U.S. patent application number 11/435559 was filed with the patent office on 2006-11-30 for method and apparatus for equalizing video transmitted over twisted pair cable.
This patent application is currently assigned to RGB Systems, Inc.. Invention is credited to Mike A. Andrews, Gary Dean Cole, Art Garcia, Manfred Schneider.
Application Number | 20060268167 11/435559 |
Document ID | / |
Family ID | 27789449 |
Filed Date | 2006-11-30 |
United States Patent
Application |
20060268167 |
Kind Code |
A1 |
Cole; Gary Dean ; et
al. |
November 30, 2006 |
Method and apparatus for equalizing video transmitted over twisted
pair cable
Abstract
A apparatus and apparatus for compensating for video insertion
loss due to transmission over long twisted pair cable lines is
presented. Transmission of video over twisted pair cable is
advantageous because of its superior cost advantage over coaxial
cable. However, twisted pair cables have significant loss
characteristics at the higher frequencies (i.e., broadband)
compared to coaxial cables. At a transmitter station, the video
signal is amplified in the high frequency region for possible skin
effect losses thereby brute forcing the high frequency components
to the receiving station. At the receiver station, the video signal
is further compensated for diffusion line and skin effect losses.
The total skin effect compensation applied in both the transmitter
and receiver stations is such that the square root of frequency
characteristics of skin effect losses is compensated for. Thus, at
the receiving station, the high frequency compensation added at the
transmitter to brute force the high frequency components to the
receiving station may be removed if found excessive. Additionally,
compensation is included to adjust for skew that may occur because
of irregularities between the various twisted pairs used to
transmit the individual video components. Non-minimum phase type
filters are used to inject delay into the faster arriving signals
so that they may coincide in phase with later arriving signals
resulting in a true reproduction of the video.
Inventors: |
Cole; Gary Dean; (Corona,
CA) ; Schneider; Manfred; (Costa Mesa, CA) ;
Garcia; Art; (San Dimas, CA) ; Andrews; Mike A.;
(Anaheim, CA) |
Correspondence
Address: |
THE HECKER LAW GROUP
1925 CENTURY PARK EAST
SUITE 2300
LOS ANGELES
CA
90067
US
|
Assignee: |
RGB Systems, Inc.
Anaheim
CA
|
Family ID: |
27789449 |
Appl. No.: |
11/435559 |
Filed: |
May 16, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09877984 |
Jun 8, 2001 |
7047556 |
|
|
11435559 |
May 16, 2006 |
|
|
|
Current U.S.
Class: |
348/398.1 ;
348/397.1; 348/553; 348/577; 348/E7.051 |
Current CPC
Class: |
H04N 7/108 20130101;
Y10S 348/914 20130101 |
Class at
Publication: |
348/398.1 ;
348/397.1; 348/553; 348/577 |
International
Class: |
H04N 11/02 20060101
H04N011/02; H04N 7/12 20060101 H04N007/12; H04N 9/64 20060101
H04N009/64; H04N 5/14 20060101 H04N005/14; H04N 11/04 20060101
H04N011/04; H04N 5/44 20060101 H04N005/44 |
Claims
1.-24. (canceled)
25. A method for equalizing video transmitted over twisted pair
cable comprising: receiving an input signal having a plurality of
video components; compensating one or more of said plurality of
video components for transmission; transmitting each of said
plurality of video components over a twisted pair cable to a
receiving station; applying compensation to each of said plurality
of video components for accumulated amplitude losses due to said
transmission over said twisted pair cable; and generating a
plurality of output video components by applying desired phase
compensation using adjustable non-minimum phase filters on one or
more of said plurality of video components, said non-minimum phase
filters having poles and zeros configured to provide said desired
phase compensation.
26. The method of claim 25, wherein said plurality of video
components comprises a color system's color components.
27. The method of claim 25, wherein each of said plurality of video
components comprises a high frequency portion and a low frequency
portion.
28. The method of claim 27, wherein said compensating said
plurality of video components for transmission comprises
application of high frequency boosting on each of said plurality of
video components.
29. The method of claim 25, wherein said applying compensation for
accumulated amplitude losses comprises: applying low frequency
shaping to account for diffusion effect losses occurring during
transmission over said twisted pair cable; and applying high
frequency shaping to account for skin effect losses occurring
during transmission over said twisted pair cable.
30. The method of claim 29, wherein said high frequency shaping
comprises a compensation network that effectively compensates for
said skin effect losses.
31. The method of claim 25, wherein said twisted pair cable is
unshielded twisted pair.
32. The method of claim 25, wherein said compensation for
accumulated amplitude losses is adjustable.
33. The method of claim 25, wherein said compensating said
plurality of video components for transmission is performed in a
transmitting station.
34. The method of claim 25, wherein said compensation for
accumulated amplitude losses is performed in said receiving
station.
35. An apparatus for equalizing video transmitted over twisted pair
cable comprising: an input signal with a plurality of video
components; a first compensation circuit coupled to said plurality
of video components; a plurality of twisted pair cables coupled to
said compensation circuit for transmitting said plurality of video
components to a receiving station; a second compensation circuit in
said receiving station coupled to said twisted pair cables for
compensation of said plurality of video components for accumulated
amplitude losses due to transmission over said twisted pair cables;
and a phase delay circuit coupled to said second compensation
circuit for application of desired phase delay to one or more of
said plurality of video components.
36. The apparatus of claim 35, wherein said plurality of video
components comprises a color system's color components.
37. The apparatus of claim 35, wherein each of said plurality of
video components comprises a high frequency region and a low
frequency region.
38. The apparatus of claim 37, wherein said first compensation
circuit comprises amplifier circuits to boost said high frequency
region of said plurality of video components.
39. The apparatus of claim 35, wherein said second compensation
circuit comprises: low frequency shaping filters to account for
diffusion effect losses occurring during transmission over said
twisted pair cable; and high frequency shaping filters to account
for skin effect losses occurring during transmission over said
twisted pair cable.
40. The apparatus of claim 39, wherein said high frequency shaping
filter comprises a compensation network that effectively
compensates for said skin effect losses.
41. The apparatus of claim 35, wherein said twisted pair cable is
unshielded twisted pair.
