3 CASE STUDY



PAPER MEMBRANE SWITCHES FOR FULLY FEATURED COMPUTER KEYBOARDS


This case study documents the successful development and evaluation of a novel membrane switch. The process results in lower environmental burdens and manufacturing costs than current screen-printed polyester solutions. Parameters governing the successful application of new materials to switchpad circuits include track resistance, migration resistance and switch contact resistances.

These have been characterised and are reported. Lifetime tests have been conducted and the printed films exposed to a regime of standard 85RH/85 °C and corrosive atmosphere tests. The lifetime of the switches has been investigated, and exceeds the normal industry specification. Several fully featured demonstration keyboards and switch membranes have been constructed and tested. These are currently in use in office environments.


INTRODUCTION


In 1999 112 million computers were produced. The requirement for cost effective manufacture in this area is intense. Early keyboards were constructed of individual mechanical switches, but these suffered from short service life. Industry required lower cost higher reliability solutions. The systems developed to fill these requirements fall into two broad categories, Membrane and Pad switches.

Pad switches typically utilise a printed circuit board within the product to carry the conductive traces and to provide mechanical support as the switch is pressed. Both switch contacts are formed on the substrate as part of the etched circuit pattern and protected by a chemically stable, conductive coating such as gold or carbon paste, to prevent corrosion. The switch is activated by pushing a conductive pad onto both contacts, enabling electrical conduction.




Simplified diagram of a pad switch.

Simplified diagram of a pad switch.


Larger, more complex keyboards and switch arrays usually utilise membrane switches. These typically consist of a polyester substrate screen-printed with silver paste to form the conductive traces and switch contacts. The substrate is folded back on itself, or faced to another printed sheet, and the conductive traces separated by an insulating spacer. The spacer is typically a polyester sheet with holes punched where the switch contacts need to touch, but may be a thick screen-printed dielectric layer. Activation of the switch is achieved by ressing the two contacts together.

The electrical characteristic of most importance to membrane switch operation is electrical resistance can be considered to comprise two elements, the line resistance and the contact resistance.




Simplified diagram of a membrane switch.

Simplified diagram of a membrane switch.


Line resistance is the total resistance of the printed track, leading to and away from the switch contacts. It is determined by the bulk resistivity of the printed ink and the physical geometry of the circuit track. The industry usually simplifies this further by using sheet resistivity (Ohms per square) figures, which assume a constant film thickness and allow easy calculations of line resistance if the length and width are known.

Contact resistance is somewhat more variable, and is influenced by the switch closing pressure, switch contact geometry, switch contact corrosion and contact topography.

Switch closing pressure is primarily determined by the force with which the key is depressed, but is also influenced by the pad design and the area of contact the key has with the top membrane. Corrosion products on the contacting surfaces will have a detrimental effect on contact resistance. For most domestic and office environments barrier coatings are not necessary. Contact topography is largely determined by the ink paste and deposition process, though it will change over the switch life. Typically, the root mean square (RMS) roughness of the printed switch contacts will be several microns (10 -6 m). Rough contacts have higher resistances than smooth counterparts, as the points of contact are fewer. Contact resistance normally increases as the switch is repeatedly used; an increase of 20% over 20 million cycles is not untypical.

The service life required of a membrane switch is dependent on the application it is used for;


Application
Life Cycles
Computer Keyboards
2.00 E 7
Typewriters
1.00 E 7
Copiers & Telephones
1.00 E 6
Calculators
5.00 E 5
Toys & Games
1.00 E 5

Typical switch life requirements


The likely cause of failure of a membrane switch is usually either electrical or mechanical wear of the switch contact surfaces, or chemical corrosion. Mechanical wear of the contact surfaces will occur if the switch contact fragments due to repeated compressive loading. Excessive electrical wear will result if the switched current is too high. Every switch cycle comprises an instant where the electrical current will pass through an infinitely small contact point. As this occurs the heat energy build up from the increased contact resistance can be enough to erode the contact. This may result in lost contact material, or a movement of material from one contact to the other.

