LMH6580VSX [TI]

IC 8-CHANNEL, CROSS POINT SWITCH, PQFP48, TQFP-48, Multiplexer or Switch;
LMH6580VSX
型号: LMH6580VSX
厂家: TEXAS INSTRUMENTS    TEXAS INSTRUMENTS
描述:

IC 8-CHANNEL, CROSS POINT SWITCH, PQFP48, TQFP-48, Multiplexer or Switch

文件: 总25页 (文件大小:709K)
中文:  中文翻译
下载:  下载PDF数据表文档文件
National Semiconductor is now part of  
Texas Instruments.  
Search http://www.ti.com/ for the latest technical  
information and details on our current products and services.  
September 2007  
LMH6580/LMH6581  
8x4 500 MHz Analog Crosspoint Switch, Gain of 1,  
Gain of 2  
General Description  
Features  
The LMH® family of products is joined by the LMH6580 and  
the LMH6581, high speed, non-blocking, analog, crosspoint  
switches. The LMH6580/LMH6581 are designed for high  
speed, DC coupled, analog signals such as high resolution  
video (UXGA and higher). The LMH6580/LMH6581 each has  
eight inputs and four outputs. The non-blocking architecture  
allows any output to be connected to any input, including an  
input that is already selected. With fully buffered inputs the  
LMH6580/LMH6581 can be impedance matched to nearly  
any source impedance. The buffered outputs of the  
LMH6580/LMH6581 can drive up to two back terminated  
video loads (75load). The outputs and inputs also feature  
high impedance inactive states allowing high performance in-  
put and output expansion for array sizes such as 8 x 8 or 16  
x 4 by combining two devices. The LMH6580/LMH6581 are  
controlled with a 4 pin serial interface that can be configured  
as a 3 wire interface. Both serial mode and addressed modes  
are available.  
8 inputs and 4 outputs  
48-pin TQFP package  
−3 dB bandwidth (VOUT = 2 VPP, RL = 1 kΩ)  
−3 dB bandwidth (VOUT = 2 VPP, RL = 150Ω)  
Fast slew rate  
Channel to channel crosstalk (10/100 MHz) −70/ −52 dBc  
All hostile crosstalk (10/100 MHz)  
Easy to use serial programming  
Two programming modes  
Symmetrical pinout facilitates expansion.  
Output current  
500 MHz  
450 MHz  
2100 V/μs  
−55/−45 dBc  
4 wire bus  
Serial & addressed modes  
±70 mA  
AV = 1 or AV = 2  
Two gain options  
Applications  
Studio monitoring/production video systems  
Conference room multimedia video systems  
KVM (keyboard video mouse) systems  
Security/surveillance systems  
Multi-antenna diversity radio  
The LMH6580/LMH6581 come in 48-pin TQFP packages.  
They also have diagonally symmetrical pin assignments to  
facilitate double sided board layouts and easy pin connec-  
tions for expansion.  
Video test equipment  
Medical imaging  
Wide-band routers & switches  
Connection Diagram  
Block Diagram  
30007211  
30007202  
LMH® is a registered trademark of National Semiconductor Corporation.  
TRI-STATE® is a registered trademark of National Semiconductor Corporation.  
© 2007 National Semiconductor Corporation  
300072  
www.national.com  
Storage Temperature Range  
Soldering Information  
Infrared or Convection (20 sec.)  
Wave Soldering (10 sec.)  
−65°C to +150°C  
Absolute Maximum Ratings (Note 1)  
If Military/Aerospace specified devices are required,  
please contact the National Semiconductor Sales Office/  
Distributors for availability and specifications.  
235°C  
260°C  
ESD Tolerance (Note 2)  
Human Body Model  
Machine Model  
Operating Ratings (Note 1)  
Temperature Range (Note 4)  
2000V  
200V  
−40°C to +85°C  
±3V to ±5.5V  
Supply Voltage Range  
VS  
±6V  
IIN (Input Pins)  
±20 mA  
(Note 3)  
Vto V+  
+150°C  
Thermal Resistance  
48-Pin TQFP  
θJA  
θJC  
IOUT  
44°C/W  
12°C/W  
Input Voltage Range  
Maximum Junction Temperature  
±3.3V Electrical Characteristics (Note 5)  
Unless otherwise specified, typical conditions are: TA = 25°C, AV = +2, VS = ±3.3V, RL = 100Ω; Boldface limits apply at the  
temperature extremes.  
Symbol  
Parameter  
Conditions  
Min  
Typ  
Max  
Units  
(Note 8)  
(Note 7)  
(Note 8)  
Frequency Domain Performance  
SSBW  
LSBW  
−3 dB Bandwidth  
VOUT = 0.5 VPP  
LMH6580 VOUT = 1 VPP  
LMH6581 V OUT = 2 VPP, RL = 1 kΩ  
LMH6580 VOUT = 1 VPP  
LMH6581 VOUT = 2 VPP, RL = 150Ω  
LMH6580 VOUT = 1 VPP  
425  
500  
,
MHz  
MHz  
,
450  
70  
GF  
0.1 dB Gain Flatness  
,
LMH6581 V OUT = 2 VPP, RL = 150Ω  
Time Domain Response  
tr  
Rise Time  
LMH6580 1V Step, LMH6581  
2V Step, 10% to 90%  
3.1  
1.4  
ns  
ns  
tf  
Fall Time  
LMH6580 1V Step, LMH6581  
2V Step, 10% to 90%  
OS  
SR  
Overshoot  
Slew Rate  
2V Step  
<1  
%
LMH6580, 2 VPP, 40% to 60%  
(Note 6)  
900  
V/µs  
Slew Rate  
LMH6581, 2 VPP, 40% to 60%  
(Note 6)  
1700  
7
V/µs  
ns  
ts  
Settling Time  
2V Step, VOUT within 0.5%  
Distortion And Noise Response  
HD2  
HD3  
en  
2nd Harmonic Distortion  
3rd Harmonic Distortion  
Input Referred Voltage Noise  
2 VPP, 10 MHz  
2 VPP, 10 MHz  
>1 MHz  
−76  
−76  
12  
dBc  
dBc  
nV/  
pA/  
dBc  
in  
Input Referred Noise Current  
>1 MHz  
2
XTLK  
ISOL  
Crosstalk  
All Hostile, f = 100 MHz  
f = 100 MHz  
−45  
−60  
Off Isolation  
dBc  
Static, DC Performance  
AV  
Gain  
LMH6581  
LMH6580  
1.986  
0.994  
2.00  
1.00  
±3  
2.014  
1.005  
±17  
VOS  
Input Offset Voltage  
mV  
µV/°C  
µA  
TCVOS  
IB  
Input Offset Voltage Average Drift (Note 10)  
38  
Non-Inverting (Note 9)  
−5  
Input Bias Current  
TCIB  
Non-Inverting (Note 10)  
−12  
Input Bias Current Average Drift  
nA/°C  
www.national.com  
2
Symbol  
VO  
Parameter  
Conditions  
LMH6581, RL = 100Ω  
Min  
(Note 8)  
Typ  
(Note 7)  
Max  
(Note 8)  
Units  
Output Voltage Range  
±1.8  
±1.24  
±2.08  
±1.25  
±2.1  
±1.3  
±2.2  
±1.3  
V
V
LMH6580, RL = 100Ω  
