AN239751 Flyback inductive sensing (ISX) design guide

About this document

Scope and purpose

This document provides a technical quick start guide for designing and implementing inductive-sensing solutions using the flyback inductive sensing technology from Infineon.

Intended audience

The intended audience for this document is primarily developers, engineers, and designers who are interested in implementing sensing solutions using Infineon’s flyback inductive sensing technology.

Note:

The document assumes a basic understanding of Infineon’s PSOC™ architecture, basic electronics, and electrical engineering principles as well as software development and programming experience. If you are new to the CAPSENSE™ and PSOC™ architecture offered by Infineon, see the

References

section to learn about these offerings.

Abbreviations

Table 1.

Abbreviations

Abbreviation	Description
ADC	Analog-to-digital converter
CMOD	Modulator capacitor
CAPDAC	Capacitive digital-to-analog converter (DAC)
C P	Parasitic capacitance
CSD	Self-capacitance sensing
C SH	Shield tank capacitor
C V REF	Voltage reference capacitor
CSX	Mutual-capacitance sensing
EMC	Electromagnetic compatibility
ESD	Electrostatic discharge
GND	Ground
GPIO	General-purpose input/output
HID	Human interface device
HMI	Human machine interface
IDAC	Current-output digital-to-analog converter
IDE	Integrated development environment
IIR	Infinite impulse response (filter)
ISX	Flyback inductive sensing method
L S	Effective inductance of an ISX sensor
MCU	Microcontroller unit
MFC	Multi frequency scan
MSC	Multi-sense converter
MSCLP	Multi-sense converter version 3 – Low Power
PRS	Pseudo random sequence
PSOC™	Programmable system on chip controllers offered by Infineon
SNR	Signal-to-noise ratio
SSC	Spread spectrum clock
VREF	Programmable reference voltage blocks available inside PSOC™ used for MSCLP and ADC operation

Introduction

Inductive sensing provides an innovative low-cost solution to detect the presence or movement of nearby metallic or conductive objects. The technology enables a robust low-cost HMI with metallic overlays that seamlessly integrates in the product. Inductive sensing addresses the short falls of capacitive sensing and enables new and creative methods for HMI.

Working principle

Inductive sensing works on the principle of electromagnetic coupling between a sensor coil and the metal/highly conductive object to be detected. When the metal target enters the electromagnetic field induced by a sensor coil, some of the electromagnetic energy is transferred into the metal target as shown in the following figure. This transferred energy causes a circulating electrical current called an eddy current. The eddy current flowing in the metal target induces a reverse electromagnetic field on the sensor coil, which results in a reduction of the effective inductance of the sensor coil.

Figure 1.

Field coupling between sensor and metal/conductive target

Applications

The following applications are the typical uses of inductive sensing:

• Proximity metal detection

• Replacing mechanical open/close switches

• Buttons (industrial keypads and ON/OFF buttons)

• Rotation detection (rotary encoders, flow meters, fan speed RPM detection, rotary control knob)

• Linear encoders

• Spring compression detection

• Replacement for Hall sensor + Magnet attachment sensing

Features

The following features make the inductive sensing an appealing sensing solution:

• Force detection – for multi-level force triggers

• Inherently liquid tolerant

• Works with gloves

• Enables hermetically sealed HMIs

• Enables metallic-surface HMIs

• Seamless integration with existing applications

• Removes the need for cut-outs for metal enclosures

Methods of inductive sensing

There are various ways to implement inductive sensing solution, the two most popular are:

• Impedance-based with resonant capacitor

• Flyback inductive sensing (ISX)

The impedance-based with resonant capacitor method is implemented with a parallel combination of sensor inductance and the external capacitor called tank circuit as shown in the following figure. The reduction in the sensor coils effective inductance causes an upward shift in the resonant frequency of the tank circuit. This shift in resonant frequency changes the amplitude of the signal across the sensor coil. The change in the amplitude of the sensor coil signal is measured by the PSOC™ MCU to detect the presence of the metal target in the proximity-sensing distance. To know more about this method, see

AN219207 - Inductive sensing design guide

.

