`
`oz» Key defining characteristic
`
`+ Touch capability is provided by a display manufacturer
`instead of a touch-module manufacturer
`
`0 Touch-module manufacturers can't do in-cell or on-cell
`
`«to Marketing Terminology Alert!
`
`+ Some display manufacturers call all their embedded touch “in—cel|”,
`even though they may be supplying hybrid or on-cell
`
`+ Some display manufacturers use a brand name to encompass all
`their embedded touch products
`
`0 For example, “Touch On Display” from lnnolux
`
`+ Some display manufacturers direct-bond or air-bond an external
`touchscreen to their display and call it “out-ce|l”
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Embedded Touch Terminology...2
`
`Touch sensor is physically inside the LCD cell
`Touch sensor can be:
`
`0 Capacitive electrodes (same as p-cap)
`
`o Light-sensing elements (rare)
`
`On-Cell Touch sensor is on top of the color-filter glass
`(LCD) or the encapsulation glass (OLED)
`
`o Capacitive electrodes (same as p-cap)
`
`Touch sensor has sense electrodes on top of the
`Hybrid
`(In-Celll color-filter glass 1 drive electrodes inside
`On-Cell) the cell
`
`0 IPS LCD: Segmented Vcom electrodes on
`the TFT glass
`
`0 Non-IPS LCD: Segmented Vcom electrodes
`on the underside of the color filter glass
`
`SID DISPLAY WEEK ‘14
`
`
`
`Early Embedded Methods All Failed
`
`oz» Attempts to develop embedded touch in 2003-2011
`
`were all trying to invent something new while
`leveraging the LCD design
`
`+ “Pressed” capacitive, first mass-produced by Samsung in 2009
`
`4 Light-sensing, first mass-produced by Sharp in 2009
`
`4 Voltage-sensing (“digital switching”), first mass-produced by
`Samsung
`
`«to But none of them was really successful
`
`+ Insufficient signal-to-noise ratio for robust operation
`
`+ The need to press the display surface, which prevented the
`use of a protective cover-glass
`
`+ The unreliability of pressing the display very close to the frame,
`where the color-filter glass has little ability to move
`
`SID DISPLAY WEEK ‘l4
`
`
`
`First Successful Embedded Touch:
`
`OLED On-Cell P-Cap
`
`«:0 Samsung S8500 Wave mobile
`phone with Super AMOLED on-cell
`p-cap touch (Feb. 2010)
`4 3.3-inch 800x480 (283 ppi) AM-OLED
`4 “Super AMOLED” is Samsung’s
`
`(odd) branding for on-cell touch
`4 Sunlight readable
`0 AR coating & no touchscreen overlay
`
`On Ce
`(5“Pe'.' AM
`- r ~
`:
`
`Window=
`direct-bond
`
`cover—glass
`
`_
`.....s.,...;
`2
`3
`'i”°'§@'
`
`.
`
`Mobile World Congress 2010
`
`Source: Samsung
`
`SID DISPLAY WEEK ‘l4
`
`
`
`On-Cell P-Cap
`
`Cover Glass ("Lens")
`Decoration
`
`Top Polarizer
`
`Metal Bridges
`1 Insulator
`
`Sense 8. Drive Electrodes (ITO)
`Color Filter Glass
`
`Color Filter
`
`Liquid Crystal
`TFTs
`
`TFT Array Glass
`
`«:0 Principle
`
`Source: The author
`
`§ ITO P-cap electrode array is deposited on top of the color filter
`glass (under the top polarizer)
`
`o Exactly the same function as discrete (standalone) p-cap
`
`o Shown above is one ITO layer with bridges; it could also be
`two layers with a dielectric instead
`
`SID DISPLAY WEEK ‘l4
`
`114
`
`
`
`The Display-Makers
`Quickly Got the Idea
`
`°:~ Don’t try to invent something new; figure out
`how to apply what already works (p-cap)!
`
`otv The result: Sony’s (JDI) “Pixel Eyes” hybrid
`in-celllon-cell mutual capacitive
`0 First successful high-volume embedded touch in LCD
`
`Segmented anti-static shield
`on color-filter glass
`(sense electrodes)
`
`VCOM electrodes ’
`on TFTg|ass
`(drive electrodes)
`
`'
`
`"
`
`M
`
`‘
`
`SID DISPLAY WEEK ‘14
`
`115
`
`Source: Japan Display; annotation by the author
`
`
`
`First Phones with Hybrid ln-Ce||/
`On-Cell Mutual-Capacitive (May 2012)
`
`«to Sony Xperia P and HTC EVO Design 4G (not the iPhone 5)
`
`oz» Similar LCDs
`
`# 4-inch 960x540
`
`LTPS (275 ppi) with
`different pixel arrays
`
`~:~ Same touch solution
`
`4 Synaptics
`C|earPad 3250
`
`(four touches)
`
`«to <100 um thinner than
`
`one-glass solution!