42. The apparatus of claim 35, wherein said phase delay circuit
comprises a non-minimum phase zero filter having adjustable poles
and zeros.
43. The apparatus of claim 35, wherein said second compensation
circuit is adjustable.
44. The apparatus of claim 36, wherein said color system is RGB.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] This invention relates to the field of video insertion loss
equalization. More specifically the invention relates to
compensating for losses in analog video quality due to transmission
over twisted pair cable.
[0003] 2. Background Art
[0004] Video cables are used to convey electronic video signals
from a source device such as a receiver to a destination, typically
a display device. A cable is supposed to accurately convey the
signal, however, losses accumulate along the cable path because of
imperfections in a transmission cable. These imperfections are not
necessarily due to manufacturing but due to the fact that a cable
is a physical device and most physical devices exhibit some losses
when a signal is conveyed through them. Thus, the longer the cable
length, the more losses accumulate. The accumulated transmission
loss is known by those of skill in the art as cable insertion loss.
Of course, other devices such as switches and splitters in a video
transmission path contribute to the total video loss however only
transmission cable insertion losses are considered in this
specification.
[0005] Video may be transmitted either in digital or analog
formats. For digital video transmission such as computer video,
cable insertion loss is generally not an issue because the digital
signal can be recovered so long as discernable digital pulses are
received at the receiving station. However, for analog signals such
as NTSC (National Television Standards Committee) video, the signal
is just voltages, and voltages are affected by wire length,
connectors, heat, cold, and other conditions. This degrading effect
on the video signal caused by the transmission cable length is
known as cable insertion loss.
[0006] Analog video such as the C-Video, S-Video, or YUV (or YIQ)
specifications may be available in any of the different color
models. A color model (also color space) facilitates the
specification of colors in some standard, generally accepted way,
for example, the RGB color model where R is for the red component,
G is for the green component, and B is for the blue component. For
high-resolution analog video, each color component is usually
transmitted separately from the receiver to the display device.
Thus, each color component must be examined for cable insertion
losses.
[0007] A coaxial cable is typically used for transmission of high
resolution (i.e., broadband) video because of its superior
performance over twisted pair cable. However, coaxial cables are
more expensive and difficult to install compared with twisted pair
cables. Historically, the significant differences between coaxial
cable and twisted pair cables limited twisted pair transmission to
low-resolution video (i.e., less than 10 MHz). However, twisted
pair has one distinct advantage over coaxial, i.e.,
cost/performance ratio. Dollar-for-dollar, twisted pairs are
significantly cheaper than coax or fiber (i.e., fiber optic cable)
to buy and install. And, while the coax and fiber offer some
advantage in bandwidth, this advantage may be offset with proper
compensation. Also, standard twisted pair, i.e., UTP/STP
(Unshielded/Shielded Twisted Pair) cable contains four pairs in a
single cable so that the actual cost per pair is one-quarter of the
per-foot price. Therefore, transmission of broadband analog video
over twisted pair cable is highly desirable.
[0008] The idea of high-data rate twisted pairs started with an
attempt to organize cables into levels. In the process, it was
recognized that twisted pairs could be made into higher grade data
versions. These data-grade twisted pairs come in two flavors:
twisted pairs commonly called UTP (unshielded twisted pairs) and
STP (shielded twisted pairs). By far, most of the domestic data
installations today employ UTP.
[0009] In the mid 1980's, twisted pair technology began to emerge
which could transmit 2 Mbps, then 4 Mbps (the original IBM data
rate), and then 10 Mbps. As these data rates increased, it became
apparent that some way of indicating the performance of the cable
was needed. It was then that the system of "Levels" was suggested.
The TIA/EIA, two groups that set standards for the data industry,
adopted the plan and separated the data rates and other parameters
into "Categories", such as Category 3, 4, and 5. Category 4 cables
are becoming extinct. Each higher numbered category has more
stringent requirements with higher data rates and higher
performance than the previous category. The specifications for
Categories are given in TIA/EIA 568A.
[0010] Current twisted pair cables look identical to the plain old
telephone service cable. They use the same color code and come in
many of the same pair counts and use the same gage conductors.
However, the specifications they are made to, the materials used to
make them, and the requirements to connect them, become more and
more critical as the data rate increases.
[0011] Twisted pair cable has many current common uses, such as in
low-resolution video display systems and in business and
corporation local area networks. With respect to low frequency
performance of video displays, twisted pair cable carries
horizontal line time, at frequencies that range from 15 to 50 KHz,
for example. In order to display a complete video picture, each
horizontal line sweep of the electron beam needs to be coordinated
with color information and vertical line information. In computer
monitors and displays, the picture information is displayed in
"progressive" scan, i.e., one horizontal line right after another.
This gives these displays excellent detail (which is required in
the computer field) but requires double the amount of information
as an interlaced analog broadcast receiver to display a picture at
the same resolution.
[0012] Timing pulses can be used to ensure that each step in the
process of transmitting video information occurs at the right time.
However, although the video signals may be properly timed, all
transmission lines have a characteristic signal loss in any system
for which they are utilized. This characteristic signal loss in
response to an applied electromagnetic wave along a cable is the
insertion loss previously discussed. The efficiency of a
transmission line carrying such a signal affects graphics image
response and the degree of streaking across a picture screen. Low
frequency streaking (or smear) across a video display screen, also
known as picture effect, creates an overall appearance of detail
loss and discontinuity critical to an observer's eye.
[0013] There are three key problems using twisted pair (e.g., UTP)
for analog video. First, the majority of video equipment uses
coaxial connectors, usually BNCs. The second is the output
impedance of the coaxial systems is 75.OMEGA. (ohms) while UTP has
an impedance of 100.OMEGA.. The third difference is that UTP being
twisted pairs is a "balanced line" system while coax, which has
only one shielded conductor, is "unbalanced". Prior art devices
attempt to solve these problems with a device called a "balun".
Balun, in fact, means BALanced to Unbalanced. The balun is a small
box which contains a transformer and other matching components
allowing the signal to be converted from 75.OMEGA. impedance to
100.OMEGA. impedance. It can also change the signal from unbalanced
to balanced; using a BNC connector (commonly used for coax) on one
side of the box (i.e., input side) and an RJ-45, the most common
UTP connector, on the other side (i.e., output side). However,
baluns can only handle very narrow bandwidth and no DC component
(basically a bandpass filter).