Figure 2.

LC tank resonance circuit

Introduction to flyback inductive sensing using ISX

ISX inductive sensing works on the principle of inductive flyback operation, where the energy is stored when current passes through the inductor coil and released when the power is removed.

A magnetic field created by the current in an inductor accumulates energy calculated by the formula as follows:

Figure 3.

Energy accumulated in an Inductor

$\mathrm{<p>E</p>}<p>=</p>\frac{\mathrm{<p>L</p>}<p>\times </p>{\mathrm{<p>I</p>}}^{\mathrm{<p>2</p>}}}{\mathrm{<p>2</p>}}$

Where,

• L – coil inductance value

• I – coil current

This energy produces a back-EMF voltage that produces a current during a flyback operation period. Multi Sense Converter – Low Power (MSCLP) converts this current to digital code (raw counts). The produced current is proportional to the sensor coil inductance. The relative change of inductance due to the high-conductive object in proximity is measured to detect the object presence.

The sensing functionality is achieved using a combination of hardware and firmware.

Figure 4

illustrates the flyback inductive sensor connection.

Figure 4.

Simplified diagram of PSOC™ 4 MSCLP with flyback inductive sensor

Direct inductance measurement using the flyback method allows wide operation frequency range and simplified schematics.

Figure 5.

Operating frequency range of flyback inductive sensing

Typically, the inductive sensors are circular metallic coils etched on a PCB.

Figure 6

illustrates an example of the inductive sensor button.

Figure 6.

Illustration of an inductive touch sensor

Figure 7

shows a plot of raw count over time for an inductive sensor. When a conductive object approaches the sensor, the effective sensor inductance (L

S

) decreases, and the converter output decreases. The firmware normalizes the raw count such that the raw counts increase when L

S

decreases. By comparing the change in the raw count to a predetermined threshold, firmware logic decides whether the sensor is active.

Figure 7.

Raw count vs. samples

The output raw count of the multi-sense converter is a digital value that is proportional to the inductance of the sensor as mentioned in the following equation.

Figure 8.

Raw count and sensor inductance relationship in ISX

$Rawcount=G_{ISX}\mathrm{.}{L}_s$

Where,

• G_ISX = Inductance to digital conversion gain

• L_S = Effective inductance of the sensor

Advantages of flyback inductive sensing

Table 2.

Advantages of flyback inductive sensing

Feature	Description	Advantage
Direct inductance measurement	ISX method measures the inductance directly without using a resonant capacitor	Simplifies sensor design requiring very few components
High resolution and sensitivity	High clock frequency and sampling rate	Better sensitivity than resonance method to sense small deflections.
Broad operating frequency	System can operate over a wide range of frequencies with support for MFS 1 , SSC, and PRS Similar sensors can have a common frequency	Simple design for EMC Little to no impact of manufacturing variations on sensor response
Linear response curve	Does not need to be tuned to operate at a resonance frequency. Change in operating frequency does not have a significant impact on response	Handles manufacturing variations. Easy tuning and development
Lower temperature drift	Slow drift due to device temperature change is compensated by base line tracking	Low sensitivity to temperature changes
Simple schematics	Simple schematics and fewer discrete components	Cost effective

1

Multi Frequency Scan (MFS) is not supported with Common mode filter feature. See

Common mode noise filter (CMF)

.

Terminology

This section defines ISX-related terms that are used throughout the document:

Table 3.

