A change in signal strength, sensed at the intersecting nodes of any vertical and horizontal element, indicates the presence or absence of a finger at that point. Resistive touch-screens or touchpads may have a simple switch shorting the x and y elements at that point, whereas “projected capacitance” touch-screens or touchpads will show a change of capacitive coupling between the x and y elements.
Each node is scanned, one after the other, to determine the presence or absence of multiple fingers over the whole array. The density of these nodes determines how close together fingers can be while still being able to distinguish individual fingers.
Because vertical (x) elements can only ever intersect horizontal (y) elements, this arrangement restricts the number of nodes to the product of the number of rows and columns (x times y). This arrangement creates “non-touch” zones along two or three edges of the touch-screen (shown shaded in figure 1) where links are made between the conductive elements and the connector.
In the “Binstead Designs” arrangement there are no fixed receiving or transmitting conductive elements, each element being considered as both transmitting or receiving at different stages in the scanning cycle. All of the conductive elements run diagonal to the orientation of the touch-screen – see figure 2.
Each element starts at the connector edge and runs diagonally through the touch-screen to a second edge which is perpendicular to the connector edge. Here it changes direction and runs through the touch-screen again until it reaches a third edge, which is parallel to and opposite the connector edge. Even numbered elements run in one direction while odd numbered elements run in an opposing direction.
Conductive element 4, in figure 2, has been emphasised to illustrate the typical route of an element. This arrangement results in each element intersecting all the other elements, resulting in a significant increase in the number of intersections. It also eliminates any “non-touch” zones along the edges of the touch-screen.
Although this lattice layout leads to a staggered arrangement of intersecting nodes, as shown circled in figure 2, these can be rearranged into a good approximation of an orthogonal x/y array by distorting the conductive elements within the sensing area.
Figure 3 shows how the lattice of figure 2 can be distorted to emulate a 7 x 4 orthogonal array, which can be treated by software as though it were a real 7 x 4 orthogonal array. Conductive element 4, in Figure 3, has been emphasised to illustrate the typical distorted route of an element. This diagonal lattice array affords at least five significant advantages over standard orthogonal x/y arrays.
1) The number of intersections is almost doubled, thereby significantly improving the ability to distinguish between fingers that are close together.
A comparison of figures 1 and 2 show that, for 8 connections, an x/y array has a maximum of 16 intersections whereas a lattice array has a maximum of 28 intersections (shown circled). If n is the total number of connections, then the standard x/y array has a maximum of n2 /4 intersections, whereas the diagonal lattice array has a maximum of (n2-n) /2 intersections.
If the touch-screen has a 9:16 format, then the diagonal lattice array has 1.93 times as many intersections as the x/y array. For example, assuming a total of 100 inputs, split 36/64 (9:16 format), the x/y array will have 2304 (36 x 64) intersections and the diagonal lattice array will have 4450 (50 x 89) intersections.
2) “Non-touch” zones along the edges of the touch-screen are eliminated.
There are no tracks along the sides of the touch-screen or the edge opposite the connector. This means that the touch-screen can sense right up to the very edge of the touch-screen film on these three sides.
If the connector “tail” is folded back along the line A – B in figures 2 and 3, then the touch-screen can effectively sense right up to all four edges of the touch-screen film.
Standard x/y arrays have multiple tracks running along the edges of a touch-screen which, as well as creating a “non-touch” zone, also causes some visual degradation. This is often hidden by a bezel or some appropriate artwork, or tolerated as a necessary evil.
This problem does not exist with a diagonal lattice array as there are no tracks along the edges of the touch-screen. This has the advantage that multiple screens can be mounted side by side and operated as one very wide touch-screen, with no visual impairment at the junction of neighbouring touch-screens, and seamless touch operation across the boundary. These large touch-screens can be tens of metres wide, on separate films, or they can be manufactured on the same film.
3) All links to the connector come directly from one single edge of the touch-screen.
Standard touch-screens may have connectors on two or more edges, and those that only have connectors on one edge need complex track routing around the edges and corners to get connections between the touch-screen elements and the connector.