[0014] Balance is also a critical parameter. The nature of a
balanced line means that the two conductors in the twisted pair are
identical in length. The more identical they are, and the closer
they are together, the easier it is for the balun to reject noise
and interference generated outside the pair. When noise hits both
conductors, and a resultant noise "pulse" is generated in both
wires, the more identical the noise on each wire at the end of the
cable, the greater the noise rejection can be (`common mode
rejection`) in a balanced line. The less identical, for example,
standard telephone service lines are often very unequal, the more
noise will get through.
[0015] Signal reconstruction at the far end depends on balance.
Twisted pair is a difference technology. The effect in the receiver
is not dependent on the voltages on the wires but the difference in
the voltages on the wires. Currents in the wires that are identical
will not contribute to the received signal. Practical circuits have
limits in voltage range and current delivery to compensate for
common mode voltages and currents. The difference in chassis ground
voltages can be on the order of 5 volts at 500 feet in most
industrial locations. Digital circuitry has large currents (20 to
50 amps in their ground planes causing 200 mV noise). Thus, to keep
common mode signals from being converted into difference signals
the impedances experienced by each line must be identical.
Common UTP/STP Properties
[0016] ACR is "attenuation-to-crosstalk ratio". By subtracting the
attenuation from the crosstalk, a number is generated which can
indicate the overall performance of a cable. Positive ACR,
especially at high frequencies, can be an indicator of cable
performance. On the other hand, it is possible to dramatically
improve crosstalk, and thereby improve ACR, by unusually tight
twisting of the conductor pairs.
[0017] Skew or Delay Skew is timing differences on a multi-pair
cable. Skew is especially interesting when using more than one
twisted pair to simultaneously deliver data, which is especially
important with analog video transmission. In such systems, it is
essential that the signals arrive at the other end of the cable at
the same time. For example, the R, G, and B components of a
high-resolution video are transmitted on separate wire pairs.
Transmission Lines
[0018] The next level of approximation deals with the behavior of
the twisted pair line when the length of the line exceeds the
longest wavelength of interest, which is almost always the case
with video transmission. Under these conditions the classical RLC
telephone equations apply for frequencies of the energy propagated
throughout the line.
[0019] There are three frequency regions of interest: The diffusion
region in which Resistance (i.e., R) per unit length is greater
than Inductance (i.e., L) per unit length, the standard
transmission telephone equation region, and the skin effect region.
In the diffusion region (i.e., where R dominates over L), the
classical telegraph equations apply. The Skin effect region, which
is above 5 MHz with CAT5 twisted pair, is characterized by a loss
that increases in proportion to the square root of frequency. In
the limit the telephone equations devolve into the telegraph
equations. Thus, the diffusion region and the telephone region may
be combined and solved as a diffusion line problem, which can be
approximated by several RC networks distributed over several
decades of frequencies.
[0020] A typical insertion loss response indicates that for a given
twisted pair cable, as much as two thirds of the insertion loss
occurs in the first few megahertz of its operating frequency range.
This loss is mainly due to diffusion in the wire. The magnitudes of
each of these forms of signal loss are dependent on the effective
capacitance of the wire. The combination of a long cable run with
higher direct current (DC) loss coupled with the changing
resistivity at low frequencies due to diffusion effect causes the
transmission line to appear more like an R-C time response. This
time response change causes medium to large details on the image to
create a trailing streak as current is changing direction in the
cable. The condition is commonly referred to as short-time or
line-time distortion when seen in a video processing circuit which
has a long response time constant, or time necessary for the system
to reach steady state. The R-C like response time constant is
related to the distributed capacitance times the distributed
resistance, and it is generally accepted that the current becomes
constant after five times the distance integral of this time
constant. It is this long time constant that degrades the image to
an unacceptable point in most cases.
Impedance
[0021] Impedance indicates the ability of a data cable to attach
and transfer energy from one box to another. The impedance of the
cable is determined by the physical construction of the system. The
TIA/EIA standard for Category 5 is 100.OMEGA..+-.15.OMEGA.. Some
Category 5 cables meet this spec. Others require the use of a
smoothing formula called "Zo-fit". This allows manufacturers to
ignore rapid changes in impedance. While many Enhanced Category 5
manufacturers say that their cable is "tested to 300 MHz" (or even
up to 400 MHz), they offer no data (impedance, crosstalk,
attenuation, ACR, skew etc.) at those high frequencies. Impedance
variation can often be 100.OMEGA..+-.50.OMEGA., rendering these
cables virtually useless for uncompensated video transmission.
Bandwidth
[0022] Bandwidth is the range of frequencies available to be used
for signal carrying. It is the "size of the tunnel". Bandwidth is a
measure of the signal-carrying capacity of different cables. For
broadband video, uniform amplitude and delays across colors are
required.
[0023] Cable insertion loss can be characterized as a function of
frequency (i.e., rate of the video input). FIGS. 1A and 1B
illustrate such characteristics for a 300 feet Category 5 twisted
pair cable. FIG. 1A shows the square root of frequency
characteristics of the skin effect loss, while FIG. 1B is a more
traditional view using a log-log scale. Insertion loss is specified
in decibels (dB). As shown in the illustrations of FIGS. 1A and 1B,
the insertion loss is approximately 0 dB when the input frequency
is near zero, i.e., the output will be approximately equal to the
input if the input is constant for a long period of time. In
addition, the insertion loss increases as the input frequency
increases. In general, insertion loss increases as the cable length
over which the analog video is transmitted increases. This is why
compensation is generally not needed and usually not applied for
short cable runs, e.g., six feet or less.
[0024] Several compensation techniques may be used to compensate
for cable insertion losses, however, in broadband systems where
signals range from the steady state to several hundred megahertz,
it is nearly unfeasible in terms of cost to compensate the entire
frequency spectrum (typically 0-300 MHz). It is particularly
problematic because of the shape of the frequency response
characteristics of the insertion loss which shows rapid drop off in
the low frequency range (0-10 MHz) and then followed by a shallow
drop over the remainder of the frequency range (see FIG. 1A). The
insertion loss characteristics at the low frequency region are due
to diffusion in the cable. Cables of different type or length have
different diffusion rates. Current industry methods effectively
compensate for the high frequency end of the spectrum but not the
low frequency end because of the complex frequency response
characteristics associated with insertion loss of twisted pair
cables. The problem with compensating for high frequency losses
without compensating for the low frequency losses in video
transmitted over long twisted pair cable is that people are more
tolerant to seeing less sharp pictures (high frequency effects) so
long as they can see the information. However, people are least
tolerant of low frequency anomalies which are characterized by
distortion or smearing type phenomenon across the video screen.