Terminology

Terms	Description
Sensor coil	Inductive coil etched on PCB
Overlay	Metal plane over the sensor coil. It is a touch surface exposed to user. This is made of materials like Aluminum, Stainless steel, copper, etc.
Spacer	Non -conductive material used to hold the overlay on top of the coil
Air gap	Air gap refers to distance between the sensor coil on the PCB to the overlay surface on the top. This cavity between the PCB and overlay is recommended to be kept unfilled with any material
Deflection	This term refers to deflection of metal overlay (target) present on the Inductive sensing sensor coil in the button
Detection distance	Detection distance is the distance where the reduction in inductance exceeds some threshold values. The detection distance depends on the sensor’s magnetic field propagation. A longer propagation distance provides a longer detection range
Target object	Object of which presence or movement is to be detected by sensor. It is made of metal or ferrite
Raw count	The effective sensor inductance is converted into a count value by the MSCLP hardware. The converted digital count value is referred to as the raw count. Processing of the raw count results in ON/OFF states for the sensor
Baseline	A value resulting from a firmware algorithm that estimates a trend in the raw count when there is no conductive object/movement of target object present over the sensor, and when the raw count is within noise thresholds. The baseline is less sensitive to sudden changes in the raw count and provides a reference point for computing the difference count
Signal (Difference count)	Subtracting the baseline level from the raw count produces the difference count that is used in the decision process. The thresholds are offset by a constant amount from the baseline level
Signal-to-noise ratio (SNR)	As the name suggests it is the measure of the quality of the information in a signal to noise present in the signal. In a capacitive proximity sensing system, SNR implies reliable and accurate detection

1

Multi Frequency Scan (MFS) is not supported with Common mode filter feature. See

Common mode noise filter (CMF)

.

Resources for getting started with ISX

This section describes the resource offered by Infineon to get started on designing inductive sensing solutions.

Online resources

• Infineon inductive sensing webpage

• Infineon ModusToolbox™ webpage

• ModusToolbox™ software help on GitHub

Getting started kits

CY8CPROTO-040T-MS Multi-Sense Prototyping Kit

The CY8CPROTO-040T-MS Multi-Sense Prototyping Kit enables evaluation and development of capacitive, inductive and liquid-level sensing solution using the PSOC™ 4000T device with fifth-generation CAPSENSE™ and multi-sense technology offering ultra-low power touch HMI solution (MSCLP).

Figure 9.

CY8CPROTO-040T-MS kit contents

For more information about the kit, see the

kit documentation

.

CY8CPROTO-041TP PSOC 4100T Plus CAPSENSE Prototyping kit

The CY8CPROTO-041TP PSOC™ 4100T Plus CAPSENSE™ Prototyping kit enables evaluation and development of capacitive, inductive, and liquid-level sensing solution, using the PSOC™ 4100T Plus device with fifth-generation CAPSENSE™ and multi-sense technology, offering ultra-low power touch HMI solution (MSCLP). The expansion boards that are part of the CY8CPROTO-040T-MS Multi-Sense Prototyping Kit packaging are also compatible with this kit.

Figure 10.

CY8CPROTO-041TP kit

For more information about the kit, see the

kit documentation

.

Development tools

ModusToolbox

ModusToolbox™ software suite is used for the development of PSOC™ 6- and PSOC™ 4-based CAPSENSE™ applications. You can download ModusToolbox™ from

Infineon Development Center

. Before you start working with this software, it is recommended that you go through the

quick start guide

and

user guide

. ModusToolbox™ software trainings can be found

here

. If you have ModusToolbox™ IDE installed in your system, you can create a CAPSENSE™ application for the devices supported in ModusToolbox™.

Sensor Designer

Sensor Designer tool simplifies the process of designing flyback inductive sensors and the mechanical stackups for the target application. With just system-level requirements, it can create optimal designs and recommendations. The tool allows export of .dxf files, which can be seamlessly integrated into your PCB and product enclosures/mechanicals.

The Sensor Designer tool can be launched from the Quick Panel in ModusToolbox™.

Figure 11

shows the tool interface.

Figure 11.

