneg
d0
a0
vneg
ip12
in11
ip28
in29
vneg
vdd
vdd
vneg
in12
ip13
in30
ip29
vneg
dgnd
dgnd
vneg
ip14
in13
ip30
in31
vneg
vocm
vocm
vneg
in14
ip15
vpos
ip31
vpos
vneg
vocm
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vocm
vneg
vpos
vpos
in15
vpos
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vpos
vpos
vpos
vpos
vpos
on14
op14
on12
op12
on10
op10
on8
op8
on6
op6
on4
op4
on2
op2
on0
op0
vpos
vpos
vpos
vpos
vpos
vpos
on15
op15
on13
op13
on11
op11
on9
op9
on7
op7
on5
op5
on3
op3
on1
op1
vpos
vpos
vpos
vpos
A
A
B
B
C
C
D
D
E
E
F
F
G
G
H
H
J
J
K
K
L
M
N
L
M
N
P
P
R
R
T
T
U
U
V
V
W
W
Y
Y
AA
AA
AB
AB
AC
AC
23
22
21
20
19
18
17
16
15
14
13
12
11
10
Figure 5. BGA Bottom View Pinout
Rev. PrA | Page 8 of 32
9
8
7
6
5
4
3
2
1
Preliminary Technical Data
AD8117/AD8118
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
vpos
vpos
vpos
on31
op31
on29
op29
on27
op27
on25
op25
on23
op23
on21
op21
on19
op19
on17
op17
vpos
vpos
vpos
vpos
A
A
vpos
vpos
vpos
vpos
on30
op30
on28
op28
on26
op26
on24
op24
on22
op22
on20
op20
on18
op18
on16
op16
vpos
vpos
vpos
vpos
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vpos
vpos
ip0
vpos
vneg
vocm
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vocm
vneg
vpos
vpos
in16
ip1
in0
vneg
vocm
vocm
vneg
in17
ip16
B
B
C
C
D
D
E
E
in1
ip2
vneg
vdd
vdd
vneg
ip17
in18
ip3
in2
vneg
dgnd
dgnd
vneg
in19
ip18
in3
ip4
vneg
data_out
resetb
vneg
ip19
in20
ip5
in4
vneg
clk
updateb
vneg
in21
ip20
F
F
G
G
H
H
J
J
in5
ip6
vneg
data_in
web
vneg
ip21
in22
ip7
in6
vpos
serbpar
PRELIMINARY
d5
vpos
in23
ip22
in7
ip8
vpos
a4
Top View
d4
vpos
ip23
in24
ip9
in8
vpos
a3
PRELIMINARY
d3
vpos
in25
ip24
in9
ip10
vneg
a2
d2
vneg
ip25
in26
ip11
in10
vneg
a1
d1
vneg
in27
ip26
in11
ip12
vneg
a0
d0
vneg
ip27
in28
ip13
in12
vneg
vdd
vdd
vneg
in29
ip28
in13
ip14
vneg
dgnd
dgnd
vneg
ip29
in30
ip15
in14
vneg
vocm
vocm
vneg
in31
ip30
in15
vpos
vpos
vneg
vocm
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vocm
vneg
vpos
ip31
vpos
vpos
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vneg
vneg
vneg
vneg
vneg
vneg
vpos
vpos
vpos
vpos
vpos
vpos
vpos
op0
on0
op2
on2
op4
on4
op6
on6
op8
on8
op10
on10
op12
on12
op14
on14
vpos
vpos
vpos
vpos
vpos
vpos
vpos
vpos
op1
on1
op3
on3
op5
on5
op7
on7
op9
on9
op11
on11
op13
on13
op15
on15
vpos
vpos
vpos
K
K
L
M
N
L
M
N
P
P
R
R
T
T
U
U
V
V
W
W
Y
Y
AA
AA
AB
AB
AC
AC
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure 6. BGA Top View Pinout
Table 6. Ball Grid Description
Ball
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
A12
Mnemonic
VPOS
VPOS
VPOS
ON31
OP31
ON29
OP29
ON27
OP27
ON25
OP25
ON23
Description
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Output number 31, negative phase.
Output number 31, positive phase.
Output number 29, negative phase.
Output number 29, positive phase.
Output number 27, negative phase.
Output number 27, positive phase.
Output number 25, negative phase.
Output number 25, positive phase.
Output number 23, negative phase.
Ball
A13
A14
A15
A16
A17
A18
A19
A20
A21
A22
A23
B1
Rev. PrA | Page 9 of 32
Mnemonic
OP23
ON21
OP21
ON19
OP19
ON17
OP17
VPOS
VPOS
VPOS
VPOS
VPOS
Description
Output number 23, positive phase.
Output number 21, negative phase.
Output number 21, positive phase.
Output number 19, negative phase.
Output number 19, positive phase.
Output number 17, negative phase.
Output number 17, positive phase.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
AD8117/AD8118
Ball
B2
B3
B4
B5
B6
B7
B8
B9
B10
B11
B12
B13
B14
B15
B16
B17
B18
B19
B20
B21
B22
B23
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
D1
D2
D3
D4
D5
Mnemonic
VPOS
VPOS
VPOS
ON30
OP30
ON28
OP28
ON26
OP26
ON24
OP24
ON22
OP22
ON20
OP20
ON18
OP18
ON16
OP16
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
VNEG
VNEG
VNEG
VNEG
VNEG
VNEG
VPOS
VPOS
VPOS
VNEG
VNEG
VNEG
VNEG
VNEG
VNEG
VPOS
VPOS
VPOS
VPOS
VPOS
IP0
VPOS
VNEG
VOCM
Preliminary Technical Data
Description
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Output number 30, negative phase.
Output number 30, positive phase.
Output number 28, negative phase.
Output number 28, positive phase.
Output number 26, negative phase.
Output number 26, positive phase.
Output number 24, negative phase.
Output number 24, positive phase.
Output number 22, negative phase.
Output number 22, positive phase.
Output number 20, negative phase.
Output number 20, positive phase.
Output number 18, negative phase.
Output number 18, positive phase.
Output number 16, negative phase.
Output number 16, positive phase.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Input number 0, positive phase.
Analog positive power supply.
Analog negative power supply.
Output common-mode reference supply.
Ball
D6
D7
D8
D9
D10
D11
D12
D13
D14
D15
D16
D17
D18
D19
D20
D21
D22
D23
E1
E2
E3
E4
E20
E21
E22
E23
F1
F2
F3
F4
F20
F21
F22
F23
G1
G2
G3
G4
G20
G21
G22
G23
H1
H2
H3
H4
H20
H21
H22
H23
Rev. PrA | Page 10 of 32
Mnemonic
VNEG
VNEG
VNEG
VNEG
VNEG
VPOS
VPOS
VPOS
VNEG
VNEG
VNEG
VNEG
VNEG
VOCM
VNEG
VPOS
VPOS
IN16
IP1
IN0
VNEG
VOCM
VOCM
VNEG
IN17
IP16
IN1
IP2
VNEG
VDD
VDD
VNEG
IP17
IN18
IP3
IN2
VNEG
DGND
DGND
VNEG
IN19
IP18
IN3
IP4
VNEG
DATA_OUT
RESETB
VNEG
IP19
IN20
Description
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Output common-mode reference supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
Input number 16, negative phase.
Input number 1, positive phase.
Input number 0, negative phase.
Analog negative power supply.
Output common-mode reference supply.
Output common-mode reference supply.
Analog negative power supply.
Input number 17, negative phase.
Input number 16, positive phase.
Input number 1, negative phase.
Input number 2, positive phase.
Analog negative power supply.
Logic positive power supply.
Logic positive power supply.
Analog negative power supply.
Input number 17, positive phase.
Input number 18, negative phase.
Input number 3, positive phase.
Input number 2, negative phase.
Analog negative power supply.
Logic negative power supply.
Logic negative power supply.
Analog negative power supply.
Input number 19, negative phase.
Input number 18, positive phase.
Input number 3, negative phase.
Input number 4, positive phase.
Analog negative power supply.
Control pin: serial data out.
Control pin: second rank data reset.
Analog negative power supply.
Input number 19, positive phase.
Input number 20, negative phase.