Corrosion products can cause failure by two mechanisms, contact insulation and migration. Contact insulation can occur if elements of the ink film react with atmospheric pollutants. Sulphurous gasses can be particularly problematic, though the level found in most modern environments is generally insufficient to cause problems. Careful design of membrane and track geometries has also been employed to reduce exposure. Migration occurs if a potential gradient and critical quantities of moisture occur between two silver loaded circuit tracks. It is an electrochemical reaction resulting in the growth of conductive silver dendrites between the tracks, eventually resulting in a short circuit. Careful circuit design, minimising steep potential gradients in the same plane, minimises dendrite formation. The voltage and current a keyboard membrane switch operates at is determined by the circuitry it is in electrical contact with. In most cases this is an encoder IC. The encoder determines which switch on the matrix has been closed and sends the appropriate serial data to the CPU. Encoders present very high impedance to the membrane circuitry. This ensures only minute current flow occurs when the switch contacts are closed, and enables the use of moderately high resistance membrane circuitry. Typical current flow in contemporary computer keyboards is normally less than 0.1 mA, and the potential applied is normally 3-9 volts. The closed switch resistance permissible may be as high as 10 Kohms, though in practice the combined printed track and switch contact resistances are seldom higher than 500 Ohms. Open switch resistance needs to remain higher than the maximum permissible closed switch resistance if the encoder is not to erroneously trigger. In practice open switch resistances are seldom lower than tens of MegaOhms unless there are high levels of moisture present between circuit tracks or migration has occurred.


Method
To assess the application of CLF circuitry printed on paper for keyboard membrane switches, the screen-printed polyester components of an existing product were replaced. The Cherry® Business Line® keyboard was selected as it offers a highly optimised design. A rubber spring acts as the key return and determines the tactile feedback the keyboard user experiences. This design enables the mechanical aspects of each key to comprise just three separate parts, two of which are common to the whole product; the product housing, and a rubber spring, the third being each individual key cap. This contrasts to many other designs, which use individual spring and key cap mountings for each key.




Simplified diagram of the switch mechanism employed in the Cherry® keyboard.

Simplified diagram of the switch mechanism employed in the Cherry® keyboard.
Both of the Membrane Switch Substrates are formed from a single piece of material,
which is printed, die cut, and folded over so the printed side contacts itself
through the holes in the paper spacer.





Simplified Membrane Circuit to PWB Connection.

Simplified Membrane Circuit to PWB Connection. The printed paper circuit
is connected to the encoder PWB by screwing the PWB into place, trapping
the membrane circuitry between the PWB and an elastomer.


The printed paper membranes were prepared using an identical circuit layout to the screen-printed originals. The printed track width was 0.6 mm. The substrate chosen was a standard Gloss Art 80 gsm filled paper, usually used for glossy brochures and magazines. The substrate was imaged using a Heidelberg offset lithographic press, though Roland presses have also been successfully employed. No post printing curing phase was implemented. If it is considered necessary, the sheet resistivity of the printed films can be reduced by about 30 percent by hot air curing at 100 °C for 5 minutes.




Paper membrane switch constituent parts.

Paper membrane switch constituent parts, to right:
spring support, switch circuitry, spacer, switch circuitry




Lithographically printed paper switch membrane.



Final assembly of the encoder PWB and paper switch membrane.

Final assembly of the encoder PWB and paper switch membrane.


Results
Circuit Track Resistance The sheet resistivity of the paper membrane switch circuitry is ~ 150 mOhms/sq. This compares to about 75 mOhms/sq. for the screen-printed original. Resistivity measurements were averaged measurements obtained using a four-point probe Low OhmMeter. The resultant track resistances are about twice that of the original, just over 400 ohms in the worst case. The contact resistance is also about twice that of the original circuitry. The total switch resistance, worst case scenario, is ~ 600 ohms, comprising contact resistances of ~ 1 ohm, dependent upon key pressure.

Migration Resistance There are as yet no international standards for testing the susceptibility of printed tracks to migration. Initial work at Brunel indicates migration resistance of the lithographic ink films and the conventionally manufactured screen-printed films are comparable.

Lifecycle Performance
The lifecycle performance of membrane switches is normally determined by placing a unit in a mechanical jig, which activates the switch, whilst under electrical load, and increments a counter if a closed circuit is detected. These jigs typically operate at 3 cycles per second and will tolerate 2 consecutive null returns, i.e. open circuit, before terminating the test. The contact life of the CLF paper membrane switch has been tested to 25 million cycles, a figure in excess of most keyboard specifications, without failure. In addition to the mechanical lifecycle testing, two keyboards have been in use in real world office environments. One of these has been in place for 6 months without failure.

Conclusions
This work has demonstrated the suitability of lithographically printed paper membrane switches for computer keyboard applications. Acceptable electrical performance and mechanical longevity has been demonstrated. Paper/cellulose based substrates can achieve acceptable levels of reliability for application in contemporary electronic products. Coupled with the high production speed and low volume /unit ink requirements of offset lithographic printing this fabrication process can offer environmental benefits of: Reduced non-renewable resource use, as polyester is replaced by paper/cellulose material; Lower energy per unit manufacturing cost; Reduced raw material use as the quantity of deposited silver conductor is reduced.



 

Previous

Next page

 

Electronics Manufacturing Renewables Design

 

Brunel University

Cleaner Electronics Research Group © 2008