LMH6581, RL = Ω, (Note 11)  
LMH6580 RL = Ω,  
VO  
Output Voltage Range  
PSRR  
ICC  
Power Supply Rejection Ratio  
Positive Supply Current  
−45  
50  
dBc  
mA  
60  
56  
13  
RL = ∞  
RL = ∞  
IEE  
Negative Supply Current  
Tri State Supply Current  
50  
10  
mA  
mA  
RST Pin > 2.0V  
Miscellaneous Performance  
RIN  
CIN  
RO  
RO  
Input Resistance  
Non-Inverting  
Non-Inverting  
Closed Loop, Enabled  
LMH6580  
100  
1
kΩ  
pF  
Input Capacitance  
Output Resistance Enabled  
Output Resistance Disabled  
300  
50  
mΩ  
kΩ  
V
LMH6581  
1100  
2.0  
1350  
±1.3  
1500  
CMVR  
IO  
Input Common Mode Voltage  
Range  
Output Current  
Sourcing, VO = 0 V  
±50  
mA  
Digital Control  
VIH  
VIL  
Input Voltage High  
V
V
Input Voltage Low  
Output Voltage High  
Output Voltage Low  
Switching Time  
Setup Time  
0.8  
VOH  
VOL  
>2.0  
<0.4  
15  
V
V
ns  
ns  
ns  
TS  
TH  
7
Hold Time  
7
±5V Electrical Characteristics (Note 5)  
Unless otherwise specified, typical conditions are: TA = 25°C, AV = +2, VS = ±5V, RL = 100Ω; Boldface limits apply at the tem-  
perature extremes.  
Symbol  
Parameter  
Conditions  
Min  
Typ  
Max  
Units  
(Note 8)  
(Note 7)  
(Note 8)  
Frequency Domain Performance  
SSBW  
−3 dB Bandwidth  
VOUT = 0.5 VPP (Note 11)  
450  
500  
LMH6580 VOUT = 1 VPP  
LMH6581 VOUT = 2 VPP, RL = 1 kΩ  
LMH6580 VOUT = 1 VPP  
LMH6581 VOUT = 2 VPP, RL = 150Ω  
LMH6580, VOUT = 1 VPP  
,
MHz  
MHz  
LSBW  
GF  
,
450  
100  
0.1 dB Gain Flatness  
,
LMH6581, VOUT = 2 VPP, RL = 150Ω  
RL = 150Ω, 3.58 MHz/4.43 MHz  
RL = 150Ω, 3.58 MHz/4.43 MHz  
DG  
DP  
Differential Gain  
.05  
.05  
%
Differential Phase  
deg  
Time Domain Response  
tr  
Rise Time  
LMH6580 2V, Step, 10% to 90%  
LMH6581 2V, Step, 10% to 90%  
2V Step, 10% to 90%  
2.8  
1.2  
1.6  
<1  
ns  
tf  
Fall Time  
ns  
%
OS  
Overshoot  
2V Step  
3
www.national.com  
Symbol  
SR  
Parameter  
Conditions  
Min  
(Note 8)  
Typ  
(Note 7)  
Max  
(Note 8)  
Units  
V/µs  
V/µs  
ns  
Slew Rate  
Slew Rate  
LMH6580, 2 VPP, 40% to 60%  
(Note 6)  
1200  
2100  
6
SR  
LMH6581, 6 VPP, 40% to 60%  
(Note 6)  
ts  
Settling Time  
2V Step, VOUT Within 0.5%  
Distortion And Noise Response  
HD2  
HD3  
en  
2nd Harmonic Distortion  
3rd Harmonic Distortion  
Input Referred Voltage Noise  
2 VPP, 5 MHz  
2 VPP, 5 MHz  
>1 MHz  
−80  
−70  
12  
dBc  
dBc  
nV/  
pA/  
dBc  
in  
Input Referred Noise Current  
Cross Talk  
>1 MHz  
2
XTLK  
All Hostile, f = 100 MHz  
Channel to Channel, f = 100 MHz  
f = 100 MHz  
−45  
−52  
−65  
dBc  
dBc  
ISOL  
Off Isolation  
Static, DC Performance  
AV  
Gain  
LMH6581  
LMH6580  
1.986  
0.995  
2.00  
1.00  
±2  
2.014  
1.005  
±17  
Vos  
Input Offset Voltage  
mV  
µV/°C  
µA  
TCVos  
IB  
Input Offset Voltage Average Drift (Note 10)  
38  
Non-Inverting (Note 9)  
−5  
±12  
Input Bias Current  
TCIB  
VO  
Non-Inverting (Note 10)  
LMH681, RL = 100Ω  
LMH6580, RL = 100Ω  
LMH6581, RL = Ω  
−12  
±3.6  
nA/°C  
Input Bias Current Average Drift  
Output Voltage Range  
±3.4  
±2.9  
±3.7  
±2.9  
V
V
±3.0  
±3.9  
±3.0  
VO  
Output Voltage Range  
LMH6580, RL = Ω  
PSRR  
XTLK  
OISO  
ICC  
Power Supply Rejection Ratio  
DC Crosstalk Rejection  
DC Off Isloation  
DC  
−42  
−62  
−60  
−45  
−90  
−90  
54  
dBc  
dBc  
dBc  
mA  
DC, Channel to Channel  
DC  
Positive Supply Current  
66  
62  
17  
RL = ∞  
RL = ∞  
IEE  
Negative Supply Current  
Tri State Supply Current  
50  
14  
mA  
mA  
RST Pin > 2.0V  
Miscellaneous Performance  
RIN  
CIN  
RO  
RO  
Input Resistance  
Non-Inverting  
100  
1
kΩ  
pF  
Input Capacitance  
Non-Inverting  
Output Resistance Enabled  
Output Resistance Disabled  
Closed Loop, Enabled  
LMH6580, Resistance to Ground  
LMH6581, Resistance to Ground  
300  
50  
mΩ  
kΩ  
V
1100  
1300  
±3.0  
1500  
CMVR  
IO  
Input Common Mode Voltage  
Range  
Output Current  
Sourcing, VO = 0 V  
±60  
2.0  
±70  
mA  
Digital Control  
VIH  
VIL  
Input Voltage High  
V
V
Input Voltage Low  
Output Voltage High  
Output Voltage Low  
Switching Time  
Setup Time  
0.8  
VOH  
VOL  
>2.4  
<0.4  
15  
V
V
ns  
ns  
ns  
TS  
TH  
5
Hold Time  
5
www.national.com  
4
Note 1: Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. Operating Ratings indicate conditions for which the device is  
intended to be functional, but specific performance is not guaranteed. For guaranteed specifications, see the Electrical Characteristics tables.  
Note 2: Human Body Model, applicable std. MIL-STD-883, Method 3015.7. Machine Model, applicable std. JESD22-A115-A (ESD MM std. of JEDEC)  
Field-Induced Charge-Device Model, applicable std. JESD22-C101-C (ESD FICDM std. of JEDEC).  
Note 3: The maximum output current (IOUT) is determined by device power dissipation limitations.  
Note 4: The maximum power dissipation is a function of TJ(MAX), θJA. The maximum allowable power dissipation at any ambient temperature is  
PD = (TJ(MAX) – TA)/ θJA. All numbers apply for packages soldered directly onto a PC Board.  
Note 5: Electrical Table values apply only for factory testing conditions at the temperature indicated. No guarantee of parametric performance is indicated in the  
electrical tables under conditions different than those tested.  
Note 6: Slew Rate is the average of the rising and falling edges.  
Note 7: Typical values represent the most likely parametric norm as determined at the time of characterization. Actual typical values may vary over time and will  
also depend on the application and configuration. The typical values are not tested and are not guaranteed on shipped production material.  