Preliminary Technical Data
Ball
J1
J2
J3
J4
J20
J21
J22
J23
K1
K2
K3
K4
K20
K21
K22
K23
L1
L2
L3
L4
L20
L21
L22
L23
M1
M2
M3
M4
M20
M21
M22
M23
N1
N2
N3
N4
N20
N21
N22
N23
P1
P2
P3
P4
P20
P21
P22
P23
R1
R2
Mnemonic
IP5
IN4
VNEG
CLK
UPDATEB
VNEG
IN21
IP20
IN5
IP6
VNEG
DATA_IN
WEB
VNEG
IP21
IN22
IP7
IN6
VPOS
SERBPAR
D5
VPOS
IN23
IP22
IN7
IP8
VPOS
A4
D4
VPOS
IP23
IN24
IP9
IN8
VPOS
A3
D3
VPOS
IN25
IP24
IN9
IP10
VNEG
A2
D2
VNEG
IP25
IN26
IP11
IN10
AD8117/AD8118
Description
Input number 5, positive phase.
Input number 4, negative phase.
Analog negative power supply.
Control pin: serial data clock.
Control pin: second rank write strobe.
Analog negative power supply.
Input number 21, negative phase.
Input number 20, positive phase.
Input number 5, negative phase.
Input number 6, positive phase.
Analog negative power supply.
Control pin: serial data in.
Control pin: first rank write strobe.
Analog negative power supply.
Input number 21, positive phase.
Input number 22, negative phase.
Input number 7, positive phase.
Input number 6, negative phase.
Analog positive power supply.
Control pin: serial/parallel mode select.
Control pin: input address bit 5.
Analog positive power supply.
Input number 23, negative phase.
Input number 22, positive phase.
Input number 7, negative phase.
Input number 8, positive phase.
Analog positive power supply.
Control pin: output address bit 4.
Control pin: input address bit 4.
Analog positive power supply.
Input number 23, positive phase.
Input number 24, negative phase.
Input number 9, positive phase.
Input number 8, negative phase.
Analog positive power supply.
Control pin: output address bit 3.
Control pin: input address bit 3.
Analog positive power supply.
Input number 25, negative phase.
Input number 24, positive phase.
Input number 9, negative phase.
Input number 10, positive phase.
Analog negative power supply.
Control pin: output address bit 2.
Control pin: input address bit 2.
Analog negative power supply.
Input number 25, positive phase.
Input number 26, negative phase.
Input number 11, positive phase.
Input number 10, negative phase.
Ball
R3
R4
R20
R21
R22
R23
T1
T2
T3
T4
T20
T21
T22
T23
U1
U2
U3
U4
U20
U21
U22
U23
V1
V2
V3
V4
V20
V21
V22
V23
W1
W2
W3
W4
W20
W21
W22
W23
Y1
Y2
Y3
Y4
Y5
Y6
Y7
Y8
Y9
Y10
Y11
Y12
Rev. PrA | Page 11 of 32
Mnemonic
VNEG
A1
D1
VNEG
IN27
IP26
IN11
IP12
VNEG
A0
D0
VNEG
IP27
IN28
IP13
IN12
VNEG
VDD
VDD
VNEG
IN29
IP28
IN13
IP14
VNEG
DGND
DGND
VNEG
IP29
IN30
IP15
IN14
VNEG
VOCM
VOCM
VNEG
IN31
IP30
IN15
VPOS
VPOS
VNEG
VOCM
VNEG
VNEG
VNEG
VNEG
VNEG
VPOS
VPOS
Description
Analog negative power supply.
Control pin: output address bit 1.
Control pin: input address bit 1.
Analog negative power supply.
Input number 27, negative phase.
Input number 26, positive phase.
Input number 11, negative phase.
Input number 12, positive phase.
Analog negative power supply.
Control pin: output address bit 0.
Control pin: input address bit 0.
Analog negative power supply.
Input number 27, positive phase.
Input number 28, negative phase.
Input number 13, positive phase.
Input number 12, negative phase.
Analog negative power supply.
Logic positive power supply.
Logic positive power supply.
Analog negative power supply.
Input number 29, negative phase.
Input number 28, positive phase.
Input number 13, negative phase.
Input number 14, positive phase.
Analog negative power supply.
Logic negative power supply.
Logic negative power supply.
Analog negative power supply.
Input number 29, positive phase.
Input number 30, negative phase.
Input number 15, positive phase.
Input number 14, negative phase.
Analog negative power supply.
Output common-mode reference supply.
Output common-mode reference supply.
Analog negative power supply.
Input number 31, negative phase.
Input number 30, positive phase.
Input number 15, negative phase.
Analog positive power supply.
Analog positive power supply.
Analog negative power supply.
Output common-mode reference supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
AD8117/AD8118
Ball
Y13
Y14
Y15
Y16
Y17
Y18
Y19
Y20
Y21
Y22
Y23
AA1
AA2
AA3
AA4
AA5
AA6
AA7
AA8
AA9
AA10
AA11
AA12
AA13
AA14
AA15
AA16
AA17
AA18
AA19
AA20
AA21
AA22
AA23
AB1
AB2
AB3
AB4
AB5
AB6
Mnemonic
VPOS
VNEG
VNEG
VNEG
VNEG
VNEG
VOCM
VNEG
VPOS
IP31
VPOS
VPOS
VPOS
VPOS
VPOS
VNEG
VNEG
VNEG
VNEG
VNEG
VNEG
VPOS
VPOS
VPOS
VNEG
VNEG
VNEG
VNEG
VNEG
VNEG
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
OP0
ON0
OP2
Preliminary Technical Data
Description
Analog positive power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Output common-mode reference supply.
Analog negative power supply.
Analog positive power supply.
Input number 31, positive phase.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog negative power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Output number 0, positive phase.
Output number 0, negative phase.
Output number 2, positive phase.
Ball
AB7
AB8
AB9
AB10
AB11
AB12
AB13
AB14
AB15
AB16
AB17
AB18
AB19
AB20
AB21
AB22
AB23
AC1
AC2
AC3
AC4
AC5
AC6
AC7
AC8
AC9
AC10
AC11
AC12
AC13
AC14
AC15
AC16
AC17
AC18
AC19
AC20
AC21
AC22
AC23
Rev. PrA | Page 12 of 32
Mnemonic
ON2
OP4
ON4
OP6
ON6
OP8
ON8
OP10
ON10
OP12
ON12
OP14
ON14
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
VPOS
OP1
ON1
OP3
ON3
OP5
ON5
OP7
ON7
OP9
ON9
OP11
ON11
OP13
ON13
OP15
ON15
VPOS
VPOS
VPOS
Description
Output number 2, negative phase.
Output number 4, positive phase.
Output number 4, negative phase.
Output number 6, positive phase.
Output number 6, negative phase.
Output number 8, positive phase.
Output number 8, negative phase.
Output number 10, positive phase.
Output number 10, negative phase.
Output number 12, positive phase.
Output number 12, negative phase.
Output number 14, positive phase.
Output number 14, negative phase.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Output number 1, positive phase.
Output number 1, negative phase.
Output number 3, positive phase.
Output number 3, negative phase.
Output number 5, positive phase.
Output number 5, negative phase.
Output number 7, positive phase.
Output number 7, negative phase.
Output number 9, positive phase.
Output number 9, negative phase.
Output number 11, positive phase.
Output number 11, negative phase.
Output number 13, positive phase.
Output number 13, negative phase.
Output number 15, positive phase.
Output number 15, negative phase.
Analog positive power supply.
Analog positive power supply.
Analog positive power supply.
Preliminary Technical Data
AD8117/AD8118
Table 7. Operation Truth Table
UPDATE
X
CLK
X
DATA IN
X
DATA
OUT
X
WE
X
RESET
0
SER/PAR
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1
1
1
1
1
X
X
X
X
X
Figure 7. Logic Diagram
Rev. PrA | Page 13 of 32
Operation/Comment
Asynchronous reset. All outputs
are disabled.