`
`Source: Sony
`
`Source: HTC
`
`SID DISPLAY WEEK ‘14
`
`intel‘
`
`
`
`Apple iPhone 5: First Fully In-Cell
`Mutual Capacitive (Sept. 2012)
`
`03° Structure
`
`+ Both sense and drive electrodes are in the TFT array, created by
`switching existing traces so they become multi-functional
`
`4 Apple has said they may change
`to Innolux “Touch On Display”
`(TOD, |nnolux’s brand name for ALL
`of their embedded touch structures)
`in iPhone 6
`
`0 That doesn't actually tell us
`anything, since TOD includes
`
`all three embedded structures...
`
`SID DISPLAY WEEK ‘l4
`
`Source: CNET
`
`
`
`Apple’s iPhone-5 Electrode Structure
`
`a
`
`éé
`
`Suhpnalc =
`
`1
`
`tr;tHw».& -~
`
`ULU __UUU
`
`‘>
`
`I UUU II ULM
`
`Source: BOE Technology Group's Central Research Institute
`
`SID DISPLAY WEEK ‘14
`
`
`
`Other In-Cell Electrode Structures
`
`(Based On Patents)
`
`«:0 Apple & Samsung
`
`+ Drive electrodes are segmented VCOM
`
`+ Sense electrodes are metal overlaid on the CF black matrix
`
`«to Apple & Samsung
`
`+ Drive electrodes are ITO stripes deposited on top of a dielectric
`layer over the color filter material
`
`+ Sense electrodes as above
`
`~:~ Sharp
`
`+ Both drive & sense electrodes are deposited on the bare CF-glass,
`before the black matrix and color-filter material are applied
`
`«to LG Displays
`
`4 Self-capacitive method using just segmented VCOM
`
`SID DISPLAY WEEK ‘14
`
`
`
`Summary of Sensor Locations
`
`Discrete sensor
`(separate glass)
`
`To of cover— lass
`
`Bottom of cover—glass
`(OGS = G2)
`
`lndustry standard
`Glass or PET
`Easy to add shield layer
`Display unconstrained
`
`Good for sensing
`Widest sensing area
`Disla unconstrained
`
`Thickness & weight
`
`lmractical
`
`Complex lens (yield)
`Limited durability
`
`To of olarizer
`
`lmractical
`
`(1 or 2 layers)
`
`Lower cost (1 layer)
`
`Limited to display size
`
`Both sides of CF glass
`(hybrid for non-IPS)
`
`Slightly thinner
`Slightly lower cost
`
`Top of CF glass and
`in TFT array
`h brid for IPS
`
`Highest performance
`Slightly thinner
`Sli htl
`lower cost
`
`2-sided CF process
`Limited to display size
`Re uires disla interation
`
`2-sided CF process
`Limited to display size
`Re uires disla interation
`
`In cell (on TFT array
`for IPS; split between
`TFT and CF for non—lPS
`
`High performance
`Thinnest
`Potentiall
`
`lowest cost
`
`Limited to display size
`Requires display integration
`Com lex desi n
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Integrating the Touch Controller
`and the Display Driver lC...1
`
`Host Processor
`
`and Display
`Accelerator
`
`SID DISPLAY WEEK ‘14
`
`
`
`Integrating the Touch Controller
`and the Display Driver lC...2
`
`«:0 Advantages
`
`+ Full synchronization of touch and DDI
`
`+ Can work with any sensor (discrete, OGS, on-cell, in-cell, hybrid)
`
`+ Reduced latency
`
`0 70 ms to 20 ms
`
`4 Capable of user-input and feedback without CPU involvement
`
`0 Done by programming the display configuration blocks of flash memory
`
`a Overlay capability plus image fade-in/out, animation, translation, etc.
`
`+ Can support wake-on-touch
`
`o Can display sprites or graphics for log-in screen
`
`»:« Disadvantages
`
`# Design is LCD-specific (resolution & pixel layout)
`
`+ Substantial NRE; appropriate only for high-volume
`
`SID DISPLAY WEEK ‘14
`
`
`
`Comparison of Discrete (e.g., OGS)
`Touch with Embedded Touch...1
`
`«to Cost: is embedded touch really “free”? N0!
