MULTIPLEX NOISE-RESISTANT KEYBOARD WITH ZERO CURRENT CONSUMPTION

ABSTRACT

This article discusses the keyboard device, which has increased noise immunity in combination with a two-wire connection and zero consumption current at rest. The text of the article in the form of schematic diagrams shows the circuitry of the device with a listing of the main advantages and disadvantages for each option and the development as a whole.

Keywords: keyboard, key, line, current consumption, multiplexing, noise immunity.

There are situations in which the use of a keyboard input device is required at a great distance from the registration unit. In this use case, the task is to minimize the length of the cable being pulled and to suppress interference induced in the line. The presented article describes a device that is resistant to interference with a keyboard with a two-wire connection to the analog input of the target microprocessor device, and also describes the algorithm of the reading program and the limitations associated with the use of this development. The scope of the described device may include: intercom systems, industrial control panels, IoT devices and other systems in which the use of digital keypads is not economically justified.

Functionally, the presented keyboard consists of keys and associated signal conditioning circuits, the shape of which, and, consequently, the frequency spectrum is individual for each key. The signal is generated when the key is pressed and completed before it is released, the signal itself is a damped pulse generated in this implementation using an RC circuit and transmitted in the form of current along the supply line of the device.

Figure 1. Structural - functional diagram of the device. Designations: PGD – Pulse Generation Device, K - Key

Unlike a circuit in which a certain amount of current or voltage is unique for identifying a key number that is pressed for a given key, this circuit is more noise-resistant due to the encoding of the key pressed, not only in amplitude but also in waveform (a parameter characterizing the shape) in addition, none of the presented circuits practically consumes current in the case when the key is pressed for a long time and does not consume current at all in the absence of pressing, since the circuit is open. The disadvantage of the proposed schemes is the difficulty in processing several simultaneously pressed keys.

In practice, processing of several simultaneously pressed buttons is usually not required, in this case a partial union of the nodes responsible for the formation of the pulse is possible, namely the use of one capacitor (or resistor) for all keys (in the case of using an RC circuit). In this case, the pulse parameters are set by a resistor or capacitor connected in series with the key.

Figure 2. Various implementation options

The presented schemes have the following features:

1. Scheme 1 is the most functional, but requires 1C2R elements per key in order to realize the possibility of registering multiple pressures due to an individual forming device for each button.

2. Resistors R1 and R2 are needed to discharge capacitors when the key is pressed, returning the circuit to its initial state. Resistors R3 and R4 are used to set the peak current.

3. Scheme 2 requires 1C1R elements per key and allows you to register only one press, but while ensuring the same peak current in each arm due to the presence of a common current-limiting resistor R6.

4. Scheme 3 requires 1R elements per key, also allowing only one press to be recorded, however, the same resistor is used to set the parameter τ and the peak current, as a result of which the peak current in the arms is different.

Due to the high complexity of the multi-press processing program and the high cost of the capacitors included in the circuit, circuit 3, which allows only one press at a time, with setting the pulse parameter τ and the peak current by the resistor, is of the greatest practical importance.

The algorithm of the recording click program can be implemented as follows:

1. Sampling data to a buffer whose length is 3 ∙ max (τ) and then passing it through a digital filter, the duration of the pulse response of which is equal to the position of the nearest measuring point (see Clause 3) in order to reduce distortions introduced by the convolution operation.

2. Current measurement at time  and  ().

The position of the measuring points is selected from the considerations of the presence in the signal of a certain fraction of noise, which will affect the accuracy of determining the parameters, as well as the inaccuracies in determining the beginning of the pulse.

Thus, it is more rational to choose:

3. Fiding  and  of the signal parameters based on the system of equations for the current in the RC circuit

4. Based on the found values of the parameters  and , the number of the pressed key is determined by the greatest coincidence with the reference values recorded in the table.

In all schemes, the maximum current in the circuit flows at the moment the key is pressed, which is convenient from the software implementation of the above algorithm, since there is no need to interrogate the keyboard in a loop or by interrupting the timer. An abrupt increase in current at the moment of pressing allows the event to be recorded in synchronous mode, using the signal front for synchronization purposes. Snap to the front limits the maximum length of the line due to the requirements for its integrating properties.

Figure 3. Current graphs depending on the pressed key and on the interval between presses for circuit 3. Scale: X = 200ms / Div, Y = 2mA / Div.

Summing up the above, we can list the following disadvantages of the scheme:

1. The difficulty of determining the set of pressed keys and the high number of elements for a circuit that supports this functionality: 1C2R per button, versus 1R1C and 1R for circuits 2 and 3, respectively.
2. The difficulty of obtaining a large number of buttons due to increasing requirements for line parameters and reducing noise immunity.
3. The complexity of the registration program clicks.

The obvious advantages compared to a scheme that uses the coding of a pressed key by level include:

1. Greater noise immunity due to the use of two signal parameters.
2. Less current consumption (key holding current is small).
3. The ability to synchronously process keystrokes and release keys on the front / bottom.

Conclusion

The device described in the article, despite the complexity of implementation, due to its low cost and increased noise immunity compared to level coding, can be used in a wide range of applied solutions.

References:

1. Бессонов Л. А. Теоретические основы электротехники. Электрические цепи. – М.: «Высшая школа». - 1996. - С. 560.
2. Угрюмов Е. П. Цифровая схемотехника. – СПб.: БВХ-Петербург, 2004. - C. 528.: ил.
3. Болл Стюарт Р. Аналоговые интерфейсы микроконтроллеров. – М.: Издательский дом «Додэка-XXI», 2007. – 360 с.: ил. (Серия «Программируемые системы»).