tbd
tbd
tbd
tbd
tbd
AD8117/AD8118
Preliminary Technical Data
OPn
2500Ω
Ω
IPn
3.4pF
OPn,ONn
0.4pF
1.3pF
30k
0.3pF
1.3pF
3.4pF
INn
2500Ω
Ω
ONn
a. AD8117/AD8118 Enabled Output
(see also ESD Protection Map)
2500Ω
Ω
IPn
b. AD8117/AD8118 Disabled Output
(see also ESD Protection Map)
5075Ω
Ω
c. AD8117 Receiver (see also ESD Protection Map)
2500Ω
Ω
1.3pF
0.3pF
1.6pF
0.3pF
1.3pF
2500Ω
Ω
2500Ω
Ω
1.3pF
INn
5075Ω
Ω
d. AD8118 Receiver (see also ESD Protection Map)
2538Ω
Ω
IPn
IPn
1.3pF
INn
2538Ω
Ω
e. AD8117/AD8118 Receiver Simplified Equivalent
Circuit When Driving Differentially
3.33kΩ
Ω AD8117 G=+1
3.76kΩ
Ω AD8118 G=+2
INn
f. AD8117/AD8118 Receiver Simplified Equivalent
Circuit When Driving Single-Ended
VDD
25kΩ
Ω
VOCM
RESET
h. Reset Input (see also ESD Protection Map)
VDD
VPOS
DATA OUT
DGND
j. Logic Output (see also ESD Protection Map)
VDD
CLK,
RESET,
SER/PAR,
WE,
UPDATE,
DATA IN,
DATA OUT,
A[4:0],
D[5:0]
IPn,
INn,
OPn,
ONn,
VOCM
VNEG
DGND
k. ESD Protection Map
Figure 8. I/O Schematics
Rev. PrA | Page 14 of 32
1kΩ
Ω
DGND
DGND
VNEG
g. VOCM input (see also ESD Protection Map)
CLK, SER/PAR, WE,
UPDATE, DATA IN,
A[4:0], D[4:0]
1kΩ
Ω
i. Logic Input (see also ESD Protection Map)
Preliminary Technical Data
AD8117/AD8118
TYPICAL PERFORMANCE CHARACTERISTICS
4
0
VOUT = 1 VPP
2
Differential In/Out
Single-ended
0
-20
CROSSTALK (dB)
-2
GAIN (dB)
-4
-6
-8
-10
-40
-60
-80
-12
-14
-16
300 k
1M
10 M
100 M
FREQUENCY (Hz)
1G
-100
300 k
8G
Figure 9. AD8117 Large Signal Frequency Response
1M
10 M
100 M
FREQUENCY (Hz)
1G 2G
Figure 12. AD8117 Crosstalk, One Adjacent Channel, Differential Mode
0
0
Differential In/Out
Single-Ended In/Out
-10
-20
CROSSTALK (dB)
CMRR (dB)
-20
-30
-40
-40
-60
-50
-80
-60
-70
300 k
1M
10 M
100 M
FREQUENCY (Hz)
-100
300 k
1G 2G
Figure 10. AD8117 Common-Mode Rejection
10 M
100 M
FREQUENCY (Hz)
1G 2G
Figure 13. AD8117 Crosstalk, One Adjacent Channel, Single-Ended Mode
0
140
Differential In/Out
Differential Out
120
-20
100
CROSSTALK (dB)
NOISE SPECTRAL DENSITY (nV/rtHz)
1M
80
60
-40
-60
40
-80
20
0
1k
10 k
100 k
FREQUENCY (Hz)
-100
300 k
1M
1M
10 M
100 M
FREQUENCY (Hz)
1G 2G
Figure 14. AD8117 Crosstalk, All-Hostile, Differential Mode
Figure 11. AD8117 Noise Spectral Density, Differential Mode
Rev. PrA | Page 15 of 32
AD8117/AD8118
Preliminary Technical Data
0
5k
Single-Ended In/Out
-40
-60
-80
-100
300 k
Differential In
INPUT IMPEDANCE (Ohms)
CROSSTALK (dB)
-20
1M
10 M
100 M
FREQUENCY (Hz)
4k
3k
2k
1k
0
300 k
1G 2G
Figure 15. AD8117 Crosstalk, All-Hostile, Single-Ended Mode
1M
10 M
FREQUENCY (Hz)
1G
Figure 18. AD8117 Input Impedance, Differential Mode
0
3.5k
Differential In/Out
Single-Ended In
3.0k
INPUT IMPEDANCE (Ohms)
-20
CROSSTALK (dB)
100 M
2.5k
-40
2.0k
1.5k
-60
1.0k
-80
0.5k
-100
300 k
1M
10 M
100 M
FREQUENCY (Hz)
0
300 k
1G 2G
Figure 16. AD8117 Crosstalk, Off-Isolation, Differential Mode
1M
10 M
FREQUENCY (Hz)
1G
Figure 19. AD8117 Input Impedance, Single-Ended Mode
0
30k
Single-Ended In/Out
Differential Out
OUTPUT IMPEDANCE (Ohms)
-20
CROSSTALK (dB)
100 M
-40
-60
-80
25k
20k
15k
10k
5k
-100
300 k
1M
10 M
100 M
FREQUENCY (Hz)
0
100 k
1G 2G
Figure 17. AD8117 Crosstalk, Off-Isolation, Single-Ended Mode
1M
10 M
FREQUENCY (Hz)
100 M
1G
Figure 20. AD8117 Output Impedance, Disabled, Differential Mode
Rev. PrA | Page 16 of 32
Preliminary Technical Data
AD8117/AD8118
25k
OUTPUT IMPEDANCE (Ohms)
Single-Ended Out
20k
15k
10k
5k
0
300 k
1M
10 M
FREQUENCY (Hz)
100 M
1G
Figure 21. AD8117 Output Impedance, Disabled, Single-Ended Mode
Rev. PrA | Page 17 of 32
AD8117/AD8118
Preliminary Technical Data
THEORY OF OPERATION
The AD8117 is a fully-differential crosspoint array with 32
outputs, each of which can be connected to any one of 32
inputs. Organized by output row, 32 switchable input
transconductance stages are connected to each output buffer to
form 32-to-1 multiplexers. There are 32 of these multiplexers,
each with its inputs wired in parallel, for a total array of 1,024
transconductance stages forming a multicast-capable crosspoint
switch.
Decoding logic for each output selects one (or none) of the
transconductance stages to drive the output stage. The enabled
transconductance stage drives the output stage, and feedback
forms a closed-loop amplifier with a differential gain of one (the
difference between the output voltages is equal to the difference
between the input voltages). A second feedback loop controls
the common-mode output level, forcing the average of the
differential output voltages to match the voltage on the VOCM
reference pin. Although each output has an independent
common-mode control loop, the VOCM reference is common
for the entire chip, and as such needs to be driven with a low
impedance to avoid crosstalk.
Each differential input to the AD8117 is buffered by a receiver.
The purpose of this receiver is to provide an extended input
common-mode range, and to remove this common-mode from
the signal chain. Like the output multiplexers, the input
receiver has both a differential loop and a common-mode
control loop. A mask-programmable feedback network sets the
closed-loop differential gain. For the AD8117, this differential
gain is one, and for the AD8118, this is a differential gain of
two. The receiver has an input stage that does not respond to
the common-mode of the signal. This architecture, along with
the attenuating feedback network, allows the user to apply input
voltages that extend from rail-to-rail. Excess differential loop
gain-bandwidth product reduces the effect of the closed-loop
gain on the bandwidth of the device.
The output stage of the AD8117 is designed for low differential
gain and phase error when driving composite video signals. It
also provides slew current for fast pulse response when driving
component video signals. Unlike many multiplexer designs,
these requirements are balanced such that large signal
bandwidth is very similar to small signal bandwidth. The
design load is 150 Ω, but provisions are made to drive loads as
low as 75 Ω so long as on-chip power dissipation limits are not
exceeded.
differential outputs. This high impedance allows multiple ICs
to be bussed together without additional buffering. Care must
be taken to reduce output capacitance, which will result in more
overshoot and frequency-domain peaking. A series of internal
amplifiers drive internal nodes such that a wide-band highimpedance is presented at the disabled output, even while the
output bus is under large signal swings. When the outputs are
disabled and driven externally, the voltage applied to them
should not exceed the valid output swing range for the AD8117
in order to keep these internal amplifiers in their linear range of
operation. Applying excess differential voltages to the disabled
outputs can cause damage to the AD8117 and should be
avoided (see the Absolute Maximum Ratings section of this
datasheet for guidelines).
The connection of the AD8117 is controlled by a flexible TTL
compatible logic interface. Either parallel or serial loading into
a first rank of latches preprograms each output. A global
update signal moves the programming data into the second
rank of latches, simultaneously updating all outputs. In serial
mode, a serial-out pin allows devices to be daisy chained
together for single-pin programming of multiple ICs. A poweron reset pin is available to avoid bus conflicts by disabling all
outputs. This power-on reset clears the second rank of latches,
but does not clear the first rank of latches. In parallel mode, to
quickly clear the first rank, a broadcast parallel programming
feature is available. In serial-mode, pre-programming
individual inputs is not possible and the entire shift register
needs to be flushed.
The AD8117 can operate on a single +5 V supply, powering
both the signal path (with the VPOS/VNEG supply pins), and
the control logic interface (with the VDD/DGND supply pins).
But in order to easily interface to ground-referenced video
signals, split supply operation is possible with ± 2.5 V supplies.
In this case, a flexible logic interface allows the control logic
supplies (VDD/DGND) to be run off +2 V/0 V to +3.3 V/0 V
while the core remains on split supplies. Additional flexibility
in the analog output common-mode level facilitates unequal
split supplies. If +3 V/–2 V supplies to +2 V/–3 V supplies are
desired, the VOCM pin can still be set to 0 V for groundreferenced video signals.