`
`+ Barrier to entry
`
`c There is much more intellectual property (IP) on embedded touch
`layer-structure & driving; making sure you don’t infringe costs money
`
`+ Development cost
`
`a Embedded touch is much more complex to develop than OGS
`
`a High volume is required (5M) to make it practical
`
`§ Cover glass, decoration & bonding
`
`0 Similar to discrete (OGS), but embedded cover-glass is just
`glass & decoration (no ITO), so it's easier to manufacture
`
`a Sheet-type OGS may not be as strong as plain cover-glass
`
`+ Touch controller
`
`0 No integration = same cost (but performance is poor)
`
`a Linked to TCON for timing control = same cost (slightly different chip)
`
`o Integrated with TCON = saves $1-$2 in material cost
`
`SID DISPLAY WEEK '14
`
`inte|'
`
`
`
`Comparison of Discrete (e.g., OGS)
`Touch with Embedded Touch...2
`
`°:~ Cost (continued)
`
`+ FPC to connect electrodes
`
`0 On-cell and hybrid = same
`
`0 In-cell = none if touch controller is COG; saves another $1-$2
`
`4 Electrode material
`
`0 Discrete OGS currently uses ITO; could move to printed metal-mesh,
`which could save $10+ in tablet size (once sensor competition gets real)
`
`0 On-cell = same as discrete ITO
`
`0 Hybrid = only half as much added ITO (little material cost-difference)
`
`o In-cell = no added ITO
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Comparison of Discrete (e.g., OGS)
`Touch with Embedded Touch...3
`
`»:~ Performance
`
`+ On-cell = same as discrete or worse
`
`0 If you build the color-filter first (focus on LCD yield) then
`you can’t use high-temperature ITO so touch performance is worse
`
`0 If you build the touch electrodes first for good performance, then
`
`you can’t thin the color-filter glass
`
`+ Hybrid = same
`
`+ In-cell = worse, but should improve to be same as SNR goes up
`
`oz» Thickness
`
`0 Embedded is typically 100 um thinner than discrete OGS
`
`+ But the thickness variation between smartphone models with
`embedded touch is ~1.0 mm due to other features, so 0.1 mm
`
`doesn't mean that much to the consumer (it’s mostly marketing!)
`
`SID DISPLAY WEEK ‘14
`
`intel"
`
`
`
`Comparison of Discrete (e.g., OGS)
`Touch with Embedded Touch...4
`
`«:0 Weight
`
`+ Embedded = discrete (same number of sheets of glass)
`
`~:~ Power consumption
`
`+ On-cell & hybrid = same as discrete
`
`+ In-cell with integrated touch & TCON = probably lower, but touch
`power consumption is much lower than LCD power—consumption,
`so the decrease isn’t very significant
`
`«:0 Off-screen icons
`
`+ Discrete = no problem
`
`+ Embedded = eguires additional circuitry
`
`SID DISPLAY WEEK ‘14
`
`
`
`Embedded Touch Conc|usions...1
`
`oz» Embedded touch isn’t a clear win in either cost or
`
`technology; it’s all about who gets the touch revenue!
`
`oz» The driving force in embedded touch is the display-
`makers’ need to add value in order to increase their
`
`profitability
`
`°:~ Embedded touch provides little advantage to the
`end-user (consumer)
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Embedded Touch Conc|usions...2
`
`0:0 It’s not clear that embedded touch will offer significant
`
`cost-savings to the device OEM, since OGS can be
`further cost-reduced with ITO-replacement materials
`
`0:0 The display-makers will take some market share
`
`with embedded touch in high-volume products
`
`(Displaysearch says 25% in 2018) but embedded
`touch is unlikely to become dominant because the
`touch-panel makers won’t let their business be
`destroyed
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Large-Format P-Cap
`
`03° Introduction
`
`03° ITO Electrodes
`
`03° Wire Electrodes
`
`~'° Metal Mesh Electrodes
`
`6. Applications
`
`SID DISPLAY WEEK ‘14
`
`
`
`Introduction
`
`oz» Large-format touch is a much more wide-open space
`than consumer-electronics touch
`
`§ Multi-touch infrared (IR) has replaced traditional (single-touch) IR
`
`+ Camera-based optical has dropped substantially with the
`exit of NextWindow (SMART Technologies) from the market
`
`+ Startup: Sentons is taking a new approach to bending-wave
`
`+ Startup: RAPT is taking a new approach to in-glass optical
`
`4 P-cap with metal mesh is a threat to all other large-format
`touch technologies
`
`o Commonality of user experience (UX) with the 3 billion p-cap units
`shipped since 2007 may be the driving force
`
`0 Cost and complexity (as always) are the impediment
`
`SID DISPLAY WEEK ‘l4
`
`
`
`ITO Electrodes
`
`~30 3M has managed to get ITO electrodes to work
`
`in a 46-inch display (larger than any other with ITO)
`
`+ They won't disclose their secret sauce
`T V
`
`SID DISPLAY WEEK ‘l4
`
`Source: Photo by Author
`
`
`
`Wire Electrodes...1
`
`03 One more sensor variation: 10-micron wires
`
`between two sheets of PET or glass
`
`+ Commonly used for large-format touchscreens
`
`+ Two main suppliers: Visual Planet & Zytronic, both in the UK
`
`9 floor-to-ceiling
`
`Visual Planet
`
`touchscreens in
`
`the University of
`Oregon Alumni
`
`SID D|SPLAY WEEK ‘l4
`
`132
`
`ifltel
`
`Source: The University of Oregon