Note 8: Room Temperature limits are 100% production tested at 25°C. Factory testing conditions result in very limited self-heating of the device such that TJ  
=
TA. Limits over the operating temperature range are guaranteed through correlation using Statistical Quality Control (SQC) methods.  
Note 9: Negative input current implies current flowing out of the device.  
Note 10: Drift determined by dividing the change in parameter at temperature extremes by the total temperature change.  
Note 11: This parameter is guaranteed by design and/or characterization and is not tested in production.  
Ordering Information  
Package  
Part Number  
LMH6580VS  
LMH6580VSX  
LMH6581VS  
LMH6581VSX  
Package Marking  
Transport Media  
250 Units/Tray  
NSC Drawing  
LMH6580VS  
1k Tape and Reel  
250 Units/Tray  
48-Pin QFP  
VBC48A  
LMH6581VS  
1k Tape and Reel  
5
www.national.com  
Typical Performance Characteristics LMH6580  
1 VPP Frequency Response  
1 VPP Frequency Response  
30007253  
30007254  
1 VPP Frequency Response Broadcast  
1 VPP Frequency Response Broadcast  
30007255  
30007256  
Frequency Response 1 kLoad  
Frequency Response 1kLoad  
30007274  
30007252  
www.national.com  
6
Frequency Response with Input Expansion  
Frequency Response with Input Expansion  
30007275  
30007276  
2 VPP Pulse Response  
2 VPP Pulse Response  
30007265  
30007263  
2 VPP Pulse Response, Broadcast Mode  
2 VPP Pulse Response, Broadcast Mode  
30007258  
30007264  
7
www.national.com  
1 VPP Pulse Response  
1 VPP Pulse Response  
30007261  
30007259  
Channel to Channel Crosstalk  
All Hostile Crosstalk  
30007277  
30007278  
Second Order Distortion (HD2) vs. Frequency  
Third Order Distortion (HD3) vs. Frequency  
30007272  
30007270  
www.national.com  
8
Second Order Distortion (HD2) vs. Frequency  
Third Order Distortion (HD3) vs. Frequency  
30007271  
30007273  
Positive Voltage Swing over Temperature  
Negative Voltage Swing over Temperature  
30007266  
30007267  
Positive Voltage Swing over Temperature  
Negative Voltage Swing over Temperature  
30007269  
30007268  
9
www.national.com  
Enabled Output Impedance  
Disabled Output Impedance  
Switching Time  
Enabled Output Impedance  
30007281  
30007279  
Disabled Output Impedance  
30007282  
30007280  
30007257  
www.national.com  
10  
Typical Performance Characteristics LMH6581  
2 VPP Frequency Response  
2 VPP Frequency Response  
30007248  
30007249  
Large Signal Bandwidth  
Large Signal Bandwidth  
30007222  
30007223  
Small Signal Bandwidth  
Small Signal Bandwidth  
30007224  
30007225  
11  
www.national.com  
Group Delay  
Frequency Response 1 kLoad  
30007245  
30007241  
2 VPP Pulse Response  
2 VPP Pulse Response  
30007214  
30007213  
4 VPP Pulse Response  
4 VPP Pulse Response Broadcast  
30007216  
30007217  
www.national.com  
12  
4 VPP Pulse Response  
6 VPP Pulse Response  
30007215  
30007218  
Off Isolation  
All Hostile Crosstalk  
30007219  
30007221  
Second Order Distortion (HD2) vs. Frequency  
Third Order Distortion (HD3) vs. Frequency  
30007227  
30007226  
13  
www.national.com  
Second Order Distortion vs. Frequency  
Third Order Distortion vs. Frequency  
30007228  
30007229  
No Load Output Swing  
Positive Swing over Temperature  
30007234  
30007238  
Negative Swing Over Temperature  
No Load Output Swing  
30007231  
30007239  
www.national.com  
14  
Positive Swing over Temperature  
Enabled Output Impedance  
Switching Time  
Negative Swing over Temperature  
30007236  
30007237  
Disabled Output Impedance  
30007250  
30007251  
30007257  
15  
www.national.com  
Application Information  
INTRODUCTION  
The LMH6580/LMH6581 are high speed, fully buffered, non-  
blocking, analog crosspoint switches. Having fully buffered  
inputs allows the LMH6580/LMH6581 to accept signals from  
low or high impedance sources without the worry of loading  
the signal source. The fully buffered outputs will drive 75or  
50back terminated transmission lines with no external com-  
ponents other than the termination resistor. When disabled,  
the outputs are in a high impedance state. The LMH6580/  
LMH6581 can have any input connected to any (or all) output  
(s). Conversely, a given output can have only one associated  
input.  
INPUT AND OUTPUT EXPANSION  
The LMH6580/LMH6581 have high impedance inactive  
states for both inputs and outputs allowing maximum flexibility  
for crosspoint expansion. In addition the LMH6580/LMH6581  
employ diagonal symmetry in pin assignments. The diagonal  
symmetry makes it easy to use direct pin to pin vias when the  
parts are mounted on opposite sides of a board. As an ex-  
ample two LMH6580/LMH6581 chips can be combined on  
one board to form either an 8 x 8 crosspoint or a 16 x 4 cross-  
point. To make an 8 x 8 crosspoint all 8 input pins would be  
tied together (Input 0 on side 1 to input 7 on side 2 and so on)  
while the 4 output pins on each chip would be left separate.  
To make the 16 x 4 crosspoint, the 4 outputs would be tied  
together while all 16 inputs would remain independent. In the  
16 x 4 configuration it is important not to have 2 connected  
outputs active at the same time. With the 8 x 8 configuration,  
on the other hand, having two connected inputs active is a  
valid state. Crosspoint expansion as detailed above has the  
advantage that the signal will go through only one crosspoint.  
Expansion methods that have cascaded stages will suffer  
bandwidth loss far greater than the small loading effect of  