The outputs of the AD8117 can be disabled to minimize onchip power dissipation. When disabled, there is only a
common-mode feedback network of 30kΩ between the
Rev. PrA | Page 18 of 32
Preliminary Technical Data
AD8117/AD8118
sequence will be 192 bits times the number of devices in the
chain.
APPLICATIONS
PROGRAMMING
The AD8117/AD8118 have two options for changing the
programming of the crosspoint matrix. In the first option a
serial word of 192 bits can be provided that will update the
entire matrix each time. The second option allows for changing
a single output’s programming via a parallel interface. The
serial option requires fewer signals, but more time (clock
cycles) for changing the programming, while the parallel
programming technique requires more signals, but can change
a single output at a time and requires fewer clock cycles to
complete programming.
Serial Programming Description
The serial programming mode uses the device pins CLK,
DATA IN, UPDATE and SER/PAR. The first step is to assert a
LOW on SER/PAR in order to enable the serial programming
mode. The parallel clock, WE should be held HIGH during the
entire serial programming operation.
The UPDATE signal should be high during the time that data is
shifted into the device’s serial port. Although the data will still
shift in when UPDATE is LOW, the transparent, asynchronous
latches will allow the shifting data to reach the matrix. This will
cause the matrix to try to update to every intermediate state as
defined by the shifting data.
The data at DATA IN is clocked in at every falling edge of CLK.
A total of 192 bits must be shifted in to complete the
programming. For each of the 32 outputs, there are five bits
(D0–D4) that determine the source of its input followed by one
bit (D5) that determines the enabled state of the output. If D5 is
LOW (output disabled), the four associated bits (D0–D4) do
not matter, because no input will be switched to that output.
The most-significant-output-address data is shifted in first,
then following in sequence until the least-significant-outputaddress data is shifted in. At this point UPDATE can be taken
low, which will cause the programming of the device according
to the data that was just shifted in. The UPDATE latches are
asynchronous and when UPDATE is low they are transparent.
If more than one AD8117 device is to be serially programmed
in a system, the DATA OUT signal from one device can be
connected to the DATA IN of the next device to form a serial
chain. All of the CLK, UPDATE, and SER/PAR pins should be
connected in parallel and operated as described above. The
serial data is input to the DATA IN pin of the first device of the
chain, and it will ripple through to the last. Therefore, the data
for the last device in the chain should come at the beginning of
the programming sequence. The length of the programming
Parallel Programming Description
When using the parallel programming mode, it is not necessary
to reprogram the entire device when making changes to the
matrix. In fact, parallel programming allows the modification
of a single output at a time. Since this takes only one
WE/UPDATE cycle, significant time savings can be realized by
using parallel programming.
One important consideration in using parallel programming is
that the RESET signal does not reset all registers in the
AD8117. When taken LOW, the RESET signal will only set
each output to the disabled state. This is helpful during powerup to ensure that two parallel outputs will not be active at the
same time.
After initial power-up, the internal registers in the device will
generally have random data, even though the RESET signal has
been asserted. If parallel programming is used to program one
output, then that output will be properly programmed, but the
rest of the device will have a random program state depending
on the internal register content at power-up. Therefore, when
using parallel programming, it is essential that all outputs be
programmed to a desired state after power-up. This will ensure
that the programming matrix is always in a known state. From
then on, parallel programming can be used to modify a single
output or more at a time.
In similar fashion, if UPDATE is taken LOW after initial
power-up, the random power-up data in the shift register will
be programmed into the matrix. Therefore, in order to prevent
the crosspoint from being programmed into an unknown state,
do not apply a low logic level to UPDATE after power is
initially applied. Programming the full shift register one time to
a desired state, by either serial or parallel programming after
initial power-up, will eliminate the possibility of programming
the matrix to an unknown state.
To change an output’s programming via parallel programming,
SER/PAR and UPDATE should be taken HIGH. The serial
programming clock, CLK, should be left HIGH during parallel
programming. The parallel clock, WE, should start in the
HIGH state. The 5-bit address of the output to be programmed
should be put on A0–A4. The first five data bits (D0–D4)
should contain the information that identifies the input that
gets programmed to the output that is addressed. The sixth data
bit (D5) will determine the enabled state of the output. If D5 is
LOW (output disabled), then the data on D0–D4 does not
matter.
Rev. PrA | Page 19 of 32
AD8117/AD8118
Preliminary Technical Data
After the desired address and data signals have been
established, they can be latched into the shift register by a high
to low transition of the WE signal. The matrix will not be
programmed, however, until the UPDATE signal is taken low.
It is thus possible to latch in new data for several or all of the
outputs first via successive negative transitions of WE while
UPDATE is held HIGH, and then have all the new data take
effect when UPDATE goes LOW. This is the technique that
should be used when programming the device for the first time
after power-up when using parallel programming.
Reset
When powering up the AD8117, it is usually desirable to have
the outputs come up in the disabled state. The RESET pin,
when taken LOW, will cause all outputs to be in the disabled
state. However, the RESET signal does not reset all registers in
the AD8117. This is important when operating in the parallel
programming mode. Please refer to that section for
information about programming internal registers after powerup. Serial programming will program the entire matrix each
time, so no special considerations apply.
fashion. This presents several options for circuit configurations
that will require different gains and treatment of terminations, if
they are used.
Differential Input
The AD8117/AD8118 has differential input receivers. These
receivers allow the user to drive the inputs with a differential
signal with an uncertain common-mode voltage, such as from a
remote source over twisted pair. The receivers will respond
only to the difference in input voltages, and will restore a
common-mode voltage suitable for the internal signal path.
Noise or crosstalk that is present in both inputs will be rejected
by the input stage, as specified by its common-mode rejection
ratio (CMRR). Differential operation offers a great noise
benefit for signals that are propagated over distance in a noisy
environment.
RF
IN+
RG
OUT-
OUT+
INRG
Since the data in the shift register is random after power-up, it
should not be used to program the matrix, or the matrix can
enter unknown states. To prevent this, do not apply a logic low
signal to UPDATE initially after power-up. The shift register
should first be loaded with the desired data, and then UPDATE
can be taken LOW to program the device.
The RESET pin has a 20 kΩ pull-up resistor to VDD that can be
used to create a simple power-up reset circuit. A capacitor from
RESET to ground will hold RESET low for some time while the
rest of the device stabilizes. The low condition will cause all the
outputs to be disabled. The capacitor will then charge through
the pull-up resistor to the high state, thus allowing full
programming capability of the device.
Broadcast
The AD8117 logic interface has a broadcast mode, in which all
first rank latches can be simultaneously parallel-programmed to
the same data in one write-cycle. This is especially useful in
clearing random first rank data after power-up. To access the
broadcast mode, the part is parallel-programmed using the
device pins WE, A0–A4, D0–A5 and UPDATE. The only
difference is that the SER/PAR pin is held LOW, as if serial
programming. By holding CLK high, no serial clocking will
occur, and instead the WE can be used to clock all first rank
latches in the chip at once.
OPERATING MODES
to switch matrix
RCVR
VOCM
RF
Figure 22. Input Receiver Equivalent Circuit
The circuit configuration used by the differential input receivers
is similar to that of several Analog Devices general-purpose
differential amplifiers, such as the AD8131. It is a voltagefeedback amplifier with internal gain setting resistors. The
arrangement of feedback makes the differential input
impedance appear to be 5 kΩ across the inputs.
RIN, dm = 2 × RG = 5 kΩ
This impedance will create a small differential termination
error if the user does not account for the 5 kΩ parallel element,
although this error will be less than 1% in most cases.
Additionally, the source impedance driving the AD8117
appears in parallel with the internal gain-setting resistors, such
that there may be a gain error for some values of source
resistance. The AD8117/AD8118 are adjusted such that their
gains will be correct when driven by a back-terminated 75 Ω
source impedance at each input phase (37.5 Ω effective
impedance to ground at each input pin, or 75 Ω differential
source impedance across pairs of input pins). If a different
source impedance is presented, the differential gain of the
AD8117/AD8118 can be calculated by
The AD8117/AD8118 has fully-differential inputs and outputs.
The inputs and outputs can also be operated in a single-ended
Rev. PrA | Page 20 of 32
Gdm =
VOUT, dm
VIN, dm
=
RF
RG + RS
Preliminary Technical Data
AD8117/AD8118
where RG is 2.5 kΩ, RS is the user single-ended source resistance
(such as 37.5 Ω for a back-terminated 75 Ω source), and RF is
2.538 kΩ for the AD8117 and 5.075 kΩ for the AD8118.