`
`
`
`Wire Electrodes...2
`
`03° Zytronic’s new multi-touch large-format p-cap
`
`+ Previous Zytronic products were self-capacitive (2-touch max)
`
`o Binstead’s frequency-variation patent was the basis of sensing
`
`+ New product is mutual-capacitive with very dense electrode pattern
`
`a Traditional measurement of capacitance reduction caused by finger
`
`o ~1.5 mm electrode spacing in 6 mm x 6 mm cell
`
`> Density reduces visibility because the human visual system sees
`a more uniform contrast
`
`o 10-micron insulated copper wires allow crossover (“single layer”)
`
`>100’s Q/m at 10 pm
`
`0 Can be applied to glass or film (including curved surfaces)
`
`0 Initial controller handles all sizes up to 72”; 100”+ possible
`
`a Minimum 10 touches with palm rejection
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Wire E|ectrodes...3
`
`«:« Jeff Han from Perceptive Pixel (acquired by Microsoft
`
`in mid-2012) showed an 82” at CES 2012 (with active
`stylus) and a 72” at Digital Signage Expo (DSE) 2012
`
`§ Metal electrodes (not ITO) — although Jeff wou|dn’t talk about the
`electrode material or who is manufacturing the touchscreens
`
`Source: Photos by Author
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Wire E|ectrodes...4
`
`0:0 Both the 72” & 82” look much better than the
`
`traditional Zytronic zig-zag 10-micron wire pattern
`
`72" at DSE 2012
`
`Source: Photos by Author
`
`— oN U
`N‘
`
`J
`U)
`
`D '
`
`56
`
`in
`N
`
`SID D|SPLAY WEEK ‘14
`
`
`
`Metal-Mesh Electrodes
`
`03° “lnvisible” metal-mesh electrodes are the biggest
`threat & opportunity in large-format p-cap
`
`+ Many suppliers are working on this
`
`+ Few (if any) have made formal product announcements
`
`4 Display sizes of 42” to 55” are frequently mentioned
`
`0 There are significant challenges
`
`0 Total number of connections is large (~250 + ~150 = 400 for 55”)
`
`0 Multiple ganged controllers are required
`
`0 Longer electrodes means slower sensing (larger RC time-constant)
`
`o Much larger number of electrodes takes longer to sense
`
`0 Number of suppliers able to print on 1,200 mm web is limited
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Applications...1
`
`oz» Large-format multi-touch applications
`
`SID DISPLAY WEEK ‘l4
`
`Source:
`
`Zytronic
`
`
`
`App|ications...2
`
`~30 Applications for curved large-format touchscreens
`
`Kiosks
`Public Information
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Applications...3
`
`«to BUT, stepping back from a technology focus, is the
`large-format touch market likely to start shrinking?
`
`+ Interactive media walls — touch is very necessary
`
`a Mu|tiTaction makes the best vision-based touch today
`(author’s opinion)
`
`0 Point-of-information — touch still seems necessary
`
`+ Digital signage - interaction via smartphone
`
`+ Education — interaction via tablets (including multi-user!)
`
`4 TV — interaction via mobile & motion-based devices
`
`+ Horizontal home-gaming tables — will they ever exist?
`
`+ Other large-format app|ications??
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Stylus Technologies
`
`0:» History
`
`oto Use Cases
`
`»:~ Passive Stylus
`
`oto Electromagnetic Resonance (EMR) Stylus
`
`0:» Active P-Cap Stylus
`
`03° Prediction
`
`«to Other Active Stylus Technologies
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Stylus History...1
`
`«to Microsoft Tablet PCs, PDAs, and early smartphones
`(e.g., Trio) always had styli (1989 to 2007), so why are
`we so finger-focused now?
`
`0 Steve Jobs and the iPhone in 2007 — “Who needs a stylus?”
`
`9 Microsoft’s failure to make the stylus-based Tablet PC a success
`with consumers caused them to de-emphasize the stylus and
`focus on finger-touch in Windows 7; that has continued and
`become even stronger in Windows 8
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Stylus History...2
`
`03° Is the stylus coming back into the consumer space?
`
`YES!
`
`+ All the major p-cap controller suppliers support active & passive
`
`4 PC OEMs want to differentiate their products from Apple’s
`
`+ Legacy Windows software on a Win8 tablet needs a stylus
`
`+ Android (in Ice Cream Sandwich) supports stylus messages
`
`+ Samsung has shipped >15M Galaxy Notes in two sizes
`
`+ Consumption isn't enough; a stylus is great for creation
`
`Source: Almel
`
`SID DISPLAY WEEK ‘14
`
`
`
`Stylus Use-Cases...1
`
`oz» Taking notes (in both Windows and Android)
`# Notes are automatically converted into text in background; being
`able to search your “ink” notes is very powerful
`
`oz» Annotating documents
`4 Typically Office or PDF
`
`«:0 Quick sketches
`
`0 Typical whiteboard-type sketches
`
`oto Precision pointing device, e.g. with Windows 8 Desktop
`+ When you’re trying to select tiny Ul elements
`
`«to Artistic drawings
`+ |t’s unbelievable what a real artist can do...
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Stylus Use Cases...2
`
`t
`.am-m-«
`23232211?
`lfluouszotnau mum
`3”“-M
`3 oaolnn mun Lynn
`_3oms2naa-ugmn
`
`_
`
`‘
`
`'
`
`%
`
`,
`V;
`I
`t
`“
`._ _
`‘I ,
`
`..
