Keyboard Technology

Key switches

In the first electronic keyboards in the early 1970s, the key switches were individual switches inserted into holes in metal frames. These keyboards cost from 80–120 US dollars and were used in mainframe data terminals. The most popular switch types were reed switches (contacts enclosed in a vacuum in a glass capsule, affected by a magnet mounted on the switch plunger).

In the mid-1970s, lower-cost direct-contact key switches were introduced, but their life in switch cycles was much shorter (rated ten million cycles) because they were open to the environment. This became more acceptable, however, for use in computer terminals at the time, which began to see increasingly shorter model lifespans as they advanced.

In 1978, Key Tronic Corporation introduced keyboards with capacitive-based switches, one of the first keyboard technologies to not use self-contained switches. There was simply a sponge pad with a conductive-coated Mylar plastic sheet on the switch plunger, and two half-moon trace patterns on the printed circuit board below. As the key was depressed, the capacitance between the plunger pad and the patterns on the PCB below changed, which was detected by integrated circuits (IC). These keyboards were claimed to have the same reliability as the other "solid-state switch" keyboards such as inductive and Hall-Effect, but competitive with direct-contact keyboards. Prices of $60 for keyboards were achieved and Key Tronic rapidly became the largest independent keyboard manufacturer.

Meanwhile, IBM made their own keyboards, using their own patented technology: Keys on older IBM keyboards were made with a "buckling spring" mechanism, in which a coil spring under the key buckles under pressure from the user's finger, triggering a hammer that presses two plastic sheets (membranes) with conductive traces together, completing a circuit. This produces a clicking sound, and gives physical feedback for the typist indicating that the key has been depressed.^[13]

The first electronic keyboards had a typewriter key travel distance of 0.187 inches (4.75 mm), keytops were a half-inch (12.7 mm) high, and keyboards were about two inches (5 cm) thick. Over time, less key travel was accepted in the market, finally landing on 0.110 inches (2.79 mm). Coincident with this, Key Tronic was the first company to introduce a keyboard which was only about one inch thick. And now keyboards measure only about a half-inch thick.

Keytops are an important element of keyboards. In the beginning, keyboard keytops had a "dish shape" on top, like typewriters before them. Keyboard key legends must be extremely durable over tens of millions of depressions, since they are subjected to extreme mechanical wear from fingers and fingernails, and subject to hand oils and creams, so engraving and filling key legends with paint, as was done previously for individual switches, was never acceptable. So, for the first electronic keyboards, the key legends were produced by two-shot (or double-shot, or two-color) molding, where either the key shell or the inside of the key with the key legend was molded first, and then the other color molded second. But, to save cost, other methods were explored, such as sublimation printing and laser engraving, both methods which could be used to print a whole keyboard at the same time.

Initially, sublimation printing, where a special ink is printed onto the keycap surface and the application of heat causes the ink molecules to penetrate and commingle with the plastic modules, had a problem because finger oils caused the molecules to disperse, but then a necessarily very hard clear coating was applied to prevent this. Coincident with sublimation printing, which was first used in high volume by IBM on their keyboards, was the introduction by IBM of single-curved-dish keycaps to facilitate quality printing of key legends by having a consistently curved surface instead of a dish. But one problem with sublimation or laser printing was that the processes took too long and only dark legends could be printed on light-colored keys. On another note, IBM was unique in using separate shells, or "keycaps", on keytop bases. This might have made their manufacturing of different keyboard layouts more flexible, but the reason for doing this was that the plastic material that needed to be used for sublimation printing was different from standard ABS keytop plastic material.

Three final mechanical technologies brought keyboards to where they are today, driving the cost well under $10:

1. "Monoblock" keyboard designs were developed where individual switch housings were eliminated and a one-piece "monoblock" housing used instead. This was possible because of molding techniques that could provide very tight tolerances for the switch-plunger holes and guides across the width of the keyboard so that the key plunger-to-housing clearances were not too tight or too loose, either of which could cause the keys to bind.
2. The use of contact-switch membrane sheets under the monoblock. This technology came from flat-panel switch membranes, where the switch contacts are printed inside of a top and bottom layer, with a spacer layer in between, so that when pressure is applied to the area above, a direct electrical contact is made. The membrane layers can be printed by very-high volume, low-cost "reel-to-reel" printing machines, with each keyboard membrane cut and punched out afterwards.

Plastic materials played a very important part in the development and progress of electronic keyboards. Until "monoblocks" came along, GE's "self-lubricating" Delrin was the only plastic material for keyboard switch plungers that could withstand the beating over tens of millions of cycles of lifetime use. Greasing or oiling switch plungers was undesirable because it would attract dirt over time which would eventually affect the feel and even bind the key switches (although keyboard manufacturers would sometimes sneak this into their keyboards, especially if they could not control the tolerances of the key plungers and housings well enough to have a smooth key depression feel or prevent binding). But Delrin was only available in black and white, and was not suitable for keytops (too soft), so keytops use ABS plastic. However, as plastic molding advanced in maintaining tight tolerances, and as key travel length reduced from 0.187-inch to 0.110-inch (4.75 mm to 2.79 mm), single-part keytop/plungers could be made of ABS, with the keyboard monolocks also made of ABS.