With the lattice array, links between the connector and the first edge of the touch-screen are short and direct, thereby greatly reducing the resistance of conductive elements between the connector and the touch-sensing area of the touch-screen. The connector tail is easily swapped between the long side or the short side of a rectangular touch-screen.
4) Fewer, and/or lower pin-count chips and connectors are required, leading to cost savings and improved reliability.
For a given number of intersections, about 25% to 30% less connections are required for a diagonal lattice array than are required for an x/y array. This means that chips with lower pin counts can be used. For example, The Microchip touch controller MTCH6301 is a 44 pin TQFP device with 18 transmit pins and 13 receive pins. This could theoretically be replaced with a 36 pin device with no loss of functionality.
5) All conductive tracks are of similar length and follow very similar routes.
Track lengths can vary considerably when the standard x/y layout is used. For example, track 8, in figure 1, is considerably longer than track 5. If a 9:16 format is used, then the horizontal and vertical conductive elements can be nearly twice as long in one direction as they are in the other direction. Track lengths between the touch-sensing area and the connector are also extremely variable.
These factors can lead to some variation in sensitivity over the surface of the touch-screen that is difficult to predict. With the diagonal lattice layout all conductive elements, within the touch-screen sensing area, are of similar length, creating much more uniform touch sensing over the touch-screen surface. Any loss of sensitivity, due to the resistance of the conductive elements, will be very predictable and easily compensated for in software.
Although, in a 9:16 format touch-screen, the conductive elements within the “touch” zone are about 12% longer than the longest elements used in an x/y array, the extra resistance may be offset by the lower resistance of the conductors in the “non-touch” zone leading to the connector.
There are a number of other advantages derived from the elements being diagonal to the touch-screen. These include:
a) Improved immunity to LCD interference.
Experiments have shown that diagonal elements have about 30% less LCD scan line interference compared to vertical/horizontal elements.
b) Improved sensing in “self capacitance” mode.
Conductive elements at 45 degrees to the LCD panel have similar sensitivity to each other, whereas, when some elements are horizontal and others are vertical, the horizontal elements can be much less sensitive than the vertical elements.
Although these products can be manufactured using standard techniques involving materials such as Indium Tin Oxide (ITO) or fine conductive mesh, Binstead Designs has developed and patented a much simpler and more environmentally friendly technique using very fine insulation coated wire.
This involves three simple steps; forming the desired pattern of wires on a clear plastic film, laminating a second film over the wires, and finally terminating the wires to a connector.
The process is readily automated for “reel to reel” mass production, yet, because the design is controlled by software, each touch-screen can, if required, be tailored to suit individual customer needs. The resulting touch-screens are very thin and flexible, being about 100 microns thick, with superb clarity.
They can be rolled up for posting. They can even be creased with no loss of touch sensing functionality, a feature which can be exploited when folding the tail around the back of the display.
When mounted on a sheet of glass, and used with a new method of touch-sensing, also developed and patented by Binstead Designs, these touch-screens can sense many fingers simultaneously through 0.5mm of glass to more than 20mm of glass.
Although the standard orthogonal x/y layout works very well, bezels are becoming narrower so manufacturers are constantly being required to reduce the width of the “non-touch” zone around the sides of the touch-screen.
Also, the requirement to be able to recognise separate fingers when they are brought within a centimetre of each other, even in very large touch-screens, means that the density of intersecting nodes has become an important issue.
Manufacturers overcome the finger resolution issue by increasing the number of inputs to the touch-screen to a point where there may be hundreds of inputs in a very large touch-screen.
However, this solution exacerbates the problem regarding the width of the “non-touch” zone, by pushing even more tracks into it. The technology described here totally alleviates the problem of the “non-touch” zone by eliminating any tracking around the edges of the touch-screen.
It also alleviates, to some extent, the finger resolution problem by significantly increasing the number of touch-sensing nodes that can be derived from a fixed number of inputs. All the advantages listed above are compelling reasons for touch-screen manufacturers to consider adopting this new layout.
About the author:
Ron Binstead is the owner/manager of Binstead Designs Ltd. – www.binsteaddesigns.com