`
`;;
`'
`
`‘
`
`3
`
`--
`
`Created with
`
`an N-Trig active
`stylus on a
`Fujitsu Lifebook
`using ArtRage
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Passive Stylus...1
`
`03° A passive stylus can be any conductive object
`
`+ Metal rod
`
`+ Conductive plastic
`
`+ Ballpoint pen
`
`+ #2 pencil (shown at CES 2014)
`
`0 Long fingernail
`
`4 And those horrible 7 mm conductive-rubber-tipped styli
`
`o Needed for backwards compatibility with early tablets with low SNR
`
`~:« Tip diameter
`
`+ State of the art is 1.5 to 2.0 mm
`
`0 Next generation is 1.0 mm
`
`+ Essentially every controller supplier supports this now
`but not many have made it out into shipping products yet
`
`SID DISPLAY WEEK ‘14
`
`
`
`Passive Stylus...2
`
`«:0 Advantages
`
`+ Extremely low cost
`
`+ Easily replaceable
`
`+ Can be made any size and comfort level by low-tech methods
`
`+ Improves as SNR increases
`
`«to Disadvantages
`
`# No hover that meets Microsoft’s specification
`
`+ There's no OS support (yet) for differentiating between
`finger & stylus
`
`+ No pressure-sensing, so art and handwriting aren't as good
`
`+ Resolution can't be better than a finger
`
`SID DISPLAY WEEK ‘14
`
`
`
`Electromagnetic Resonance
`(EMR) Stylus...1
`
`«:0 Key characteristics
`
`+ Second sensor under the LCD
`
`4 Batteryless electronic stylus
`
`Cover glass
`
`Sensor
`
`Acer TM100
`
`/(The first
`
`Microsoft
`
`Tablet PC
`
`convertible)
`
`SID DISPLAY WEEK ‘l4
`
`147
`
`ifltel
`
`Source: Wacom
`
`
`
`EMR Stylus...2
`
`Penabled
`
`UJBCOITT
`
`Pressure-sensitive
`
`capacitor (Cnp)
`
`Cordless pen
`
`without battery
`
`Pen equivalent circuit
`
`Sensor grid Schematic
`
`Transmitted RF_.-" "-._Received RF
`
`Many wires
`
`Serial/USB
`
`interface
`
`to host
`
`SID DISPLAY WEEK ‘l4
`
`III—II_
`III—III
`I
`III—III
`II
`
`
`
`EMR Stylus...3
`
`«:« Variations
`
`+ Sensor substrate (rigid FR4 vs. flexible 0.3 - 0.6 mm PET)
`+ Pen diameter (3.5 mm “PDA pen” to 14 mm “executive” pen)
`
`0:» Size range
`+ 2” to 14"
`
`«to Controllers
`
`4 Proprietary
`
`Controller for 10.4”
`'3’ Advantages
`S°”'°"3"""°°'"
`A
`+ Very high resolution (1,000 dpi)
`4 Pen “hover” (mouseover = move cursor without clicking)
`+ Sensor is behind LCD = high durability & no optical degradation
`
`+ Batteryless, pressure-sensitive pen
`
`__
`r§3§’.i‘°$§"n‘S.
`
`T
`
`T
`
`H
`
`T
`
`T
`
`T
`
`Single controller can
`run both pen digitizer
`& p-cap finger touch
`
`SID DISPLAY WEEK ‘l4
`
`
`
`EMR Stylus...4
`
`oz» Disadvantages
`# Electronic pen = disables product if lost; relatively expensive
`+ Difficult integration requires lots of shielding in mobile computer
`4 Sensor can’t be integrated with some LCDs
`+ Single-source for mobile CE devices (Wacom) = relatively high cost
`
`«to Applications
`+ Phablets and tablets
`
`# E-book readers
`
`0 Opaque desktop graphics tablets
`+ Integrated tablet (pen) monitors
`
`«to Suppliers
`
`+ Wacom, Hanvon, Waltop,
`UC-Logic/Sunrex, KYE
`
`SID DISPLAY WEEK ‘l4
`
`
`
`EMR Stylus...5
`
`do Samsung Galaxy Note sketching demo at CES 2012
`
`The Galaxy
`Notes use
`
`both a
`
`P‘°aP
`
`touchscreen
`
`AND a
`
`Wacom
`
`EMR
`
`stylus
`(2 sensors!)
`
`SID DISPLAY WEEK ‘l4
`
`Source: Photos by Author
`
`
`
`Active P-Cap StyIus...1
`
`3) Stylus LF transmission
`containing:
`- TIP Pressure level
`- Buttons status
`
`2) Every cycle:
`- Atrigs drivelreceive analog signal
`to/from the ITO grid lines
`Finger touching the sensor affect
`signal level, allowing for touch
`detection
`
`-
`
`Atrigs also pick-up Slylus signal
`and determine its location based
`
`on signal level distribution across
`the ITO lines
`
`Source: N—Tn'g
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Active P-Cap Sty|us...2
`
`03- Variations
`
`+ One-way digital RF transmission from stylus to p-cap sensor,
`with both sense & drive electrodes acting as antennas
`o N-Trig has by far the most-developed user experience
`
`4 Two-way transmission between stylus and p-cap sensor
`o Stylus receives p-cap sensor drive-signal, amplifies it, adds digitally
`
`encoded stylus information, and transmits it back to sensor
`0 Atmel was the first to put this into production, but their user
`experience is still very immature
`
`+ Stylus generates intense e-field at tip
`a E-field Q capacitance to p-cap sensor operating as usual
`(finger subtracts capacitance)
`
`o Unclear if anyone is actually doing this...