In the case of the AD8117, this is
the AD8117/AD8118 will create a high-pass filter with the input
impedance of the device. This capacitor will need to be sized
such that the corner frequency is low enough for frequencies of
interest.
Single-ended Input
Gdm =
2.538 kΩ
2.5 kΩ + RS
In the case of the AD8118, this is
Gdm =
5.075 kΩ
2.5 kΩ + RS
When operating with a differential input, care must be taken to
keep the common-mode, or average, of the input voltages
within the linear operating range of the AD8117/AD8118
receiver. This common-mode range can extend rail-to-rail,
provided the differential signal swing is small enough to avoid
forward biasing the ESD diodes (it is safest to keep the
common-mode plus differential signal excursions within the
supply voltages of the part).
The differential output of the AD8117/AD8118 receiver is
linear for a peak of 1.4V of output voltage difference (1.4 V
peak input difference for the AD8117, and 0.7 V peak input
difference for the AD8118). Taking the output differentially,
using the two output phases, this allows 2.8 VPP of linear output
signal swing. Beyond this level, the signal path will saturate and
limit the signal swing. This is not a desired operation, as the
supply current will increase and the signal path will be slow to
recover from clipping. The absolute maximum allowed
differential input signal is limited by long-term reliability of the
input stage. The limits in the Absolute Maximum Ratings
section of the datasheet should be observed in order to avoid
degrading device performance permanently.
INn
50Ω
Ω
OPn
RCVR
When operating the AD8117/AD8118 receiver single-endedly,
the observed input resistance at each input pin is higher than in
the differential input case, due to a fraction of the receiver
internal output voltage appearing as a common-mode signal on
its input terminals, bootstrapping the voltage on the input
resistance. This single-ended input resistance can be calculated
by the formula
RG + RS
RIN =
1–
RF
2 × (RG + RS + RF)
+ RF
where RG is 2.5 kΩ, RS is the user single-ended source resistance
(such as 37.5 Ω for a back-terminated 75 Ω source), and RF is
2.538 kΩ for the AD8117 and 5.075 kΩ for the AD8118.
In most cases, a single-ended input signal will be referred to
mid-supply, typically ground. In this case, the undriven
differential input could be connected to ground. For best
dynamic performance and lowest offset voltage, this unused
input should be terminated with an impedance matching the
driven input, instead of being directly shorted to ground. Due
to the differential feedback of the receiver, there is highfrequency signal current in the undriven input and it should be
treated as a signal line in the board design.
AD8117
IPn
The AD8117/AD8118 input receiver can also be driven singleended (unbalanced). From the standpoint of the receiver, there
is very little difference between signals applied positive and
negative in two phases to the input pair, versus a signal applied
to one input only with the other input held at a constant
potential. One small difference is that the common-mode
between the input pins will be changing if only one input is
moving, and there is a very small common-mode to differential
conversion gain in the receiver that will add an additional gain
error to the output (see the common-mode rejection ratio
specifications for the input stage). For low frequencies, this
gain error is negligible. The common-mode rejection ratio
degrades with increasing frequency.
ONn
50Ω
Ω
Figure 23. Example of Input Driven Differentially
AC-Coupling
It is possible to AC-couple the inputs of the AD8117/AD8118
receiver. This is simplified in that bias current does not need to
be supplied externally. A capacitor in series with the inputs to
Rev. PrA | Page 21 of 32
AD8117/AD8118
Preliminary Technical Data
The VOCM reference is a high-speed signal input, common to
all output stages on the device. It requires only small amounts
of bias current, but noise appearing on this pin will be buffered
to the outputs of all the output stages. As such, the VOCM
node should be connected to a low-noise, low-impedance
voltage to avoid being a source of noise, offset and crosstalk in
the signal path.
AD8117
IPn
INn
75Ω
Ω
OPn
RCVR
ONn
75Ω
Ω
Ω)
(or 37.5Ω
Figure 24. Example of Input Driven Single-Ended
Termination
Differential Output
Benefits of Differential Operation
The AD8117/AD8118 has a fully-differential switch core, with
differential outputs. The two output voltages move in opposite
directions, with a differential feedback loop maintaining a fixed
output stage differential gain of +1 (the different overall signal
path gains between the AD8117 and AD8118 are set in the
input stage for best signal-to-noise ratio). This differential
output stage provides a benefit of crosstalk-canceling due to
parasitic coupling from one output to another being equal and
out of phase. Additionally, if the output of the device is utilized
in a differential design, noise, crosstalk and offset voltages
generated on-chip that are coupled equally into both outputs
will be cancelled by the common-mode rejection ratio of the
next device in the signal chain. By utilizing the
AD8117/AD8118 outputs in a differential application, the best
possible noise and offset specifications can be realized.
Differential Gain
The specified signal path gain of the AD8117/AD8118 refers to
its differential gain. For the AD8117, the gain of +1 means that
the difference in voltage between the two output terminals is
equal to the difference applied between the two input terminals.
For the AD8118, the ratio of output difference voltage to
applied input difference voltage is +2.
The common-mode, or average voltage of the pair of output
signals is set by the voltage on the VOCM pin. This voltage is
typically set to mid-supply (often ground), but may be moved
approximately ± 0.5 V in order to accommodate cases where
the desired output common-mode voltage may not be midsupply (as in the case of unequal split supplies). Adjusting
VOCM beyond ± 0.5 V can limit differential swing internally
below the specifications on the datasheet.
The AD8117/AD8118 is designed to drive 150 Ω on each
output (or an effective 300 Ω differential) while meeting
datasheet specifications, but the output stage is capable of
supplying the current to drive 100 Ω loads (200 Ω differential)
over the specified operating temperature range. If care is taken
to observe the maximum power derating curves, the output
stage can drive 75 Ω loads with slightly reduced slew rate and
bandwidth (an effective 150 Ω differential load).
Termination at the load end is recommended for best signal
integrity. This load termination is often a resistor to a ground
reference on each individual output. By terminating to the
same voltage level that drives the VOCM reference, the power
dissipation due to DC termination current will be reduced. In
differential signal paths, it is often desirable to terminate
differentially, with a single resistor across the differential
outputs at the load end. This is acceptable for the
AD8117/AD8118, but when the device outputs are placed in a
disabled state, a small amount of DC bias current is required if
the output is to present as a high-impedance over an excursion
of output bus voltages. If the AD8117/AD8118 disabled outputs
are floated (or simply tied together by a resistor), internal nodes
will saturate and an increase in disabled output current may be
observed.
For best pulse response, it is often desirable to place a series
resistor in each output to match the characteristic impedance
and termination of the output trace or cable. This is known as
back-termination, and helps shorten settling time by
terminating reflected signals when driving a load that is not
accurately terminated at the load end. A side-effect of backtermination is an attenuation of the output signal by a factor of
two. In this case, a gain of two is usually necessary somewhere
in the signal path to restore the signal.
AD8117
Regardless of the differential gain of the device, the commonmode gain for the AD8117 and AD8118 is +1 to the output.
This means that the common-mode of the output voltages will
directly follow the reference voltage applied to the VOCM
input.
Rev. PrA | Page 22 of 32
OPn
50Ω
Ω
100Ω
Ω
ONn
50Ω
Ω
Preliminary Technical Data
AD8117/AD8118
observed. An AD8117 taken with single-ended output will
appear to have a gain of +0.5. An AD8118 will be a singleended gain of +1.
Figure 25. Example of Back-Terminated Differential Load
Single-ended Output
Usage
The AD8117/AD8118 output pairs can be used single-endedly,
taking only one output and not using the second. This is often
desired to reduce the routing complexity in the design, or
because a single-ended load is being driven directly. This mode
of operation will produce good results, but has some
shortcomings when compared to taking the output
differentially. When observing the single-ended output, noise
that is common to both outputs appears in the output signal.
This includes thermal noise in the chip biasing, as well as
crosstalk that coupled into the signal path. This component
noise and crosstalk is equal in both outputs, and as such can be
ignored by a differential receiver with high common-mode
rejection ratio. But when taking the output single-ended, this
noise is present with respect to the ground (or VOCM)
reference and is not rejected.
When observing the output single-ended, the distribution of
offset voltages will appear greater. In the differential case, the
difference between the outputs when the difference between the
inputs is zero will be a small differential offset. This offset of
created from mismatches in components of the signal path
which must be corrected by the finite differential loop gain of
the device. In the single-ended case, this differential offset is
still observed, but an additional offset component is also
relevant. This additional component is the common-mode
offset, which is a difference between the average of the outputs
and the VOCM reference. This offset is created by mismatches
that affect the signal path in a common-mode manner, and is
corrected by the finite common-mode loop gain of the device.