[0025] FIG. 2 is an illustration of a setup for a video display
from a source signal to a destination display device. The input
video signal is V.sub.IN, and the output video signal displayed on
display screen 200 is V.sub.OUT. As discussed earlier, video input
signal V.sub.IN is an analog voltage signal thus video amplifier
204 is used in a video receiver to condition the voltage to the
level desired by the display device. Generally, video amplifier 204
conditions the analog video input signal to compensate for
transmission line loss such that the proper video signal reaches
display device 200. Block 208 is the operational amplifier gain and
block 206 is the current feedback gain (generally representing
resistive dividers). The feedback is from the output of the
operational amplifier through resistors (represented by block B
206) to the negative input terminal of the operational amplifier.
Thus, in this configuration, the transfer function between the
input and output voltage of the video amplifier is A/[1+AB].
[0026] Twisted pair cable line 202 represents the total cable
length the video must travel from the input source to display
device 200. Thus, assuming there is no change in voltage at VGA to
TP Converter 204, FIG. 1 would represent the frequency responses
between input signal V.sub.IN and output signal V.sub.OUT for a
300-foot twisted pair cable. FIG. 3 shows the transient response
characteristics of a fixed length twisted pair cable. Using FIG. 2
as an illustration, a step input (e.g., 300), which is
characterized as having frequency content from zero to
approximately infinity, is introduced at V.sub.IN. The transient
response output V.sub.OUT (e.g., 310) is shown in FIG. 3. Region
320 represents the low frequency diffusion effects while region 330
represents the high frequency skin effects of the transmission
cable. The effect of the low frequency characteristic of the
transmission cable is also evidenced by the large rise time, i.e.,
time it takes the output to reach 90% of steady state. The gradual
rise of the response (i.e., 310) over a long period of time appears
as a smear or shadow effect on the uncompensated video signal at
display 200. Large rise time is synonymous with low bandwidth.
Thus, the twisted pair cable introduces a low bandwidth filter
effect on the input video signal passing through it. Generally,
rise time increases with decreasing characteristic bandwidth and
longer cable lengths tend to produce lower characteristic
bandwidth.
[0027] To illustrate the filter effect of the twisted pair cable,
FIG. 4 shows a low pass filter being used as an example to
represent the effect of cable line 202 of FIG. 2. That is, the
low-pass filter represents the transfer function between the input
and output of the twisted pair cable. Using low pass filter 400
with bandwidth .omega. representing the characteristic bandwidth of
the cable length, V.sub.OUT is given by the equation: V OUT = 1 [ 1
+ AB ] .times. 1 [ 1 .omega. .times. s + 1 ] .times. V IN
##EQU1##
[0028] Thus, the goal is to make bandwidth .omega. as large as
possible so that the frequency response between the input and
output videos remains as flat as possible to a frequency high
enough that any distorting effects caused by the transmission line
are not discernable by the human eye. One way to compensate for the
effect of the low bandwidth .omega. of the cable characteristics is
to include the twisted pair cable line loss in the feedback loop as
shown in FIG. 5. Thus, instead of the current feedback originating
directly from the output of video amplifier 504, cable
characteristic 400 is included in the closed loop and thereby
compensated for in video amplifier 504. The resulting transfer
function between input V.sub.IN and output V.sub.OUT is given as
follows: A V OUT = [ 1 + AB ] .times. 1 [ 1 ( 1 + AB ) .times.
.omega. .times. s + 1 ] .times. V IN ##EQU2##
[0029] Thus, the new bandwidth between the input and output voltage
is increased by the factor (1+AB) and thus directly controllable by
the gains chosen for the video amplifier feedback. However, this
implementation would require feedback from a long (e.g., 200 feet)
cable back to the video (i.e., operational) amplifier 504 which is
located at the source. Firstly, this is impractical and secondly
there is a significant amount of time delay involved in feeding
back from the long cable runs. Long feedback delays tend to cause
loop instability and thus limits loop gains that can be used in the
video amplifier. The most practical way of compensating for the
cable insertion loss characteristics is to duplicate those
characteristics in the feedback path as shown in FIG. 6 thus
eliminating the time delay involved in feeding back from a long
cable run.
[0030] FIG. 6 illustrates the most practical way of compensating
for cable insertion loss. Mathematically, this provides the same
results as FIG. 5 above, i.e., the loop bandwidth is increased by
[1+AB]. Assuming the gain B is unity, then V.sub.OUT approaches
V.sub.IN as A increases. In a perfect situation, Block 600
replicates the inverse of the cable characteristics. However, cost
prohibitions may make it impractical to create analog circuitry for
block 602 in video amplifier 600 that duplicates the cable
insertion loss characteristics 400 in a wide frequency range as
shown in FIG. 1. Thus, prior art implementations tend to only
compensate for the high frequency losses by adding "peaking"
compensation, i.e., pole-zero pair to boost the response at the
high frequencies.
SUMMARY OF THE INVENTION
[0031] This invention defines a method and apparatus for enhancing
or improving the quality of high-resolution video images by
compensating for the losses caused by transmission of the video
signal over long twisted pair cables. The loss that occurs during
transmission of the video signal is known as cable insertion loss.
Transmission of video over twisted pair cable is advantageous
because of its superior cost advantage over coaxial cable. However,
twisted pair cables have significant loss characteristics at the
higher frequencies (i.e., broadband) compared to coaxial cables. In
one embodiment of the invention, the video signal is
pre-compensated at a transmitter station for possible skin effect
losses before being injected into the twisted pair line for
transmission to a receiver station. The pre-compensation at the
transmitter station amplifies the high frequency region to brute
force the high frequency signals to the receiver station.