`
`SID DISPLAY WEEK ‘14
`
`
`
`Active P-Cap Sty|us...3
`
`»:~ Advantages
`+ Uses existing (single) p-cap sensor
`4 Pen “hover” (mouseover = move cursor without clicking)
`+ Stylus tip can be very small (< 1 mm)
`+ High resolution and accuracy
`
`«to Disadvantages
`+ Stylus requires power source (battery or super-capacitor),
`which requires charging contacts in stylus-garage and
`charging circuit in host computer
`+ Stylus technology is unique to each p-cap controller supplier
`c Total lack of interoperability will probably prevent active stylus
`from ever becoming mainstream
`
`o OEMs’ desire to obtain high margin on accessories makes the
`problem even worse
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Active vs. Passive Stylus Summary
`
`«z» This battle’s been going on since the 1990s...
`
`versus
`
`+ Very |ow—cost
`
`+ More expensive
`
`+ “Good enough”
`+ Improves as SNR increases
`
`+ #2 pencil is the gold standard
`+ “Artificial finger” in Windows
`
`+ More flexibility in Android
`
`+ Pressure-sensing
`4 Hover (required for Windows)
`
`4 Higher resolution
`+ Customizable features
`
`P—cap (powered)
`
`EMR (batteryless)
`
`+ N-Trig leads
`+ Wacom leads
`+ Others following
`+ Others insignificant
`+ N0 interoperability
`+ 2"“ sensor
`as
`Cost in high-volume is surprisingly close
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Prediction
`
`03° Passive stylus is going to win (become mainstream)
`
`+ Being “good enough” is very important in the touch industry!
`
`+ |t’s the lowest-cost solution
`
`+ However...
`
`0 There is still some chicken-and-egg regarding good support
`for stylus in application software
`
`a Some OEMs haven't bought into the need for a stylus yet
`(more chicken-and-egg)
`
`«to Active stylus will remain a niche
`
`+ Active stylus’ total lack of interoperability and very high
`price as a replacement accessory will prevent it from ever
`becoming mainstream
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Other Active-Stylus Technologies
`
`«to Combination ultrasonic & infrared
`
`# Used in many clip-on and clipboard-style digital note-taking
`accessories; also available for iPad
`
`0:. Embedded CMOS-camera stylus by Anoto
`+ Widely licensed for digital-pen note-taking accessories and
`form—fil|ing applications
`+ Used by LG Displays in large-format touch
`4 Used in Panasonic 4K 20" professional tablet shown at CES 2013
`
`«:0 Infrared LED light-pen
`+ Used by iDT| in their light-sensing in-cell touch monitor
`
`«:0 Visible laser-pointer
`4 Used by isiQiri in |arge—format touch
`+ Also works with iDT| light-sensing in-cell touch
`
`SID DISPLAY WEEK ‘l4
`
`
`
`92° Multi-Touch
`
`«:0 OS Application-Development Support
`
`«to Middleware
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Multi-Touch
`
`»:~ Multi-touch is defined as the ability to recognize
`two or more simultaneous touch points
`
`«to Multi-touch was invented in 1982 at the
`
`University of Toronto (not by Apple in 2007!)
`
`oz» “Pinching” gestures were first defined in 1983
`(not by Apple in 2007!)
`
`oz» Windows 7 (2009) & Windows 8 (2012) both support
`multi-touch throughout the OS and are architected to
`support an “unlimited” number (~100) of simultaneous
`touch points
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Multi-Touch Architecture
`
`’ Capable of decoding multiple
`
`streams of moving points and
`taking actions in response
`
`a
`
`Capable of delivering multiple
`
`streams of moving points (and
`acting on a defined subset of them)
`
`Capable of delivering sets of
`simultaneous points to the OS
`
`Capable of sensing multiple
`simultaneous points
`
`Source: The author
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Why Multi-Touch Has
`Become So lmportant...1
`
`«:0 Apple
`+ Apple established multi-touch as a “must-have” for coolness.
`The result is that people of all ages expect every display they
`see to be touchable with multiple fingers
`
`«:0 Gaming
`0 Gaming is a natural for multi-touch. Try playing air hockey
`without multi-touch. ..
`
`«to Unintended touches
`
`0 One of the major values of multi-touch is to allow the system
`to ignore unintended touches (“palm rejection”, “grip suppression”,
`etc.). As desktop screens become more horizontal (recline)
`this will become even more important.