A differential receiver would reject this common-mode offset
voltage, but in the single ended case this offset is observed with
respect to the signal ground. The single-ended output sums
half the differential offset voltage and all of the common-mode
offset voltage for a net gain in observed random offset.
Single-Ended Gain
The AD8117/AD8118 operates as a closed-loop differential
amplifier. The primary control loop forces the difference
between the output terminals to be a ratio of the difference
between the input terminals. One output will increase in
voltage, while the other decreases an equal amount to make the
total difference correct. The average of these output voltages is
forced to the voltage on the VOCM terminal by a second
control loop. If only one output terminal is observed with
respect to the VOCM terminal, only half of the difference
voltage will be observed. This implies that when using only one
output of the device, half of the differential gain will be
This factor of one-half in the gain increases the noise of the
device when referred to the input, contributing to higher noise
specifications for single-ended output designs.
Termination
When operating the AD8117/AD8118 with a single-ended
output, the preferred output termination scheme is a resistor at
the load end to the VOCM voltage. A back-termination may be
used, at an additional cost of one half the signal gain.
In single-ended output operation, the second phase of the
output is not used, and may or may not be terminated locally.
Termination of the unused output is not necessary for proper
device operation, so total design power dissipation can be
reduced by floating this output. However, there are several
reasons for terminating the unused output with a load
resistance equal to the signal output.
One component of crosstalk is magnetic, coupling by mutual
inductance between output package traces and bond wires that
carry load current. In a differential design, there is coupling
from one pair of outputs to other adjacent pairs of outputs. The
differential nature of the output signal simultaneously drives the
coupling field in one direction for one phase of the output, and
in an opposite direction for the other phase of the output.
These magnetic fields do not couple exactly equal into adjacent
output pairs due to different proximities, but they do
destructively cancel the crosstalk to some extent. If the load
current in each output is equal, this cancellation will be greater
and less adjacent crosstalk will be observed (regardless if the
second output is actually being used).
A second benefit of balancing the output loads in a differential
pair is to reduce fluctuations in current requirements from the
power supply. In single-ended loads, the load currents alternate
from the positive supply to the negative supply. This creates a
parasitic signal voltage in the supply pins due to the finite
resistance and inductance of the supplies. This supply
fluctuation appears as crosstalk in all outputs, attenuated by the
power supply rejection ratio (PSRR) of the device. At low
frequencies, this is a negligible component of crosstalk, but
PSRR falls off as frequency increases. With differential,
balanced loads, as one output draws current from the positive
supply, the other output draws current from the negative supply.
When the phase alternates, the first output draws current from
the negative supply and the second from the positive supply.
The effect is that a more constant current is drawn from each
supply, such that the crosstalk-inducing supply fluctuation is
minimized.
Rev. PrA | Page 23 of 32
AD8117/AD8118
Preliminary Technical Data
A third benefit of driving balanced loads can be seen if one
considers that the output pulse response will change as load
changes. The differential signal control loop in the
AD8117/AD8118 forces the difference of the outputs to be a
fixed ratio to the difference of the inputs. If the two output
responses are different due to loading, this creates a difference
that the control loop will see as signal response error, and it will
attempt to correct this error. This will distort the output signal
from the ideal response if the two outputs were balanced.
operation as there is a common-mode to differential gain
conversion that becomes greater at higher frequencies.
During operation of the AD8117/AD8118, transient currents
will flow into the VOCM net from the amplifier control loops.
Although the magnitude of these currents are small (10 – 20 µA
per output), they can contribute to crosstalk if they flow
through significant impedances. Driving VOCM with a lowimpedance, low-noise source is desirable.
AD8117
Power Dissipation
OPn
ONn
75Ω
Ω
Calculation of Power Dissipation
75Ω
Ω
8.0
150Ω
Ω
MAXIMUM POWER – Watts
TJ = 150 C
Figure 26. Example of Back-Terminated Single-Ended Load
Decoupling
The signal path of the AD8117/AD8118 is based on high open
loop gain amplifiers with negative feedback. Dominant-pole
compensation is used on-chip to stabilize these amplifiers over
the range of expected applied swing and load conditions. To
guarantee this designed stability, proper supply decoupling is
necessary with respect to both the differential control loops and
the common-mode control loops of the signal path. Signalgenerated currents must return to their sources through lowimpedance paths at all frequencies in which there is still loop
gain (up to 700 MHz at a minimum). Refer to the example
Evaluation Board schematic as an example of wideband parallel
capacitor arrangements that can properly decouple the
AD8117/AD8118.
The signal path compensation capacitors in the
AD8117/AD8118 are connected to the VNEG supply. At high
frequencies, this limits the power supply rejection ratio (PSRR)
from the VNEG supply to a lower value than that from the
VPOS supply. If given a choice, an application board should be
designed such that the VNEG power is supplied from a lowinductance plane, subject to a least amount of noise.
The VOCM should be considered a reference pin and not a
power supply. It is an input to the high-speed, high-gain
common-mode control loop of all receivers and output drivers.
In the single-ended output sense, there is no rejection from
noise on the VOCM net to the output. For this reason, care
must be taken to produce a low-noise VOCM source over the
entire range of frequencies of interest. This is not only
important to single-ended operation, but to differential
7.0
6.0
5.0
4.0
15
25
35
45
55
65
AMBIENT TEMPERATURE – C
75
85
Figure 27. Maximum Die Power Dissipation vs. Ambient Temperature
The above curve was calculated from
PD, MAX =
(TJUNCTION, MAX – TAMBIENT)
θJA
As an example, if the AD8117/AD8118 is enclosed in an
environment at 45°C (TA), the total on-chip dissipation under
all load and supply conditions must not be allowed to exceed
7.0 W.
When calculating on-chip power dissipation, it is necessary to
include the rms current being delivered to the load, multiplied
by the rms voltage drop on the AD8117/AD8118 output
devices. For a sinusoidal output, the on-chip power dissipation
due the load can be approximated by
PD, OUT = (VPOS – VOUTPUT, RMS) × IOUTPUT, RMS
For nonsinusoidal output, the power dissipation should be
calculated by integrating the on-chip voltage drop multiplied by
the load current over one period.
The user may subtract the quiescent current for the Class AB
output stage when calculating the loaded power dissipation. For
Rev. PrA | Page 24 of 32
Preliminary Technical Data
AD8117/AD8118
each output stage driving a load, subtract a quiescent power
according to
PDQ, OUTPUT = (VPOS – VNEG) × IO, QUIESCENT
PDQ, OUTPUT = (2.5 V – (–2.5V)) × (1.65 mA) = 8.25 mW
PD, OUT, Q = (VPOS – VNEG) × IOUTPUT, QUIESCENT
There are 32 output pairs, or 64 output currents.
For the AD8117/AD8118, IOUTPUT, QUIESCENT = 1.65 mA for each
single-ended output pin.
For each disabled output, the quiescent power supply current in
VPOS and VNEG drops by approximately 9 mA.
nPD, OUTPUT = 64 × 8.25 mW = 0.53 W
Step 4. Verify that the power dissipation does not exceed
maximum allowed value.
PD, ON-CHIP = PD, QUIESCENT + nPD, OUTPUT + nPDQ, OUTPUT
VPOS
PD, ON-CHIP = 2.5 W + 0.96 W – 0.53 W = 2.9 W
IO,QUIESCENT
QNPN
From the figure or the equation, this power dissipation is
below the maximum allowed dissipation for all ambient
temperatures up to and including 85°C.
VOUTPUT
QPNP
IOUTPUT
IO,QUIESCENT
Short Circuit Output Conditions
VNEG
Figure 28. Simplified Output Stage
An example: AD8117, in an ambient temperature of 85°C,
with all 32 outputs driving 1 V rms into 100 Ω loads. Power
supplies are ± 2.5 V.
Step 1. Calculate power dissipation of AD8117 using data
sheet quiescent currents. We are neglecting VDD
current as it is insignificant.
PD, QUIESCENT = (VPOS × IVPOS) + (VNEG × IVNEG)
PD, QUIESCENT = (2.5 V × 500 mA) + (2.5 V × 500 mA) = 2.5 W
Step 2. Calculate power dissipation from loads. For a
differential output and ground-referenced load, the
output power is symmetrical in each output phase.
PD, OUTPUT = (VPOS – VOUTPUT, RMS) × IOUTPUT, RMS
PD, OUTPUT = (2.5 V – 1 V) × (1 V/100
V/150 Ω) = 15
10 mW
Although there is short-circuit current protection on the
AD8117 outputs, the output current can reach values of 80 mA
into a grounded output. Any sustained operation with too
many shorted outputs can exceed the maximum die
temperature and can result in device failure (see Absolute
Maximum Ratings).
