¿ìè¶ÌÊÓƵ

(see Graphic)
(see Graphic)
Just over the horizon is a promised land where people watch TV on wide,
cinema-quality screens in the comfort of their homes, business executives
jot their notes into hand-held electronic notepads, and children escape to
other lands by looking through electronic spectacles. But before this
nirvana can become reality, the consumer electronics industry needs to
produce reliable, high-resolution, colour flat screens.

Japanese electronics giants are already claiming victory in the race to
produce one key technology. Over the past five years a handful of these
companies have invested $3 billion in factories to build flat
screens, and in April NEC announced that it had produced a thin screen
measuring 33 cm in diagonal that displays full-colour pictures with a finer
definition than those produced by conventional televisions with cathode-ray
tubes (CRT).

But despite such successes, the race to dominate the flat screen market is
far from over. The potential rewards of designing flat screens, suitable for
devices as diverse as high definition television and electronic books, are
so great that in many countries flat screens are an issue of national pride.
This has prompted researchers in the US and Europe – as well as some
maverick groups in Japan – to back rival technologies, in a bid to capture
the market with lower costs, and address the need for very large screens.
There are also still some question marks over the underlying technologies
used in flat-panel displays. These are often difficult and expensive to
manufacture, and the results don’t always satisfy today’s sophisticated
customers.

Shooting from behind

One of the major technological challenges facing flat screen researchers is
that the CRT displays they want to replace are good at what they do: they
can display rapidly changing pictures in natural-looking colours, they are
visible in daylight, and they come in a wide range of sizes, from pocket
displays to almost the size needed for HDTV, which can be a metre in
diagonal. Above all, CRTs are cheap to make. Producing ordinary colour CRTs
is essentially a 1950s technology and plants in Southeast Asia can churn out
standard screens for a few dollars apiece.

The main problem with CRTs is their depth – the electron ‘guns’ are best
placed some distance behind the screen so that the beams painting the
picture can scan across the face of the tube at a constant velocity.
Researchers have tried to reduce the space required by CRTs, for example, by
firing rays from the side of the screen and then deflecting the image. But
the resulting displays were too small and the images too indistinct. And
when engineers tried to increase the size and introduce colour, the complex
optics degraded the image hopelessly.

By the early 1980s, manufacturers had turned their attention to replacing
CRTs altogether. The driving force came from two new applications for
displays, portable computers and HDTV. Engineers at Japan’s state-owned
broadcasting corporation, NHK, decided that the next generation of TV sets
would have rectangular screens over a metre in diagonal, showing pictures
made up of twice the number of lines of conventional screens. If this new
era of television was to rely on the existing CRT technology then the
resulting products would be impossibly large and bulky – a fact that became
apparent with the release of the first HDTVs that appeared in Japanese shops
last year, which were the size of chest freezers and cost Pounds Sterling
4000. NHK recognised that a flat screen would be essential if HDTV was to
become anything more than the latest toy for the rich, consumer electronics
buff.

At about the same time as NHK was making its plans for flat screens,
computer manufacturers were getting excited about portable computers. By the
mid-1980s, some personal computer manufacturers had attempted to produce
portable machines using CRT displays, but the results were disappointing.
The machines were bulky, heavy and expensive, and became the butt of jokes
rather than an essential item in the executive’s briefcase. Like NHK,
computer manufacturers such as Toshiba decided some form of flat-panel
display was the answer.

One technology that caught the eye of these companies was the liquid-crystal
display (LCD). This display is based on a discovery made by RCA, the
American company now owned by General Electric, in the 1960s. It found that
the alignment of liquid crystals – chemicals that behave optically as a
crystal when liquid – could be controlled with an electric field. So under
certain conditions the crystals could be lined up to let a dot of light
through in response to the electrical signal.

Some of the first companies to commercialise this idea were the Japanese
consumer electronics firms that made cheap watches and calculators. They
decided to use LCD displays because they consumed far less electricity than
the alternative light-emitting diode displays. However, the quality of these
early LCDs was questionable; poor contrast and a narrow viewing angle made
them unsuitable for anything larger than a watch face. And it wasn’t until
companies began to look at ways of improving the underlying design of LCDs
that it became practical to consider the technology for larger screens.

By the time Toshiba launched the T1000 family (Dynabook in Japan) – the
first real notebook computers complete with full-size screen and disc drives
– in 1989, LCD design had seen a number of enhancements. ‘Supertwisted
nematic’ displays superseded the simple passive-matrix LCD, which contained
no active electronics, consisting soley of a thin film of liquid crystal
placed between two polarising filters and a grid of electrodes running
vertically and horizontally across the device. These supertwisted nematic
displays used chemicals with nematic or rod-shaped crystals, which turned
through 180degree, 240degree or 270degree. These were arranged in spiral
patterns so that an electrical signal changed the alignment of the crystals,
altering the polarisation and the amount of light transmitted. The overall
result was a monochrome display that looked almost as clear as a CRT.