`
`SID DISPLAY WEEK ‘14
`
`
`
`Why Multi-Touch Has
`Become So lmportant...2
`
`°:~ Multi-user collaboration
`
`+ When two people want to collaborate on a large screen (e.g.,
`a student and teacher on an interactive “whiteboard” LCD),
`multi-touch is essential
`
`0 Identifying which touch belongs to which user is still unsolved
`
`o It IS currently possible to uniquely identify multiple simultaneous styli
`
`SID DISPLAY WEEK ‘l4
`
`
`
`How Many Touches Are Enough?...1
`
`oz» The industry has multiple answers
`0 Microsoft settled for 5 touches in Win8 (they wanted 10)
`o But now under pressure from OEMs they have buckled and
`reduced it to TWO touches for All-in-One desktops (BIG mistake!)
`+ The p-cap touchscreen suppliers under 30” either say “10” or
`
`“as many as possible” (e.g., 3M’s p-cap supports 60+ touches)
`+ The large-format touchscreen suppliers say that 40 is enough
`
`~:~ In practice it depends on the hardware and
`
`controller firmware implementation
`
`0 Ideally the touchscreen should ignore all other touches beyond
`however many the product is guaranteeing
`
`+ This is usually called “palm rejection” and its implementation is
`
`absolutely critical to the user experience
`
`SID DISPLAY WEEK ‘l4
`
`
`
`How Many Touches Are Enough?...2
`
`»:~ The answer actually depends on the application
`
`+ For a small mobile device, 2-5 (one hand) are enough
`
`+ For a single-user app on El device (even an 82” screen),
`it’s hard to see why more than 10 (two hands) are needed
`
`+ For a multi-user app, it depends...
`o For a 55-inch gaming table, 40 (8 hands) is not unreasonable
`> The key touchscreen specification is probably response time (latency)
`
`o For a 65-inch interactive “whiteboard” LCD, 20 (4 hands) is
`probably enough, although an argument can be made for 40
`> BUT, the key touchscreen specifications are entirely different:
`minimum stylus tip size, pre-touch, jitter, ink-lag, etc., can all be critical
`
`From a video of a very
`cool muIti—pIayer game
`on the FIatFrog website
`
`Source: FIatFrog
`
`SID DISPLAY WEEK ‘l4
`
`
`
`#1 Reference On Multi-Touch
`
`»:~ “Mu|ti-Touch Systems that
`I Have Known and Loved”
`
`+ wvvw.bilIbuxton.com/mu|titouchOverview.html
`
`“If you can only manipulate one
`point
`you are restricted to the
`gestural vocabulary of a fruit fly.
`We were given multiple limbs
`for a reason. It is nice to be
`
`able to take advantage of them.”
`
`SID DISPLAY WEEK ‘l4
`
`Bill Buxton, 2008
`
`Principal Researcher,
`Microsoft Research
`
`
`
`For Windows, the “Logo”
`Is the Starting Point
`
`«:0 A set of touch performance standards designed
`to ensure a high-quality user experience
`
`+ 5 touch-point minimum
`+ Touchscreen jitter
`
`+ Extra input behavior
`+ High-resolution timestamp
`+ Input separation
`+ Noise suppression
`+ Physical input position
`+ Reporting rate
`+ Response latency
`4 Cold boot latency
`+ Touch resolution
`
`+ User experience
`4 Pre-touch
`
`+ Pen tests
`
`SID DISPLAY WEEK ‘l4
`
`Windows 8
`
`Compatible ]
`
`
`
`Windows 8 Touch
`
`I: Windows 8
`
`«to The Win8 Touch Logo specification is based on p-cap
`4 Win7 spec was based on optical, which had little relevance
`4 Win8 spec creates a common touch capability for mobile phones,
`tablets, notebooks, and desktops
`o This may be very significant for multi-platform applications!
`
`0:» Basic spec requirements
`4 Minimum of 5 simultaneous touches; must ignore an additional 5
`4 Tablets must be zero-bezel; otherwise 20 mm border minimum
`
`4 Respond to first touch in < 25 ms
`4 Subsequent touches must be < 15 ms at 100 Hz for all touches
`4 Better than 0.5 mm accuracy with < 2 mm offset from actual location
`4 No jitter when stationary; < 1 mm when moving 10 mm
`4 Pre-touch < 0.5 mm
`
`4 Finger separation >= 12 mm horizontal/vertical, 15 mm diagonal
`0 But on-screen keyboards and normal human behavior violates this!
`
`SID DISPLAY WEEK ‘l4
`
`intel
`
`
`
`Windows 8 Touch
`Application Development I=gWindows8
`
`«to There are multiple development environments
`commonly used in Windows 8, each of
`which handles touch differently
`
`+ Native C++ (Win32/COM)
`
`§ Managed environment (.NET Framework)
`
`§ Silverlight & WPF (Windows Presentation Foundation)
`
`+ Adobe Flash
`
`+ Modem (Win-8) using C# and XAML or HTML5 and JavaScript
`
`o Modern apps today only represent one aspect of business computing:
`
`reporting/dashboards, with moderate-to-light data updating
`
`0:0 From my perspective...