parallel expansion.  
30007242  
FIGURE 1. Output Expansion  
Output expansion as shown in Figure 1 is very straight for-  
ward. Connecting the inputs of two crosspoint switches has a  
very minor impact on performance. Input expansion requires  
more planning. Input expansion, as show in Figure 2 and  
Figure 3 gives the option of two ways to connect the outputs  
of the crosspoint switches. In Figure 2 the crosspoint switch  
outputs are connected directly together and share one termi-  
nation resistor. This is the easiest configurarion to implement  
and has only one drawback. Because the disabled output of  
the unused crosspoint (only one output can be active at a  
time) has a small amount of capacitance, the frequency re-  
sponse of the active crosspoint will show peaking. This is  
illustrated in Figure 4 and Figure 5. In most cases this small  
amount of peaking is not a problem.  
As illustrated in Figure 3 each crosspoint output can be given  
its own termination resistor. This results in a frequency re-  
sponse nearly identical to the non expansion case. There is  
one drawback for the gain of 2 crosspoint, and that is gain  
error. With a 75termination resistor the 1250resistance  
of the disabled crosspoint output will cause a gain error. In  
order to counter act this the termination resistors of both  
crosspoints should be adjusted to approximately 80. This  
will provide very good matching, but the gain accuracy of the  
system will now be dependent on the process variations of  
the crosspoint resistors which have a variability of approxi-  
mately ±20%.  
30007243  
FIGURE 2. Input Expansion with Shared Termination  
Resistors  
www.national.com  
16  
30007247  
FIGURE 5. Input Expansion Frequency Response  
DRIVING CAPACITIVE LOADS  
Capacitive output loading applications will benefit from the  
use of a series output resistor ROUT. Capacitive loads of  
5 pF to 120 pF are the most critical, causing ringing, frequency  
response peaking and possible oscillation. Since most ca-  
pacitive loading is due to undesired parasitic capacitances the  
values of the capacitive loading will not usually be known ex-  
actly. It is best to start with a conservative value of ROUT and  
decrease the value until the bandwidth shows slight peaking.  
At this point the value of the isloation resistor will be deter-  
mined by whether flat frequency response or maximum band-  
width is the desired goal. Smaller values of ROUT will produce  
some peaking, but maximum bandwidth. Larger resistor val-  
ues will decrease bandwidth and suppress peaking.  
30007244  
FIGURE 3. Input Expansion with Separate Termination  
Resistors  
As starting values, a capacitive load of 5 pF should have  
around 75 of isolation resistance. A value of 120 pF would  
require around 12. When driving transmission lines, the out-  
put termination resistor is normally sufficient.  
30007246  
FIGURE 4. Input Expansion Frequency Response  
17  
www.national.com  
USING OUTPUT BUFFERING TO ENHANCE BANDWIDTH  
AND INCREASE REBIABILITY  
DIGITAL CONTROL  
Block Diagram  
The LMH6580/LMH6581 crosspoint switch can offer en-  
hanced bandwidth and reliability with the use of external  
buffers on the outputs. The bandwidth is increased by un-  
loading the outputs and driving the high impedance of an  
external buffer. See the Frequency Response 1 kLoad  
curve in the Typical Performance section for an example of  
bandwidth achieved with less loading on the outputs. For this  
technique to provide maximum benefit a very high speed am-  
plifier such as the LMH6703 should be used. As shown in  
Figure 6 the resistor RL is placed between the crosspoint out-  
put and the buffer amplifier. This resistor will provide a load  
for the crosspoint output buffer and reduce peaking caused  
by the buffer input capacitance. A recommended value for  
RL is 500to 1000. Higher values of RL will give higher  
bandwidth, but also higher peaking. The optimum value of  
RL will depend greatly on board layout and the input capaci-  
tance of the buffer amplifier.  
30007211  
Besides offering enhanced bandwidth performance using an  
external buffer provides greater system reliability. The first  
advantage is to reduce thermal loading on the crosspoint  
switch. This reduced die temperature will increase the life of  
the crosspoint. The second advantage is enhanced ESD re-  
liability. It is very difficult to build high speed devices that can  
withstand all possible ESD events. With external buffers the  
crosspoint switch is isolated from ESD events on the external  
system connectors.  
FIGURE 7.  
Edge  
Logic Pins  
Pin Name Level  
Triggered by  
Sensitive  
Triggered  
CLK  
CS  
Yes  
Yes  
Yes  
Yes  
CLK rising  
edge  
DATA IN  
CLK falling  
edge  
DATA  
OUT  
CLK rising  
edge  
CFG  
Yes  
Yes  
Yes  
Yes  
MODE  
RST  
30007240  
BCST  
FIGURE 6. Buffered Output  
There are two modes for programing the LMH6580/  
LMH6581, Serial Mode and Addressed Mode. The LMH6580/  
LMH6581 have internal control registers that store the pro-  
gramming states of the crosspoint switch. The logic is two  
staged to allow for maximum programming flexibility. The first  
stage of the control logic is tied directly to the crosspoint  
switching matrix. This logic consists of one register for each  
output that stores the on/off state and the address of which  
input to connect to. These registers are not directly accessible  
to the user. The second level of logic is another bank of reg-  
isters identical to the first, but set up as shift registers. These  