But even as Toshiba was basking in the glory of the Dynabook’s launch, the
company’s engineers, along with the research laboratories of rival Japanese
computer manufacturers, were searching for the key to colour LCDs – very
important in a country where nine out of ten computers have colour screens.
Technically it is possible to make colour screens with passive-matrix LCDs –
simply lay out the elements in groups of three, covered by red, green and
blue filters. But there are limitations: as the number of pixels increases
there is more interference so the image quality and viewing angle decreases,
and response times degrade because it takes more time to address all the
pixels. This means showing moving pictures or following a computer cursor is
impractical.

By the end of the 1980s these limitations were causing computer
manufacturers concern. Computer programs with graphical user interfaces were
on the market, and the whole industry was buzzing with the idea of
multimedia applications that combine text, graphics, sound and video. It
wasn’t long before some research laboratories began to claim victory in the
search for a better colour LCD technology. The answer was the active-matrix
LCD, in which each element in the display connects to a transistor that
turns the element off and on. This increases the speed and isolates the
liquid crystal from interference, greatly improving the quality and contrast
of the image.

Active research

Most of the portable computers with colour screens on the market today have
active-matrix LCD displays. However, this does not mean the race to find a
replacement for CRTs is over. At present only Japanese companies are capable
of mass-producing active-matrix displays, a fact that has proved politically
sensitive. When the US tried to protect its electronics industry by taxing
imported active-matrix LCD displays, the move backfired because US computer
manufacturers found themselves competing with foreign companies that were
free to import their laptops, complete with active-matrix LCDs. Some US
companies even moved their manufacturing operations to other countries to
take advantage of this loophole. European countries also tried to protect
their flat screen manufacturers by denying international companies access to
European research.

There are, however, some signs that these tensions are easing. The American
tax was lifted at the end of June, and the European Commission has agreed to
give local subsidiaries of Japanese companies access to European research.
But even in this warmer political climate there is no guarantee that
active-matrix LCDs will be the automatic replacement for CRTs. No one has
yet found a way of mass-producing LCDs larger than about 40 centimetres in
diagonal, and even at this size the costs are terrifying. Building an
active-matrix display involves over 100 steps and takes more than four weeks
in a process that is similar to making dynamic random access memory chips
(DRAMs). However, forming the four million components of an active matrix
display is a far trickier proposition than etching the rows of a 4-megabit
(the most complex DRAMS in mass production) memory chip on wafers of crystal
silicon. A single dust particle in the wrong place can wreck a whole
display. Manufacturers insure against this by having what they call Class
100 clean rooms, where fewer than 100 dust particles per cubic foot of air
are permitted. But even with such precautions analysts say that factories
reject more than half the LCD panels they make because they have flaws.

If these problems seem formidable now, they become far worse with every
increase in the size of the screen. Hidehiko Catoh, chief manager of NEC’s
colour LCD division, explains that the main limit on producing bigger
screens is the size of the glass base, which must be extremely flat. Today,
companies such as NEC can produce displays which are 33 cm in diagonal, and
some claim to have prototype flat screens which measure 43 cm in diagonal.
Catoh predicts that 50 cm displays should be possible ‘but not easy’ within
five years, and he estimates that with current technology it may be possible
to build active-matrix displays which measure 75 cm.

Reaching the limits

By then, however, manufacturers will have run into another problem – the
speed at which individual pixels can switch on and off. The circuits on
today’s active-matrix screens are etched onto a thin film of amorphous
(non-crystalline) silicon, deposited at temperatures below 400 degreeC,
thereby reducing the risk of damaging the glass substrate. The downside of
this technique is that if screens get much bigger, the slow response time of
amorphous silicon is likely to become a handicap. This has led some
companies to look at alternative technologies.

The annual conference of the Society for Information Display, held in
Seattle in May, provided a showcase for some of these. Sharp, which
pioneered the manufacture of large active-matrix screens, described a
display measuring 63 cm in diagonal which was made using the more responsive
polycrystalline silicon. The Canadian branch of the American company Litton
Systems suggested cadmium selenide as a faster switching alternative to
amorphous silicon. And Tektronix, which is based in Oregon, provided details
of a 41 cm full-colour, active-matrix display in which the pixels are
controlled by plasma switches rather than transistors. Resolution is not as
high as with some transistor-controlled displays, but Tektronix believes it
can achieve the 1280 by 1024 resolution needed for workstation screens, and
some companies believe plasma-switched displays may eventually prove
suitable for HDTV.

But even with these enhancements, not all display manufacturers support
active matrix-display technology. The American companies In Focus and
Motorola, for example, which have formed a joint venture company called
Motif, prefer the older passive-matrix LCD approach. Motif claims that a
technique confusingly called active addressing can solve the response
problems of these passive displays. In actively addressed displays circuits
send a stream of pulses to turn picture elements off and on, instead of the
one pulse per scan used with conventional passive-matrix displays. The
result, according to Motif, is a display with a performance similar to
active-matrix LCDs, and a price comparable to passive-matrix technology,
although initially Motif’s displays will be much smaller – around 13 cm in
diagonal – than most active-matrix products.