`
`+ As a hardware person, I find the level of detail required
`to do anything significant in touch software to be excruciating
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Android Touch
`Application Development
`
`"'
`
`«to Android has an extensive and growing API
`for touch & stylus
`
`+ I hear complaints about the degree of bugginess
`
`+ From what I can tell, the level of tediousness is a
`
`little better than Windows
`
`+ The Android API supports up to 256 touches, but the actual number
`depends on the hardware & firmware implementation
`in the device — 2 to 5 isn’t unusual
`
`+ Fragmentation of Android (different versions from each OEM)
`appears to make developing a robust run-on—anything Android
`touch application very difficult
`
`~:~ The language decision is easy — it’s Java or nothing
`
`SID DISPLAY WEEK ‘l4
`
`
`
`iOS Touch
`Application Development
`
`.
`IDS 7
`
`«:0 iOS seems to have the most constrained touch
`
`application development environment
`
`+ But it’s not any easier than Android -- in the chapter on touch in
`“Programming iOS 5" (an O’Rei||y book), the words “messy” and
`“tricky” seem to occur a lot
`
`~:~ The language decision is easy — it’s Objective-C
`or nothing
`
`SID DISPLAY WEEK ‘14
`
`
`
`Middleware...1
`
`(Consumer Electronics)
`
`oz» The best example of middleware in CE devices is
`from Myscript (formerly “Vision Objects”)
`
`+ This is what makes the Samsung Galaxy Notes possible
`
`+ Extremely powerful, configurable capabilities
`
`o Note-taking, handwriting recognition, mathematics (including
`equations), music notation, even “ink as a data—type” (same
`
`concept as in Windows, stores both ink and ASCII text)
`
`Ul: A thin layer of Samsung‘ look 8: feel
`
`Myscripti Mziddlew-are
`
`(CO_ntai'n-s most of the Notes’ functionality)
`
`Android
`
`Source: The author
`
`Samsung Galaxy Notes’ software stack
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Middleware...2
`(Large-Format I Commercial)
`
`«to The best middleware for large-format applications
`(in the author’s opinion) is Snowflake
`
`+ Good starting point for commercial applications
`
`+ Includes 30+ multi-touch apps (entertainment, presentation,
`creativity, media-browsing, etc.)
`
`+ Includes an SDK
`
`+ Runs on Win 8/7Nista/XP, Mac OS X Lion & Snow Leopard,
`and Linux Ubuntu
`
`«to Snowflake simplifies handling...
`
`4 Touch & gesture events, audio, video, images
`
`+ PDFs, 3D, on-screen keyboards, web browsing
`
`+ Multiple languages, QuickTIme integration, etc.
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Midd|eware...3
`
`0:0 Snowflake home screen
`1
`
`g-I
`
`@Eflw&%®%
`
`‘I
`
`3D ‘wzmir
`
`-"Hr Hr-nr-V f«lx».?r*: Ra ccrr.
`
`Bra-.-sf-0:-r
`
`E3-.143”, mars Colic-rt Conntedge
`
`Curling
`
`weuwmeea
`
`Frngqy
`
`(3-ooalc Maps
`
`‘£5;-(J. -,fl=..r: 1'
`
`$3.’.-‘:v.'*I3raw.I.J.'1 ‘.~’:de=o
`
`Urumfi
`
`SID DISPLAY WEEK ‘l4
`
`Source: NuiTeq
`
`
`
`Middleware...4
`
`«:0 Other alternative “middleware” for large-format
`
`+ Omnitapps
`
`o Less complete, Windows only, no SDK, more for product marketing
`
`+ Intuilab
`
`a Commercial multi-touch application platform with Kinect, RFID, etc.
`
`+ GestureWorks (ldeum)
`
`o Robust Flash multi-touch development environment
`
`4 22 Miles
`
`a Sales productivity application for iOS, Android, Windows & Mac
`
`+ Sotouch
`
`0 Application platform for wayfinding and presentations
`
`+ Fingertapps (Unlimited Realities)
`
`o Multi-touch demo software
`
`SID DISPLAY WEEK ‘14
`
`
`
`Conclusions
`
`«:0 Future Trends & Directions
`
`«to Suggested Reading on Touch
`
`0:0 Recommended Conferences & Trade Shows on Touch
`
`SID DISPLAY WEEK ‘l4
`
`
`
`Future Trends & Directions...1
`
`03° P-cap is here to stay
`
`+ It is totally dominating consumer electronics
`
`+ Consumer p-cap is getting much closer to meeting commercial
`application requirements
`
`0 For example, glove-touch and water-resistance
`
`+ P-cap’s capabilities are becoming increasingly attractive in
`commercial applications
`
`0 Curved touch-panels, particularly in automotive
`
`0 Light touch expected by ALL touch-panel users
`
`0 Flat-bezel in customer-facing applications
`
`0 Mu|ti—touch wherever images are viewed (e.g., photo-printing kiosk)
`
`+ The forecasts for commercial penetration of p-cap are
`MUCH too conservative
`
`SID DISPLAY WEEK ‘14
`
`intel"
`
`
`
`Future Trends & Directions...2
`
`oz» ITO-replacements are going to have an
`