registers are accessed by the user via the serial input bus.  
CROSSTALK  
When designing a large system such as a video router  
crosstalk can be a very serious problem. Extensive testing in  
our lab has shown that most crosstalk is related to board lay-  
out rather than occurring in the crosspoint switch. There are  
many ways to reduce board related crosstalk. Using con-  
trolled impedance lines is an important step. Using well de-  
coupled power and ground planes will help as well. When  
crosstalk does occur within the crosspoint switch itself it is  
often due to signals coupling into the power supply pins. Using  
appropriate supply bypassing will help to reduce this mode of  
coupling. Another suggestion is to place as much grounded  
copper as possible between input and output signal traces.  
Care must be taken, though, not to influence the signal trace  
impedances by placing shielding copper too closely. One oth-  
er caveat to consider is that as shielding materials come  
closer to the signal trace the trace needs to be smaller to keep  
the impedance from falling too low. Using thin signal traces  
will result in unacceptable losses due to trace resistance. This  
effect becomes even more pronounced at higher frequencies  
due to the skin effect. The skin effect reduces the effective  
thickness of the trace as frequency increases. Resistive loss-  
es make crosstalk worse because as the desired signal is  
attenuated with higher frequencies crosstalk increases at  
higher frequencies.  
The LMH6580/LMH6581 is programmed via a serial input bus  
with the support of four other digital control pins. The Serial  
bus consists of a clock pin (CLK), a serial data in pin (DIN),  
and a serial data out pin (DOUT). The serial bus is gated by a  
chip select pin (CS). The chip select pin is active low. While  
the chip select pin is high all data on the serial input pin and  
clock pins is ignored. When the chip select pin is brought low  
the internal logic is set to begin receiving data by the first  
positive transition (0 to 1) of the clock signal. The chip select  
pin must be brought low at least 5 ns before the first rising  
edge of the clock signal. The first data bit is clocked in on the  
next negative transition (1 to 0) of the clock signal. All input  
data is read from the bus on the negative edge of the clock  
signal. Once the last valid data has been clocked in, either the  
www.national.com  
18  
chip select pin must go high or, the clock signal must stop.  
Otherwise invalid data will be clocked into the chip. The data  
clocked into the chip is not transferred to the crosspoint matrix  
until the CFG pin is pulsed high. This is the case regardless  
of the state of the MODE pin. The CFG pin is not dependent  
on the state of the Chip select pin. If no new data is clocked  
into the chip subsequent pulses on the CFG pin will have no  
effect on device operation.  
figure pin. One way around this loss of flexibility would be if  
the clock signal is generated by an FPGA or microcontroller  
where the clock signal can be stopped after the data is  
clocked in. In this case the Chip select function is provided by  
the presence or absence of the clock signal.  
SERIAL PROGRAMMING MODE  
Serial programming mode is the mode selected by bringing  
the MODE pin low. In this mode a stream of 16-bits programs  
all four outputs of the crosspoint. The data is fed to the chip  
as shown in the Serial Mode Data Frame tables below (two  
tables are required to show the entire data frame). The table  
is arranged such that the first bit clocked into the crosspoint  
register is labeled bit number 0. The register labeled Load  
Register in the block diagram is a shift register. If the chip  
select pin is left low after the valid data is shifted into the chip  
and if the clock signal keeps running then additional data will  
be shifted into the register, and the desired data will be shifted  
out.  
The programming format of the incoming serial data is se-  
lected by the MODE pin. When the MODE pin is HIGH the  
crosspoint can be programmed one output at a time by en-  
tering a string of data that contains the address of the output  
that is going to be changed (Addressed Mode). When the  
mode pin is LOW the crosspoint is in Serial Mode. In this  
mode the crosspoint accepts a 16 bit array of data that pro-  
grams all of the outputs. In both modes the data fed into the  
chip does not change the chip operation until the Configure  
pin is pulsed high. The configure and mode pins are inde-  
pendent of the chip select pin.  
Also illustrated is the timing relationships for the digital pins  
in the Timing Diagram for Serial Mode shown below. It is im-  
portant to note that all the pin timing relationships are impor-  
tant, not just the data and clock pins. One example is that the  
Chip Select pin (CS) must transition low before the first rising  
edge of the clock signal. This allows the internal timing circuits  
to synchronize to allow data to be accepted on the next falling  
edge. The chip select pin must then transition high after the  
final data bit has been clocked in and before another clock  
signal positive edge occurs to prevent invalid data from being  
clocked into the chip. Another way to accomplish the same  
thing is to strobe the clock pin with only the desired number  
of pulses starting and ending with clock in the low condition.  
The configure (CFG) pin timing is not so critical, but it does  