Not all engineers are convinced by the Motif approach, and some teams
working on rival passive-matrix technology describe active addressing as a
‘dead end’. ‘It’s a short-term fix,’ says Ashok Vaidya, commercial manager
of CRL, part of the British Thorn-EMI group. ‘It doesn’t open up the
technology, it doesn’t allow larger displays and it doesn’t improve the
°ù±ð²õ´Ç±ô³Ü³Ù¾±´Ç²Ô.’

Vaidya is promoting a rival way of improving passive-matrix screens, the
ferroelectric display. In a ferroelectric screen, liquid crystals are
aligned so they rotate from one position to another when a voltage is
applied – and then stick in that position until they receive another signal.
This gives them the useful property of bistability; they do not need
continuous power to maintain an image. It is also possible to build up
screens of thousands of picture elements in a similar way to passive –
matrix LCDs, with a result more like an active-matrix display. Another
advantage of ferroelectrics, says Vaidya, is a very wide viewing angle and
no need for the complex matrix of transistors required in active-matrix
displays.

Fragile crystal

There are some drawbacks, however: today’s ferroelectric displays do not
produce as wide a range of grey tones as most active-matrix LCDs, which
limits the technology’s use in colour video displays. In addition,
ferroelectric crystals have traditionally been too fragile to manufacture
into screens, although CRL and Canon are claiming to have cracked the
manufacturing problems. An unnamed Japanese manufacturer is producing CRL’s
ferroelectric screen designs under licence, while Canon is investing nearly
$90 million in plant.

Both CRL and Canon are confident that ferroelectric technology has an
important part to play in the flat-screen market. In the immediate future
the largest demand is expected from the computer industry, where
ferroelectric displays are destined for monochrome laptops and the new
personal digital assistants – hand-held computers with handwriting
recognition and in-built fax and modem links. Ferroelectric technology may
also prove to be one of the keys to large screens of very high resolution.

Finding a flat-screen technology that can be scaled to the size required for
HDTV is proving quite a challenge. Many observers think projection systems
will be used for large HDTV displays, but projection systems have
limitations: they are dim and best used in darkened rooms. Brighter
light-emitting displays would allow people to watch TV in daylight. At
present the largest full-colour, flat-panel displays that emit light are the
plasma-discharge type, which make up images from a matrix of tiny gas-filled
tubes. A voltage applied across the gas accelerates electrons, which excite
atoms, creating an ionised plasma that emits light as the atoms drop to
their normal energy levels. In such displays signals are transmitted along
both rows and columns. To turn on the cells, a signal from each direction is
needed, and the response will then be rapid. This allows fast displays to be
made, although they do tend to be rather power-hungry, often requiring about
150 volts to work because the conversion of light into electricity is not
very efficient.

One company, Photonics Imaging of Ohio, has just delivered a 75 cm
full-colour plasma display with a resolution of 1024 by 768 to the US
Department of Defense, as part of a demonstration project. There were also a
number of other examples of large flat screens at the Society of Information
Display meeting in May, including a colour plasma display measuring a metre
diagonally designed by a team from NHK Laboratories and Dai Nippon Printing.
And Fujitsu announced it will begin selling prototype plasma displays for
standard American televisions next year at a cost of $8000.

Peter Friedman, president of Photonics Imaging, believes that Plasma screens
have won the battle to replace CRTs for large screen television. There are
also signs that light-emitting displays may have a place in the
smaller-screen market. The first full-colour display to use
electroluminescent technology, built by Planar Systems of Oregon, attracted
attention in Seattle. The crucial development was a phosphor which emits
blue light. Monochrome electroluminescent displays, which emit yellow light
when electrons excite a phosphor in a thin film, have been on the market for
a decade, but colour had been elusive. Filters could extract red and green
light from the yellow phosphor, but the displays still looked yellowish.
Adding the blue phosphor makes the display more complex, with two thin-film
substrates, but greatly improves the colour.

The availability of such a wide variety of flat screen technologies means
the field is still wide open in the race to replace CRTs. Most engineers
agree that different types of flat screen will fill different niches in the
market. After all, the performance requirements for an aircraft cockpit,
electronic book and wide-screen TV differ vastly.

But at the same time there is still a desire to dominate the two key
markets: portable computers and HDTV. The renamed US Advanced Research
Projects Agency (ARPA) (formerly the defence research body, DARPA) is
negotiating a $20 million one – year contract with a consortium of
companies to develop active-matrix displays, as well as looking at
alternative flat screen technologies. One company that will benefit is
Optical Imaging Systems of Michigan, which plans to spend $100
million on the first high volume LCD plant in the US. The European
Commission has also pledged support for European companies working on HDTV
technologies such as flat-panel displays, while Japan’s NHK is pushing its
compatriots to produce the large screens that will make HDTV a reality.

With such big names in the running, and the stakes rising by the minute, as
market awareness of technologies such as HDTV, electronic notepads and
virtual reality increases, there is little doubt that there is still plenty
of drama to come in the race to produce the ‘big’ flat picture.

Michael Cross is a freelance journalist.