Crosstalk
Many systems, such as broadcast video and KVM switches, that
handle numerous analog signal channels, have strict
requirements for keeping the various signals from influencing
any of the others in the system. Crosstalk is the term used to
describe the coupling of the signals of other nearby channels to
a given channel.
When there are many signals in close proximity in a system, as
will undoubtedly be the case in a system that uses the
AD8117/AD8118, the crosstalk issues can be quite complex. A
good understanding of the nature of crosstalk and some
definition of terms is required in order to specify a system that
uses one or more crosspoint devices.
Types of Crosstalk
There are 32 output pairs, or 64 output currents.
Crosstalk can be propagated by means of any of three methods.
These fall into the categories of electric field, magnetic field,
and sharing of common impedances. This section will explain
these effects.
nPD, OUTPUT = 64 × 15 mW = 0.96 W
Step 3. Subtract quiescent output stage current for number
of loads (64 in this example). The output stage is
either standing, or driving a load but the current
only needs to be counted once (valid for output
voltages > 0.5 V).
Every conductor can be both a radiator of electric fields and a
receiver of electric fields. The electric field crosstalk mechanism
occurs when the electric field created by the transmitter
propagates across a stray capacitance (e.g., free space) and
couples with the receiver and induces a voltage. This voltage is
an unwanted crosstalk signal in any channel that receives it.
Rev. PrA | Page 25 of 32
AD8117/AD8118
Preliminary Technical Data
Currents flowing in conductors create magnetic fields that
circulate around the currents. These magnetic fields then
generate voltages in any other conductors whose paths they
link. The undesired induced voltages in these other channels are
crosstalk signals. The channels that crosstalk can be said to have
a mutual inductance that couples signals from one channel to
another.
The power supplies, grounds, and other signal return paths of a
multichannel system are generally shared by the various
channels. When a current from one channel flows in one of
these paths, a voltage that is developed across the impedance
becomes an input crosstalk signal for other channels that share
the common impedance.
All these sources of crosstalk are vector quantities, so the
magnitudes cannot simply be added together to obtain the total
crosstalk. In fact, there are conditions where driving additional
circuits in parallel in a given configuration can actually reduce
the crosstalk. The fact that the AD8117/AD8118 is a fullydifferential design means that many sources of crosstalk either
destructively cancel, or are common-mode to the signal and can
be rejected by a differential receiver.
Areas of Crosstalk
A practical AD8117/AD8118 circuit must be mounted to some
sort of circuit board in order to connect it to power supplies
and measurement equipment. Great care has been taken to
create a characterization board (also available as an evaluation
board) that adds minimum crosstalk to the intrinsic device.
This, however, raises the issue that a system’s crosstalk is a
combination of the intrinsic crosstalk of the devices in addition
to the circuit board to which they are mounted. It is important
to try to separate these two areas when attempting to minimize
the effect of crosstalk.
In addition, crosstalk can occur among the inputs to a
crosspoint and among the outputs. It can also occur from input
to output. Techniques will be discussed for diagnosing which
part of a system is contributing to crosstalk.
Measuring Crosstalk
Crosstalk is measured by applying a signal to one or more
channels and measuring the relative strength of that signal on a
desired selected channel. The measurement is usually expressed
as dB down from the magnitude of the test signal. The crosstalk
is expressed by
|XT| = 20 log10 (ASEL(s) / ATEST(s))
where s = jω is the Laplace transform variable, ASEL(s) is the
amplitude of the crosstalk induced signal in the selected
channel, and ATEST(s) is the amplitude of the test signal. It can be
seen that crosstalk is a function of frequency, but not a function
of the magnitude of the test signal (to first order). In addition,
the crosstalk signal will have a phase relative to the test signal
associated with it.
A network analyzer is most commonly used to measure
crosstalk over a frequency range of interest. It can provide both
magnitude and phase information about the crosstalk signal.
As a crosspoint system or device grows larger, the number of
theoretical crosstalk combinations and permutations can
become extremely large. For example, in the case of the 32 × 32
matrix of the AD8117, we can look at the number of crosstalk
terms that can be considered for a single channel, say the IN00
input. IN00 is programmed to connect to one of the AD8117
outputs where the measurement can be made.
First, the crosstalk terms associated with driving a test signal
into each of the other 31 inputs can be measured one at a time,
while applying no signal to IN00. Then the crosstalk terms
associated with driving a parallel test signal into all 31 other
inputs can be measured two at a time in all possible
combinations, then three at a time, and so on, until, finally,
there is only one way to drive a test signal into all 31 other
inputs in parallel.
Each of these cases is legitimately different from the others and
might yield a unique value, depending on the resolution of the
measurement system, but it is hardly practical to measure all
these terms and then specify them. In addition, this describes
the crosstalk matrix for just one input channel. A similar
crosstalk matrix can be proposed for every other input. In
addition, if the possible combinations and permutations for
connecting inputs to the other outputs (not used for
measurement) are taken into consideration, the numbers rather
quickly grow to astronomical proportions. If a larger crosspoint
array of multiple AD8117s is constructed, the numbers grow
larger still.
Obviously, some subset of all these cases must be selected to be
used as a guide for a practical measure of crosstalk. One
common method is to measure all hostile crosstalk; this means
that the crosstalk to the selected channel is measured while all
other system channels are driven in parallel. In general, this will
yield the worst crosstalk number, but this is not always the case,
due to the vector nature of the crosstalk signal.
Other useful crosstalk measurements are those created by one
nearest neighbor or by the two nearest neighbors on either side.
These crosstalk measurements will generally be higher than
those of more distant channels, so they can serve as a worstcase measure for any other one-channel or two-channel
crosstalk measurements.
Rev. PrA | Page 26 of 32
Preliminary Technical Data
AD8117/AD8118
Input and Output Crosstalk
Capacitive coupling is voltage-driven (dV/dt), but is generally a
constant ratio. Capacitive crosstalk is proportional to input or
output voltage, but this ratio is not reduced by simply reducing
signal swings. Attenuation factors must be changed by
changing impedances (lowering mutual capacitance), or
destructive canceling must be utilized by summing equal and
out of phase components. For high-input impedance devices
such as the AD8117/AD8118, capacitances generally dominate
input-generated crosstalk.
Inductive coupling is proportional to current (dI/dt), and will
often scale as a constant ratio with signal voltage, but will also
show a dependence on impedances (load current). Inductive
coupling can also be reduced by constructive canceling of equal
and out of phase fields. In the case of driving low-impedance
video loads, output inductances contribute highly to output
crosstalk.
The flexible programming capability of the AD8117/AD8118
can be used to diagnose whether crosstalk is occurring more on
the input side or the output side. Some examples are
illustrative. A given input pair (IN07 in the middle for this
example) can be programmed to drive OUT07 (also in the
middle). The inputs to IN07 are just terminated to ground (via
50 Ω or 75 Ω) and no signal is applied.
All the other inputs are driven in parallel with the same test
signal (practically provided by a distribution amplifier), with all
other outputs except OUT07 disabled. Since grounded IN07 is
programmed to drive OUT07, no signal should be present. Any
signal that is present can be attributed to the other 15 hostile
input signals, because no other outputs are driven (they are all
disabled). Thus, this method measures the all-hostile input
contribution to crosstalk into IN07. Of course, the method can
be used for other input channels and combinations of hostile
inputs.
For output crosstalk measurement, a single input channel is
driven (IN00, for example) and all outputs other than a given
output (IN07 in the middle) are programmed to connect to
IN00. OUT07 is programmed to connect to IN15 (far away
from IN00), which is terminated to ground. Thus OUT07
should not have a signal present since it is listening to a quiet
input. Any signal measured at the OUT07 can be attributed to
the output crosstalk of the other 16 hostile outputs. Again, this
method can be modified to measure other channels and other
crosspoint matrix combinations.
Effect of Impedances on Crosstalk
The input side crosstalk can be influenced by the output
impedance of the sources that drive the inputs. The lower the
impedance of the drive source, the lower the magnitude of the
crosstalk. The dominant crosstalk mechanism on the input side
is capacitive coupling. The high impedance inputs do not have
significant current flow to create magnetically induced
crosstalk. However, significant current can flow through the
input termination resistors and the loops that drive them. Thus,
the PC board on the input side can contribute to magnetically
coupled crosstalk.
From a circuit standpoint, the input crosstalk mechanism looks
like a capacitor coupling to a resistive load. For low frequencies
the magnitude of the crosstalk will be given by
|XT| = 20 log10[(RSCM) × s]
where RS is the source resistance, CM is the mutual capacitance
between the test signal circuit and the selected circuit, and s is
the Laplace transform variable.