need to be kept low until all data has been shifted into the  
crosspoint registers.  
THREE WIRE VS. FOUR WIRE CONTROL  
There are two ways to connect the serial data pins. The first  
way is to control all four pins separately, and the second op-  
tion is to connect the CFG and the CS pins together for a 3  
wire interface. The benefit of the 4-wire interface is that the  
chip can be configured independently using the CS pin. This  
would be an advantage in a system with multiple crosspoint  
chips where all of them could be programmed ahead of time  
and then configured simultaneously. The 4-wire solution is  
also helpful in a system that has a free running clock on the  
CLK pin. In this case, the CS pin needs to be brought high  
after the last valid data bit to prevent invalid data from being  
clocked into the chip.  
The 3-wire option provides the advantage of one less pin to  
control at the expense of having less flexibility with the con-  
30007209  
Timing Diagram for Serial Mode  
19  
www.national.com  
Serial Mode Data Frame (First Two Words)  
Output 0  
Input Address  
LSB  
Output 1  
On = 0  
Off = 1  
3
Input Address  
On = 0  
Off = 1  
7
MSB  
2
LSB  
4
LSB  
0
1
5
6
Off = TRI-STATE®, Bit 0 is first bit clocked into device.  
Serial Mode Data Frame (Continued)  
Output 2  
Output 3  
Input Address  
LSB  
On = 0  
Off = 1  
11  
Input Address  
LSB  
On = 0  
Off = 1  
15  
MSB  
10  
MSB  
14  
8
9
12  
13  
ADDRESSED PROGRAMMING MODE  
first rising edge of the clock signal. This allows the internal  
timing circuits to synchronize to allow data to be accepted on  
the next falling edge. The chip select pin must then transition  
high after the final data bit has been clocked in and before  
another clock signal positive edge occurs to prevent invalid  
data from being clocked into the chip. Also, in addressed  
mode is it necessary for the clock signal to make a low to high  
transition after the chip select pin has been brought high. If  
there is not a low to high transition of the clock after the chip  
select pin goes high subsequent data wil not be loaded into  
the chip properly. The configure (CFG) pin timing is not criti-  
cal, but it does need to be kept low until all data has been  
shifted into the crosspoint registers.  
Addressed programming mode makes it possible to change  
only one output register at a time. To utilize this mode the  
mode pin must be High. All other pins function the same as  
in serial programming mode except that the word clocked in  
is 5 bits and is directed only at the output specified. In ad-  
dressed mode the data format is shown below in the table  
titled Addressed Mode Word Format.  
Also illustrated is the timing relationships for the digital pins  
in the Timing Diagram for Addressed Mode shown below. It  
is important to note that all the pin timing relationships are  
important, not just the data and clock pins. One example is  
that the Chip Select pin (CS) must transition low before the  
30007210  
Timing Diagram for Addressed Mode  
www.national.com  
20  
Addressed Mode Word Format  
Output Address  
Input Address  
TRI-STATE  
LSB  
MSB  
1
LSB  
2
MSB  
4
1 = TRI-STATE  
0 = On  
0
3
5
Bit 0 is first bit clocked into device.  
DAISY CHAIN OPTION IN SERIAL MODE  
lated based on the typical supply current of 50 mA and a 10V  
supply voltage. This power dissipation will vary within the  
range of 0.4 W to 0.6 W due to process variations. In addition,  
each equivalent video load (150) connected to the outputs  
should be budgeted 30 mW of power. For a typical application  
with one video load for each output this would be a total power  
of 0.62 W. With a θJA of 44 °C/W this will result in the silicon  
being 27°C over the ambient temperature. A more aggressive  
application would be two video loads per output which would  
result in 0.74 W of power dissipation. This would result in a  
33°C temperature rise. For heavier loading, the TQFP pack-  
age thermal performance can be significantly enhanced with  
an external heat sink and by providing for moving air ventila-  
tion. Also, be sure to calculate the increase in ambient tem-  
perature from all devices operating in the system case.  
Because of the high power output of this device, thermal  
management should be considered very early in the design  
process. Generous passive venting and vertical board orien-  
tation may avoid the need for fan cooling or heat sinks. Also,  
the LMH6580/LMH6581 can be operated with a ±3.3V power  
supply. This will cut power dissipation substantially while only  
reducing bandwidth by about 10% (2 VPP output). The  
LMH6580/LMH6581 are fully characterized and factory tested  
at the ±3.3V power supply condition for applications where  
reduced power is desired.  
The LMH6580/LMH6581 supports daisy chaining of the serial  
data stream between multiple chips. This feature is available  
only in the Serial Programming Mode. To use this feature se-  
rial data is clocked into the first chip DIN pin, and the next chip  
DIN pin is connected to the DOUT pin of the first chip. Both chips  
may share a chip select signal, or the second chip can be  
enabled separately. When the chip select pin goes low on  
both chips a double length word is clocked into the first chip.  
As the first word is clocking into the first chip the second chip  
is receiving the data that was originally in the shift register of  
the first chip (invalid data). When a full 16 bits have been  
clocked into the first chip the next clock cycle begins moving  