Control processor

Scanning procedure

Computer keyboards include control circuitry(Keyboard controller (computing)) to convert key presses into key codes (usually scancodes) that the computer's electronics can understand. The key switches are connected via the printed circuit board in an electrical X-Y matrix where a voltage is provided sequentially to the Y lines and, when a key is depressed, detected sequentially by scanning the X lines.

The first computer keyboards were for mainframe computer data terminals and used discrete electronic parts. The first keyboard microprocessor was introduced in 1972 by General Instruments, but keyboards have been using the single-chip 8048microcontroller variant since it became available in 1978. The keyboard switch matrix is wired to its inputs, it converts the keystrokes to key codes, and, for a detached keyboard, sends the codes down a serial cable (the keyboard cord) to the main processor on the computer motherboard. This serial keyboard cable communication is only bi-directional to the extent that the computer's electronics controls the illumination of the caps lock, num lock and scroll lock lights.

One test for whether the computer has crashed is pressing the caps lock key. The keyboard sends the key code to thekeyboard driver running in the main computer; if the main computer is operating, it commands the light to turn on. All the other indicator lights work in a similar way. The keyboard driver also tracks the Shift, alt and control state of the keyboard.

Some lower-quality keyboards have multiple or false key entries due to inadequate electrical designs. These are caused by inadequate keyswitch "debouncing" or inadequate keyswitch matrix layout that don't allow multiple keys to be depressed at the same time, both circumstances which are explained below:

When pressing a keyboard key, the key contacts may "bounce" against each other for several milliseconds before they settle into firm contact. When released, they bounce some more until they revert to the uncontacted state. If the computer were watching for each pulse, it would see many keystrokes for what the user thought was just one. To resolve this problem, the processor in a keyboard (or computer) "debounces" the keystrokes, by aggregating them across time to produce one "confirmed" keystroke.

Some low-quality keyboards also suffer problems with rollover (that is, when multiple keys pressed at the same time, or when keys are pressed so fast that multiple keys are down within the same milliseconds). Early "solid-state" keyswitch keyboards did not have this problem because the keyswitches are electrically isolated from each other, and early "direct-contact" keyswitch keyboards avoided this problem by having isolation diodes for every keyswitch. These early keyboards had "n-key" rollover, which means any number of keys can be depressed and the keyboard will still recognize the next key depressed. But when three keys are pressed (electrically closed) at the same time in a "direct contact" keyswitch matrix that doesn't have isolation diodes, the keyboard electronics can see a fourth "phantom" key which is the intersection of the X and Y lines of the three keys. Some types of keyboard circuitry will register a maximum number of keys at one time. "Three-key" rollover, also called "phantom key blocking" or "phantom key lockout", will only register three keys and ignore all others until one of the three keys is lifted. This is undesirable, especially for fast typing (hitting new keys before the fingers can release previous keys), and games (designed for multiple key presses).

As direct-contact membrane keyboards became popular, the available rollover of keys was optimized by analyzing the most common key sequences and placing these keys so that they do not potentially produce phantom keys in the electrical key matrix (for example, simply placing three or four keys that might be depressed simultaneously on the same X or same Y line, so that a phantom key intersection/short cannot happen), so that blocking a third key usually isn't a problem. But lower-quality keyboard designs and unknowledgeable engineers may not know these tricks, and it can still be a problem in games due to wildly different or configurable layouts in different games.

Connection types

There are several ways of connecting a keyboard to a system unit (more precisely, to its keyboard controller) using cables, including the standard AT connector commonly found on motherboards, which was eventually replaced by the PS/2 and theUSB connection. Prior to the iMac line of systems, Apple used the proprietary Apple Desktop Bus for its keyboard connector.

Wireless keyboards have become popular for their increased user freedom. A wireless keyboard often includes a required combination transmitter and receiver unit that attaches to the computer's keyboard port. The wireless aspect is achieved either by radio frequency (RF) or by infrared (IR) signals sent and received from both the keyboard and the unit attached to the computer. A wireless keyboard may use an industry standard RF, called Bluetooth. With Bluetooth, the transceiver may be built into the computer. However, a wireless keyboard needs batteries to work and may pose a security problem due to the risk of data "eavesdropping" by hackers. Wireless solar keyboards charge their batteries from small solar panels usingsunlight or standard artificial lighting. An early example of a consumer wireless keyboard is that of the Olivetti Envision.