From the equation it can be observed that this crosstalk
mechanism has a high-pass nature; it can also be minimized by
reducing the coupling capacitance of the input circuits and
lowering the output impedance of the drivers. If the input is
driven from a 75 Ω terminated cable, the input crosstalk can be
reduced by buffering this signal with a low output impedance
buffer.
On the output side, the crosstalk can be reduced by driving a
lighter load. Although the AD8117 is specified with excellent
differential gain and phase when driving a standard 150 Ω
video load, the crosstalk will be higher than the minimum
obtainable due to the high output currents. These currents will
induce crosstalk via the mutual inductance of the output pins
and bond wires of the AD8117.
From a circuit standpoint, this output crosstalk mechanism
looks like a transformer with a mutual inductance between the
windings that drives a load resistor. For low frequencies, the
magnitude of the crosstalk is given by
|XT| = 20 log10 (MXY × s/RL)
where MXY is the mutual inductance of output X to output Y
and RL is the load resistance on the measured output. This
crosstalk mechanism can be minimized by keeping the mutual
inductance low and increasing RL. The mutual inductance can
be kept low by increasing the spacing of the conductors and
minimizing their parallel length.
PCB Layout
Extreme care must be exercised to minimize additional
crosstalk generated by the system circuit board(s). The areas
Rev. PrA | Page 27 of 32
AD8117/AD8118
Preliminary Technical Data
that must be carefully detailed are grounding, shielding, signal
routing, and supply bypassing.
The packaging of the AD8117/AD8118 is designed to help keep
the crosstalk to a minimum. On the BGA substrate, each pair is
carefully routed to predominately couple to each other, with
shielding traces separating adjacent signal pairs. The ball grid
array is arranged such that similar board routing can be
achieved. Only the outer two rows are used for signals, such
that vias can be used to take the input rows to a lower signal
plane if desired.
The input and output signals will have minimum crosstalk if
they are located between ground planes on layers above and
below, and separated by ground in between. Vias should be
located as close to the IC as possible to carry the inputs and
outputs to the inner layer. The input and output signals surface
at the input termination resistors and the output series backtermination resistors. To the extent possible, these signals
should also be separated as soon as they emerge from the IC
package.
PCB Termination Layout
As frequencies of operation increase, the importance of proper
transmission line signal routing becomes more important. The
bandwidth of the AD8117/AD8118 is large enough that using
high impedance routing will not provide a flat in-band
frequency response for practical signal trace lengths. It is
necessary for the user to choose a characteristic impedance
suitable for the application and properly terminate the input
and output signals of the AD8117/AD8118. Traditionally,
video applications have used 75 Ω single-ended enviroments.
RF applications are generally 50 Ω single-ended (and board
manufacturers have the most experience with this application).
CAT-5 cabling is usually driven as differential pairs of 100 Ω
differential impedance.
For flexibility, the AD8117/AD8118 does not contain on-chip
termination resistors. This flexibility in application comes with
some board layout challenges. The distance between the
termination of the input transmission line and the
AD8117/AD8118 die is a high-impedance stub, and will cause
reflections of the input signal. With some simplification, it can
be shown that these reflections will cause peaking of the input
at regular intervals in frequency, dependent on the propagation
speed (VP) of the signal in the choosen board material and the
distance (d) between the termination resistor and the
AD8117/AD8118. If the distance is great enough, these peaks
can occur in-band. In fact, practical experience shows that
these peaks are not high-Q, and should be pushed out to three
or four times the desired bandwidth in order to not have an
effect on the signal. For a board designer using FR4 (VP = 144
× 106 m/s), this means the AD8117/AD8118 should be no more
than 1.5 cm after the termination resistors, and preferably
should be placed even closer. The BGA substrate routing inside
the AD8117/AD8118 is approximately 1 cm in length and adds
to the stub length, so 1.5 cm PCB routing equates to
d = 2.5 × 10 –2 m in the calculations.
fpeak =
(2n + 1)VP
4d
, n = {0, 1, 2, 3, …}
In some cases, it is difficult to place the termination close to the
AD8117/AD8118 due to space constraints, differential routing,
and large resistor footprints. A preferable solution in this case
is to maintain a controlled transmission line past the
AD8117/AD8118 inputs and terminate the end of the line.
This is known as fly-by termination. The input impedance of
the AD8117/AD8118 is large enough and stub length inside the
package is small enough that this works well in practice.
Implementation of fly-by input termination often includes
bringing the signal in on one routing layer, then passing
through a filled-via under the AD8117/AD8118 input ball, then
back out to termination on another signal layer. In this case,
care must be taken to tie the reference ground planes together
near the signal via if the signal layers are referenced to different
ground planes.
AD8117
IPn
OPn
INn
ONn
75Ω
Ω
Figure 29. Fly-by Input Termination. Grounds for the two transmission lines
shown must be tied together close to the INn pin.
If multiple AD8117/AD8118 are to be driven in parallel, a flyby input termination scheme is very useful, but the distance
from each AD8117/AD8118 input to the driven input
transmission line is a stub that should be minimized in length
and parasitics using the discussed guidelines.
When driving the AD8117/AD8118 single-endedly, the
undriven input is often terminated with a resistance in order to
balance the input stage. It can be seen that by terminating the
undriven input with a resistor of one-half the characteristic
impedance, the input stage will be perfectly balanced (37.5 Ω,
for example, to balance the two parallel 75 Ω terminations on
the driven input). However, due to the feedback in the input
receiver, there is high-speed signal current leaving the undriven
input. In order to terminate this high-speed signal, proper
transmission-line techniques should be used. One solution is
Rev. PrA | Page 28 of 32
Preliminary Technical Data
AD8117/AD8118
to adjust the trace width to create a transmission line of half the
characteristic impedance and terminate the far end with this
resistance (37.5 Ω in a 75 Ω system). This is not often practical
as trace widths become large. In most cases, the best practical
solution is to place the half-characteristic impedance resistor as
close as possible (preferably less than 1.5 cm away) and to
reduce the parasitics of the stub (by removing the ground plane
under the stub, for example). In either case, the designer must
decide if the layout complexity created by a balanced,
terminated solution is preferable to simply grounding the
undriven input at the ball with no trace.
While the examples discussed so far are for input termination,
the theory is similar for output back-termination. Taking the
AD8117/AD8118 as an ideal voltage source, any distance of
routing between the AD8117/AD8118 and a back-termination
resistor will be a stub that will create reflections. For this
reason, back-termination resistors should also be placed close
to the AD8117/AD8118. In practice, because back-termination
resistors are series elements, their footprint in the routing is
narrower and it is easier to place them close in board layout.
Rev. PrA | Page 29 of 32
AD8117/AD8118
Preliminary Technical Data
Figure 30. Evaluation Board Simplified Schematic
Rev. PrA | Page 30 of 32
Preliminary Technical Data
AD8117/AD8118
OUTLINE DIMENSIONS
A1 CORNER
INDEX AREA
31.00
BSC SQ
23
6
4
2
22 20 18 16 14 12 10 8
21 19 17 15 13 11
9
5
3
7
1
A
B
C
D
E
F
G
H
J
K
L
M
N
P
R
T
U
V
W
Y
AA
AB
AC
BALL A1
INDICATOR
27.94
BSC SQ
TOP VIEW
BOTTOM
VIEW
1.27
BSC
DETAIL A
1.07
0.99
0.92
DETAIL A
*1.765 MAX
0.70
0.63
0.56
0.10 MIN
0.25 MIN
(4×)
0.90
0.75
0.60
BALL DIAMETER
COPLANARITY
0.20
SEATING
PLANE
*COMPLIANT TO JEDEC STANDARDS MO-192-BAN-2
WITH THE EXCEPTION TO PACKAGE HEIGHT.
Figure 31. 304-Lead Ball Grid Array, Thermally Enhanced [BGA_ED]
Dimensions shown in millimeters
ORDERING GUIDE
Model
Temperature Range
AD8117ABPZ
−40°C to +85°C
AD8118ABPZ
−40°C to +85°C
AD8117-EVAL
NOTE: Z suffix denotes lead-free package.
Package Description
304-Lead Ball Grid Array Package [BGA_ED] (31 × 31 mm)
304-Lead Ball Grid Array Package [BGA_ED] (31 × 31 mm)
AD8117 Evaluation Kit
Rev. PrA | Page 31 of 32
Package Option
SBGA-304
SBGA-304
AD8117/AD8118
Preliminary Technical Data
NOTES
© 2006 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
PR06365-0-8/06(PrA)
Rev. PrA | Page 32 of 32