the first frame of the new configuration data into the second  
chip. With a full 32 clock cycles both chips have valid data and  
the chip select pin of both chips should be brought high to  
prevent the data from overshooting. A configure pulse will ac-  
tivate the new configuration on both chips simultaneously, or  
each chip can be configured separately. The mode, chip se-  
lect, configure and clock pins of both chips can be tied to-  
gether and driven from the same sources.  
SPECIAL CONTROL PINS  
The LMH6580/LMH6581 have two special control pins that  
function independent of the serial control bus. One of these  
pins is the reset (RST) pin. The RST pin is active high mean-  
ing that at logic 1 level the chip is configured with all outputs  
disabled and in a high impedance state. The RST pin pro-  
grams all the registers with input address 0 and all the outputs  
are turned off. In this configuration the device draws only 11-  
mA. The RST pin can be used as a shutdown function to  
reduce power consumption. The other special control pin is  
the broadcast (BCST) pin. The BCST pin is also active high  
and sets all the outputs to the on state connected to input 0.  
This is sometimes referred to as broadcast mode, where input  
0 is broadcast to all eight outputs.  
PRINTED CIRCUIT LAYOUT  
Generally, a good high frequency layout will keep power sup-  
ply and ground traces away from the input and output pins.  
Parasitic capacitances on these nodes to ground will cause  
frequency response peaking and possible circuit oscillations  
(see Application Note OA-15 for more information). If digital  
control lines must cross analog signal lines (particularly in-  
puts) it is best if they cross perpendicularly. National Semi-  
conductor suggests the following evaluation boards as a  
guide for high frequency layout and as an aid in device testing  
and characterization:  
THERMAL MANAGEMENT  
Device  
Package  
Evaluation Board  
Part Number  
The LMH6580/LMH6581 are high performance devices that  
produce a significant amount of heat. With ±5V supplies, the  
LMH6580/LMH6581 will dissipate approximately 0.5 W of  
idling power with all outputs enabled. Idling power is calcu-  
LMH6580  
48–Pin  
LMH730164EF  
21  
www.national.com  
Physical Dimensions inches (millimeters) unless otherwise noted  
48-Pin QFP  
NS Package Number VBC48A  
www.national.com  
22  
Notes  
23  
www.national.com  
Notes  
THE CONTENTS OF THIS DOCUMENT ARE PROVIDED IN CONNECTION WITH NATIONAL SEMICONDUCTOR CORPORATION  
(“NATIONAL”) PRODUCTS. NATIONAL MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY  
OR COMPLETENESS OF THE CONTENTS OF THIS PUBLICATION AND RESERVES THE RIGHT TO MAKE CHANGES TO  
SPECIFICATIONS AND PRODUCT DESCRIPTIONS AT ANY TIME WITHOUT NOTICE. NO LICENSE, WHETHER EXPRESS,  
IMPLIED, ARISING BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS  
DOCUMENT.  
TESTING AND OTHER QUALITY CONTROLS ARE USED TO THE EXTENT NATIONAL DEEMS NECESSARY TO SUPPORT  
NATIONAL’S PRODUCT WARRANTY. EXCEPT WHERE MANDATED BY GOVERNMENT REQUIREMENTS, TESTING OF ALL  
PARAMETERS OF EACH PRODUCT IS NOT NECESSARILY PERFORMED. NATIONAL ASSUMES NO LIABILITY FOR  
APPLICATIONS ASSISTANCE OR BUYER PRODUCT DESIGN. BUYERS ARE RESPONSIBLE FOR THEIR PRODUCTS AND  
APPLICATIONS USING NATIONAL COMPONENTS. PRIOR TO USING OR DISTRIBUTING ANY PRODUCTS THAT INCLUDE  
NATIONAL COMPONENTS, BUYERS SHOULD PROVIDE ADEQUATE DESIGN, TESTING AND OPERATING SAFEGUARDS.  
EXCEPT AS PROVIDED IN NATIONAL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, NATIONAL ASSUMES NO  
LIABILITY WHATSOEVER, AND NATIONAL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY RELATING TO THE SALE  
AND/OR USE OF NATIONAL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR  
PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY  
RIGHT.  
LIFE SUPPORT POLICY  
NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR  
SYSTEMS WITHOUT THE EXPRESS PRIOR WRITTEN APPROVAL OF THE CHIEF EXECUTIVE OFFICER AND GENERAL  
COUNSEL OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein:  
Life support devices or systems are devices which (a) are intended for surgical implant into the body, or (b) support or sustain life and  
whose failure to perform when properly used in accordance with instructions for use provided in the labeling can be reasonably expected  
to result in a significant injury to the user. A critical component is any component in a life support device or system whose failure to perform  
can be reasonably expected to cause the failure of the life support device or system or to affect its safety or effectiveness.  
National Semiconductor and the National Semiconductor logo are registered trademarks of National Semiconductor Corporation. All other  
brand or product names may be trademarks or registered trademarks of their respective holders.  
Copyright© 2007 National Semiconductor Corporation  
For the most current product information visit us at www.national.com  
National Semiconductor  
Americas Customer  
Support Center  
National Semiconductor Europe  
Customer Support Center  
Fax: +49 (0) 180-530-85-86  
National Semiconductor Asia  
Pacific Customer Support Center  
Email: ap.support@nsc.com  
National Semiconductor Japan  
Customer Support Center  
Fax: 81-3-5639-7507  
Email:  
new.feedback@nsc.com  
Tel: 1-800-272-9959  
Email: europe.support@nsc.com  
Deutsch Tel: +49 (0) 69 9508 6208  
English Tel: +49 (0) 870 24 0 2171  
Français Tel: +33 (0) 1 41 91 8790  
Email: jpn.feedback@nsc.com  
Tel: 81-3-5639-7560  
www.national.com  