Amplifying the high frequency spectrum of the video at the
transmitter station prevents boosting of noise that may have been
picked-up during transmission over the twisted pair cables to the
receiver station, which may be up to 1000 feet from the transmitter
station.
[0032] At the receiver station, the video signal is further
compensated for diffusion line and skin effect losses. The total
skin effect compensation applied in both the transmitter and
receiver stations is such that square root of frequency
characteristics of skin effect losses is compensated for. Thus, at
the receiving station, the high frequency compensation added at the
transmitter to brute force the high frequency signals to the
receiving station may be removed if found excessive (i.e.,
unnecessary because of short transmission line length, for
example). Embodiments of the present invention add compensation as
desired to cover the frequency range of interest. For example,
several RC circuits may be cascaded to increase the video
transmission bandwidth.
[0033] An embodiment of the invention further adds compensation to
adjust for skew that may occur because of irregularities between
the various twisted pairs used to transmit the individual video
components. Non-minimum phase type filters are used to inject delay
into the faster arriving signals so that they may coincide in phase
with later arriving signals thus resulting in true reproduction of
the video.
BRIEF DESCRIPTION OF THE DRAWINGS
[0034] FIGS. 1A-1B are illustrations of the frequency response
characteristics of the insertion loss in a twisted pair cable.
[0035] FIG. 2 is an illustration of a schematic for typical video
display originating from a source signal to a destination display
device.
[0036] FIG. 3 is the transient response characteristics of a fixed
length twisted pair cable.
[0037] FIG. 4 is an illustration using a low pass filter to
represent the effect of cable insertion loss characteristics.
[0038] FIG. 5 is an illustration of a technique to compensate for
the effect of the insertion loss through the twisted pair cable by
including the entire twisted pair cable line in the video amplifier
feedback loop.
[0039] FIG. 6 is an illustration of a technique to compensate for
the effect of the insertion loss through the twisted pair cable by
including the characteristic effects of the twisted pair cable line
in the video amplifier feedback loop.
[0040] FIG. 7 is a functional block diagram illustrating
equalization of video transmitted over twisted pair cable in
accordance with an embodiment of the present invention.
[0041] FIG. 8 is a functional block diagram of the twisted pair
transmitter in accordance with an embodiment of the present
invention.
[0042] FIG. 9 is an example compensation circuitry in the
transmitter for a color component in accordance with an embodiment
of the present invention.
[0043] FIG. 10 is a functional block diagram of the twisted pair
receiver is, in accordance with an embodiment of the present
invention.
[0044] FIG. 11 is an illustration of the differential receiver
block according to an embodiment of the present invention.
[0045] FIG. 12 is a frequency response illustrating the resulting
video signal with high frequency compensation for skin effect in
accordance with an embodiment of the present invention.
[0046] FIG. 13 is a schematic illustrating low frequency
equalization in accordance with an embodiment of the present
invention.
[0047] FIG. 14 is an illustration of the effect of a non-minimum
phase zero filter on an input signal with fixed frequency.
[0048] FIG. 15 is a representative circuit diagram illustrating an
implementation of the non-minimum phase filter in accordance with
an embodiment of the present invention.
[0049] FIG. 16 is a transient response representation of the video
response with low frequency equalization.
[0050] FIG. 17 is an illustration of a feed-forward compensation
technique in accordance with an embodiment of the present
invention.
[0051] FIG. 18 is an illustration of the receiver high frequency
compensation.
DETAILED DESCRIPTION OF THE INVENTION
[0052] An embodiment of the invention comprises a method and
apparatus for equalization of video transmitted over long twisted
pair cables. In the following description, numerous specific
details are set forth to provide a more thorough description of
embodiments of the invention. It will be apparent, however, to one
skilled in the art, that the invention may be practiced without
these specific details. In other instances, well known features
have not been described in detail so as not to obscure the
invention.
[0053] An embodiment of the invention provides a method and
apparatus for enhancing and improving the quality of video images
by compensating for the loss and other effects of transmitting
video over a medium such as twisted pair cables. For instance, one
embodiment of the invention uses a configuration consisting of
compensation for the diffusion effects, skin effects, and skew due
to dissimilarities between different cable pairs.
[0054] FIG. 7 is a functional block diagram illustrating
equalization of video transmitted over twisted pair cable in
accordance with an embodiment of the present invention. The
invention comprises Twisted Pair Transmitter (TPT) 704 and Receiver
(TPR) 708 connected together by twisted pair cable 706. TPT 704
receives video signal from source 702 via cable 703. Source 702
could be a laptop computer, for example, and cable 703 could be a
short coaxial cable. Video, comprising the various color-components
(e.g., RGB) and the horizontal and vertical sync pulses, and audio
are transmitted from source 702 to TPT 704 for additional
processing. TPT 704 processes the individual color components,
including the sync and audio signals, by adding appropriate
compensation in accordance with embodiments of the present
invention. For example, TPT 704 adds high frequency compensation
for skin effect losses by adding peaking compensation. That is, the
response at the high frequency region is boosted before the signal
is inserted into the twisted pair line 706. However, this high
frequency compensation may be removed at TPR 708 if not desired.
For example, the compensation may be removed if the length of cable
706 is insufficient to justify the compensation at the TPT.
Typically, when twisted line pair 706 is of considerable length,
compensation must be added to restore the video quality. The issue
then becomes where to add the compensation?
[0055] Depending on the desired video bandwidth, the range of
frequency desiring compensation may be considerable. For example,
broadband video may require compensation from the low Kilohertz
region to above 100 MHz region. Thus, several RC compensation
networks may be required to obtain a flat frequency response over
such a broad range of frequencies. The desired compensation may be
divided into low and high frequency compensation. Care must be
exercised during high frequency compensation to prevent from
boosting noise picked up during the video transmission. An
embodiment of the prevent invention adds high frequency boosting at
the transmitter to assure reproduction of the video, not noise, at
the receiver. Thus, by adding the high frequency lead (i.e.,
peaking) compensation at the transmitter (i.e., TPT 704), the
present invention prevents potential boosting of noise at the
receiver (i.e., TPR 708) to compensate for transmission line loss.
The TPT (e.g., 704) transmits the compensated video and audio
signals to TPR 708 via twisted pair cable 706. Cable 706 comprises
a plurality of twisted pair cables that may be of the shielded
(STP) or unshielded (UTP) variety and of appropriate category
(e.g., Cat 5). Each twisted pair carries a color component from the
transmitter to the receiver.
[0056] Because twisted pair is a difference technology, signal
reconstruction at the far end (i.e., at TPR 708) of a twisted pair
system depends on balance in the transmission line (e.g., 706).
Therefore, the effect in the receiver is not dependent on the
voltages on the wires but the difference in the voltages on the
wires. Thus, currents in the wires that are identical will not
contribute to the received signal. Practical circuits have limits
in voltage range and current delivery to compensate for common mode
voltages and currents. Embodiments of the present invention employ
impedances lower than many common mode noise sources thereby
resulting in amplitudes less than 9 volts from the noise sources
thus providing better common mode noise rejection. Therefore,
embodiments of the present invention may operate reliably with up
to 9 volts peak common mode signal whereas prior art systems
operate approximately with 1 volt common mode rejection.
[0057] Many common mode noise sources have impedances higher than
that found in the present invention and as a result will not
develop much common mode voltage. For Example, Conducted EMI
(Electro-Magnetic Interference) filters connected to chassis ground
which are not connected to a common earth ground have the potential
to deliver 5 ma or 55 V RMS to the transmitter (i.e., TPT 704). The
TPT can handle approximately 20 times the power available from this
noise source with an excellent resulting picture. The receiver, TPR
708, performs additional processing after receiving the video
signals from twisted line pair 706. This processing may include:
removing some or all of the high frequency compensation added at
TPT 704 if the line length of cable 706 is insufficient to justify
the compensation, adding low frequency compensation for the
diffusion characteristics of twisted pair line 706, adding
additional high frequency compensation as desired, adding
compensation for skew that may occur between the color components
that are transmitted over different twisted pairs, DC (i.e., steady
state) restore level translation (because the video is negatively
biased at the TPT before injection into the twisted pair line), and
sync pulse recovery.
[0058] After processing the video and audio signals, TPR 708
transmits the processed signals to destination device 710 using an
appropriate transmission medium 709. Destination device 710 could
be a projector, for example. Using the methods of the present
invention, the broadband video display at destination 710 appears
clear and without the deformities that would otherwise be evident
with video signals transmitted over twisted pair cable of
significant length.
[0059] FIG. 12 is a frequency response illustrating the resulting
video signal with high frequency compensation for skin effect in
accordance with an embodiment of the present invention. The
illustration shows the effect of two simple zeros, cut in at
different frequencies, to approximately counteract the skin effect
losses. The zeros are illustrated by the humps at 1202 and 1204.
The hump at 1202 may represent the high frequency compensation
applied at the transmitter, for example, while the hump at 1204 may
represent the compensation applied at the receiver. Thus multiple
RC circuits may be used to maintain a flat frequency response up to
a desired frequency. The net effect of both zeros is to flatten the
frequency response (i.e., less than .+-.3 dB) up to 100 MHz,
approximately. Note that for good screen graphics at display 710,
good frequency response characteristics from 50 Hz to 100 MHz is
especially important.
Twisted Pair Transmitter
[0060] The twisted pair transmitter, TPT 704, provides
pre-compensation of the signal to be transmitted to the video
display, i.e., TPT 704 prepares the video and audio signals for
transmission over the twisted pair cable. FIG. 8 is a functional
block diagram of the twisted pair transmitter in accordance with an
embodiment of the present invention. At block 802, the horizontal
and vertical sync signals 801, received from an input video source,
are used to generate sync drive signals. The sync drive signals are
then summed, in block 804, with selected color components of video
input signal 803. For example, in an embodiment using the RGB color
system (i.e., video input signal 803 is in the RGB format), the
horizontal sync drive may be injected into the blue (i.e., B)
component while the vertical sync drive is injected into the green
(i.e., G) component. It will be apparent, however, to one skilled
in the art, that the choice of blue and green color components for
sync drive injection is arbitrary. An embodiment of the present
invention recovers the sync pulses at the receiver; thus, it is
desirable to know in which color components the appropriate sync
pulses are carried.
[0061] The transmitter further compensates the color components for
potential transmission line (i.e., insertion) losses at block 804.
As discussed earlier, because there is no appreciable region of the
classical LC transmission line, the video insertion loss can be
grouped into two major loss regions for compensation: the diffusion
line region, and skin effect region. From the transmitter view the
diffusion line looks like an RC circuit with the capacitor directly
loading the source impedance. The lower the source impedance, the
closer the resultant launched signal will be to the input signal.
Thus, an embodiment of the TPT drives with the lowest practical
source resistance which can maintain acceptable balance and
stability with the driving operational amplifier. If a near perfect
termination is used at the TPR, the reflected energy is minimal due
to the rapid onset of skin effect losses.
[0062] In an embodiment of the present invention, the TPT
compensates for the potential high frequency loss because boosting
at the transmitter instead of at the receiver assures that video,
not noise, is being amplified. The high frequency compensation is
for skin effect losses which generally applies for frequencies
greater than approximately 3 MHz, for a twisted pair cable. Since
video may be transmitted over twisted pair cable ranging from 50 to
1000 feet in distance, it is desirable to provide appropriate
compensation to brute force high frequency signals to the receiver.
For example, it is desirable to provide high frequency compensation
covering approximately two decades of frequency. Thus, an
embodiment of the present invention utilizes compensation in both
the transmitter and receiver to achieve the desired high frequency
coverage.
[0063] The transmitter, in accordance with an embodiment of the
present invention, uses feed forward complex RC Voltage dividers to
optimize the high frequency level for the target maximum distance.
Embodiments of the present invention utilize feed-forward
compensation instead of feedback to prevent loop instabilities and
thus allowing the use of as many RC networks as needed to provide
the desired frequency response. FIG. 17 is an illustration of a
feed-forward compensation technique in accordance with an
embodiment of the present invention.
[0064] In this illustration, a simple low-pass filter with transfer
function 1704 is used to represent the transmission line
characteristics (e.g., diffusion or skin effect) of the twisted
cable. In order to compensate for the transmission line loss (i.e.,
shift the bandwidth as high as possible), feed-forward compensation
is added in the form of high-pass filter 1702 in block 1706. Block
1706 may be contained in the receiver, for example, if used for
diffusion line compensation. It may also be contained in either the
transmitter or the receiver if used for skin effect compensation.