相关型号:

LMH6581

8x4 500MHz Analog Crosspoint Switch, Gain of 1, Gain of 2
NSC

LMH6581VS

8x4 500MHz Analog Crosspoint Switch, Gain of 1, Gain of 2
NSC

LMH6581VS/NOPB

8-CHANNEL, CROSS POINT SWITCH, PQFP48, LEAD FREE, TQFP-48
TI

LMH6581VSX

8-CHANNEL, CROSS POINT SWITCH, PQFP48
TI

LMH6582

16x8 500 MHz Analog Crosspoint Switch, Gain of 1/Gain of 2
NSC

LMH6582YA

16x8 550 MHz Analog Crosspoint Switch, Gain of 1
NSC

LMH6582_07

16x8 550 MHz Analog Crosspoint Switch, Gain of 1
NSC

LMH6582_0708

16x8 550MHz Analog Crosspoint Switch, Gain of 1
NSC

LMH6583

16x8 500 MHz Analog Crosspoint Switch, Gain of 1/Gain of 2
NSC

LMH6583

增益为 2 的 5V、交叉点/交换、16 输入 8 输出模拟开关
TI

LMH6583YA

16x8 550 MHz Analog Crosspoint Switch, Gain of 2
NSC

LMH6583YA/NOPB

增益为 2 的 5V、交叉点/交换、16 输入 8 输出模拟开关 | PAP | 64 | -40 to 85
TI