The resulting transfer function between the input, V.sub.IN, and
output, V.sub.OUT, is given by: V OUT = 1 [ .tau. 1 .times. s + 1 ]
.times. [ ( .tau. 2 + k ) .times. s + 1 ] [ .tau. 2 .times. s + 1 ]
.times. V IN ##EQU3##
[0065] Because the purpose of the compensation is to shift the
total system response bandwidth upwards, .tau..sub.2 is usually
much less than .tau..sub.1 (where .tau. is the inverse of
bandwidth). Therefore, the bandwidth of the response will be
shifted to the inverse of .tau..sub.2 when (.tau..sub.2+k) equals
.tau..sub.1. Thus, embodiments of the present invention provide the
capability to adjust the gain k (e.g., bias of the variable
transconductance amplifier discussed below) to achieve the desired
frequency response. In one or more embodiments, a series of
feed-forward compensation 1706 may be cascaded to broaden the
system bandwidth.
[0066] FIG. 9 is an example compensation circuitry in the
transmitter for a color component in accordance with an embodiment
of the present invention. For example, RC group 915 and 916 may
provide the lead (i.e., feed forward) compensation while RC group
915 and 917 control the bandwidth of the lead-compensating filter.
In other embodiments, devices such as a voltage variable
capacitance (a.k.a. hyperabrupt varicap) diode may be used to
achieve the desired lead compensation. It will be apparent to those
of skill in the art that other methods and devices may be used to
achieve the feed-forward compensation and thus is not limited to
the structure described herein. The circuit of FIG. 9 is for
illustrative purposes only and is not intended to limit the scope
herein. For example, other circuits may include different
combination and arrangement of resistors and capacitors (e.g.,
variable capacitors and/or resistors). Further, note that in this
illustration, a sync drive signal is injected into the color
component signal via the transistor at 902, into the inputs of
differential amplifier circuitry 910, before the high frequency
(i.e., skin effect) compensation is applied. Other embodiments may
inject the sync drive after the compensation is applied or at some
intermediate point. Note that when the transistor is off, it does
not contribute noise to the active video.
[0067] The transmitter converts the single ended input 901, i.e., a
color component of the video input, plus any included sync signal
902 into differential signals and thus generating negative output
920 and positive output 930 having the appropriate high frequency
amplification. Amplifiers 912 and 913 provide the appropriate
voltage drive through impedance matching resistors 908N and 908P.
Block 906 is supply filtering while block 904 is transient pulse
absorption.
[0068] Referring back to FIG. 8, the transmitter also provides
conversion (e.g., at block 806) of the single ended input signal,
which may be from a 75-Ohm source, for example, into differential
signals with appropriate impedance. The differential signals are
negatively biased before injection in order to prevent metallic ion
migration (i.e., dissipation) of physical cabling during long term
operation in a humid environment. For example, as illustrated in
FIG. 8, the DC level that was 0V at the output of block 804 is
translated to -5V at the output of block 806.
[0069] Impedance conversion in accordance with an embodiment is
illustrated in resistor blocks 908N and 908P of FIG. 9. Impedance
has two aspects: One is launching the signal into the line and the
other is receiving the signal from the line. A diffusion line has a
major portion of its wire to wire and wire to world capacitance
available to load the source resistance. To minimize the effect of
this capacitance, driving the signal into the line with the lowest
practical resistance is desired. This becomes the lowest frequency
compensation for the diffusion line problem. Since it is desirable
to keep the source impedance equal to the line impedance to
minimize reflections, and because of common mode considerations, an
embodiment of the present invention uses termination with three
parallel resistors, illustrated in blocks 908N and 908P, on each
leg of the difference line driver. For example, three resistors
having values of 100 Ohm each may be used thereby resulting in 33
Ohm driving resistance.
[0070] All the energy not absorbed by the far end terminations at
the receiver is reflected back down the transmission line to the
transmitter. Thus, it is imperative that the end termination (i.e.,
at the receiver) be as perfect as possible. When using CAT-5, for
example, the reflected energy for all frequencies above 2 MHz may
be minimized with a 100-Ohm load. For frequencies below 2 MHz, the
diffusion equations for twisted pair cables do not exhibit the
reflection phenomena because inductance is not an important current
driver.
[0071] However, it is desirable that the output differential
drivers be selected for high power and speed in order to provide
linear response with significant common mode currents and to
provide the slew rate required to overcome high skin effect losses
at appreciable distances.
[0072] Referring back to FIG. 8, additional function of the TPT
includes single ended to balanced audio drivers at block 808 and
power conversion circuit (i.e., power for the remote receiver), if
desired, at block 810. The compensated video, audio, and power
signals are then injected into the twisted pair cable at block 812
for transmission to the receiver 708 (FIG. 7).
Twisted Pair Receiver
[0073] The receiver (e.g., 708) provides the necessary compensation
at the other end of the twisted pair cable system. A functional
block diagram of the twisted pair receiver is illustrated in FIG.
10, in accordance with an embodiment of the present invention. The
twisted pair receiver comprises the differential receiver function
1002, the low frequency equalization function 1004, the skew
adjustment block 1006, high frequency equalization block 1008,
steady state (i.e., DC) restore level transition block 1010, and
sync recovery block 1012.
[0074] The differential receiver block is further illustrated in
FIG. 11. The differential receiver converts the common mode voltage
into current and then virtually grounds the current through
capacitance block 1102 to avoid input common mode restrictions of
the operational amplifiers. It uses a two-stage amplifier process
to control the peak common mode rejection. In one embodiment, a
resistor 1108 is placed between the junction of the term resistors
(e.g., 1107 and 1109) and the signal reference (e.g., 1111) to
reduce the transmitter current. The resistor restricts common
current and thus allows higher common mode voltages to avoid
transmitter saturation. Circuit 1104 also reduces the gain of the
input circuit to values less than unity thereby preventing
amplification of noise by the differential amplifier. Finally, the
common current component is passed through to differential
amplifier 1106 to allow independent setting of the common current
gain to a desired level while keeping a high difference gain.
Low Frequency Equalization
[0075] The low frequency compensation is basically an inverse
diffusion problem. Basically, one or more CR (i.e., high-pass)
filter circuits are used to compensate for the diffusion line
problem. Note that the inverse characteristic of the diffusion line
may be duplicated with proper choices of corner frequency of the
filters. As discussed earlier, the diffusion line's response is
affected by distance. Thus, the diffusion characteristics vary for
different length cables. Therefore, any diffusion compensation
should operate properly for varying cable lengths, which are
typically greater than 50 feet. In some cases, compensation may be
required for cable lengths lesser than 50 feet. One or more
embodiments of the present invention allow shifting of the
compensation network's effective zeros and poles in order to adjust
the circuit for cable length. The shifting may be accomplished by
extending the "Miller Effect" idea by adjusting the gain of the
signal delivered to the compensation network. For example, a
compensation network that uses transconductance amplifiers provides
means for adjustable input gains. FIG. 13 is a schematic
illustrating low frequency equalization in accordance with an
embodiment of the present invention. Low frequency equalization, or
inverse diffusion compensation, of input video component 1301
comprises smear-bias control block 1302, transistor 1312,
capacitors 1314, load resistor 1306, variable transconductance
amplifier 1309, and operational amplifier 1316. Smear-Bias 1302 is
used to control the gain of variable transconductance amplifier
1309 by adjusting output voltage 1304 through variable resistor
1303. That is, the loop gain may be modified by varying the bias of
variable transconductance amplifier 1309. By varying the bias of
the emitter coupled transistor pair (e.g., 1308 and 1310), current
is divided between the transistors. The amount of current directed
to the load resistor (e.g., 1306) determines the gain of the
circuit. Almost no compensation is required for short distances
therefore the gain may be reduced to almost zero (see k in FIG.
17). However, more compensation (i.e., higher gain) would be
required for longer cable lengths because of the increase in
diffusion line loss with distance. Note that if the gain of the
compensation signal is zero it has no effect on the signal path in
that the compensation network contributes no current to the virtual
ground at 1314. High linearity may be maintained by placing the
video signal 1301 in the emitter follower portion of the amplifier.
A plurality of low frequency compensation, using the technique
illustrated in FIG. 13, may be cascaded to provide the necessary
frequency range coverage.
[0076] Due to the frequency dependency of the transmission line
from a lumped, to diffusion, to low loss, then skin effect,
frequencies in these regions propagate to the load at different
rates. An embodiment of the present invention uses an anti-smear
circuitry 1302 to provide compensating phase advance thus providing
for a more faithful reproduction of the signal.
[0077] FIG. 16 is a transient response representation of the video
response with low frequency equalization. As discussed earlier,
lines 300 and 310 are the step function input and diffusion line
responses, respectively. Line 1602 represents the transient
response after inverse diffusion compensation. As shown,
embodiments of the present invention significantly improve the rise
time of the video signal thus compensating for the diffusion
effects of transmission over twisted pair cables.
Skew Adjustment
[0078] Components of the video may arrive at the receiver at
different times since multiple pairs are used to simultaneously
deliver the video from the transmitter to the receiver. These
delays or skew in data arrival time must be minimized or
compensated for in order to properly reproduce the video. Delays
may be due to reasons such as: wire length disparities, varying
impurities, switches, splitters, etc. In one embodiment of the
present invention, compensation is added using non-minimum zero
phase type filter to delay the faster arriving signal to coincide
with the phase of the later signal. The non-minimum phase zero
filter is an all-pass type filter which adds delay equal to 1/2
wave period of its pole frequency for signal energy with frequency
higher than the pole frequency. A non-minimum phase filter with the
pole (denominator) frequency equal to the zero (numerator)
frequency provides analog delay compensation equivalent to
e.sup.-TS, where T is the desired delay in seconds and is
equivalent to one-half the time constant of the filter.
[0079] FIG. 14 is an illustration of the effect of a non-minimum
phase zero filter on an input signal with fixed frequency. The
non-minimum phase filter initially reverses the phase of the
incoming signal 1402 thus resulting in an output signal 1404 with a
phase shift (i.e., delay) equivalent to 1/2 of its pole frequency.
FIG. 15 is a representative circuit diagram illustrating an
implementation of the non-minimum phase filter in accordance with
an embodiment of the present invention. Input signal 1501 is passed
through the non-minimum phase filter to add the needed skew/delay
thus adjusting the phase at output signal 1510. Embodiments of the
invention may provide for adjustment of the pole/zero frequency
using variable resistor 1502, for example.
High Frequency Equalization
[0080] High frequency compensation at the receiver comprises
compensating for the skin effect losses that occurred during
transmission of the video signal from the transmitter to the
receiver over twisted pair cable. For example, for the high
frequency compensation, the receiver may remove unneeded high
frequency compensation applied at transmitter 704, or further
assist this high frequency compensation. In either case, the
purpose of the compensation at the receiver is to match and thus
recover the square root of frequency characteristic loss caused by
skin effect. FIG. 18 is an illustration of the receiver high
frequency compensation.
[0081] This simplified illustration comprises resistor 1802,
capacitor 1801, amplifier 1803, and multiplier 1804. Recovery of
the square root of frequency characteristic loss of skin effect may
be done by modification of the effect of the capacitor 1801 in a
low-pass filter (RC network) by applying signal to its reference
leg (i.e., 1804). If no signal is present at 1804, the full effect
of the capacitor is felt in the filter. If the signal is the same
as that on the R (i.e., 1802) of the RC filter then the capacitor
contributes no current to the system and the circuit is opened. If
the signal is more than that which is applied to the R then the low
pass filter is transformed into a high pass filter. Embodiments of
the invention may employ a standard variable transconductance
amplifier to control the level of the signal. The gain may be set
by a current division in the common emitter circuit. The gain of
the stage determines the pole of the succeeding RC low pass filter.
An additional capacitor may be added in the signal path in the 50
MHz region to partially compensate for the square root of frequency
characteristic of skin effect losses.
[0082] Thus, a method and apparatus for equalization of video
insertion loss due to transmission over twisted pair cable lines
have been described in conjunction with one or more specific
embodiments. The invention is defined by the claims and their full
scope of equivalents.
* * * * *