ISX Sensor Designer tool

Based on the input parameters, *.dxf files are generated which can then be imported into the CAD software. To learn more about the tool, click the

Help

tab in the toolbar or press F1. The same tool also helps create liquid level sensor designs (see the Liquid-level sensing with PSOC™ 4 CAPSENSE™ application note for details)

CAPSENSE Configurator

The CAPSENSE™ Configurator tool in ModusToolbox™ is used to configure the hardware and software parameters of the inductive (as well as capacitive) sensing application project. For more details on configuring CAPSENSE™ in ModusToolbox™, see the ModusToolbox™ CAPSENSE™ Configurator guide and CAPSENSE™ middleware library.

Figure 12

shows how to open the CAPSENSE™ configuration tool in ModusToolbox™. Alternatively, it can also be launched from the

Quick panel

in the ModusToolbox™.

Figure 12.

ModusToolbox™ - Quick panel

CAPSENSE middleware

ModusToolbox™ provides CAPSENSE™ middleware, which can be used to create a inductive sensing and capacitive sensing application with PSOC™ devices by simply configuring parameters in the CAPSENSE™ configuration tool. The middleware also provides an application programming interface (APIs) to simplify firmware development. See the

Middleware library and APIs

section for more details.

CAPSENSE Tuner

ModusToolbox™ also offers a GUI CAPSENSE™ Tuner provides an easy-to-use interface to visualize the sensor data and tune widgets and sensors parameters as configured in the application. This tool can be opened from the Device Configurator by selecting

Launch

CAPSENSE™

Tuner

as shown in

Figure 12

. See the

CAPSENSE™ Configurator user guide

for more details.

Code examples

Two buttons keypad demo

The

PSOC™ 4: MSCLP inductive sensing touch-over-metal keypad-2

code example demonstrates the implementation and tuning procedure of two keypad ISX buttons using fifth-generation CAPSENSE™ (MSCLP).

Four buttons keypad demo

The

PSOC™ 4: MSCLP inductive sensing touch-over-metal keypad-4 demo

code example demonstrates the implementation of four keypad ISX buttons using fifth-generation CAPSENSE™ (MSCLP).

For more code examples, see

GitHub

.

Capacitive and inductive sensing buttons example

The

PSOC™ 4: MSCLP capacitive and inductive sensing buttons

code example demonstrates the implementation of a CSD button, a CSX button, and four keypad ISX buttons using fifth-generation CAPSENSE™ (MSCLP) in the same project.

For more code examples, see

GitHub

.

1

Multi Frequency Scan (MFS) is not supported with Common mode filter feature. See

Common mode noise filter (CMF)

.

Designing an ISX inductive-sensing solution

Figure 13

illustrates the typical flow of an ISX inductive-solution design. This flow is like any other electronic product design flow except that these designs involve an additional step called “Tuning”.

Like any other product, defining the requirement specifications is recommended before designing an inductive-sensing solution. These specifications will act as design objective to be met for the system. The specifications for inductive-sensing design can be listed as follows:

• Type/material of the object to be detected

• Type of sensors, for example,

  • Touch-over-metal (ToM) buttons

  • Force-touch

  • Slider

  • Rotary encoder

  • Presence detection, etc.

• Maximum sensing distance required

• Behavior in the presence of nearby metal

• Operation in extreme environmental conditions

• Operation in areas with high EMI noise

Figure 13.

Recommended ISX system design flow

Application use cases

Based on the features offered by the flyback inductive sensing, the use cases and target applications can be categorized as follows:

Touch-over-metal

Flyback inductive sensing enables touch-over-metal buttons offering:

• Enhanced design of everyday objects with seamless, hermetically sealed metallic surfaces with integrated inductive touch buttons on metal surfaces.

• Cost-effective and highly reliable buttons that are immune to moisture and dirt.

• Flexible sensor design supporting a wide variety of sensor sizes and stack-ups.

Figure 14.

Example application use cases of metal over touch

The use cases and applications of metal over touch can be:

• Coffee machines

• Microwave

• Oven

• Refrigerator

• Washing machine

• Dish washer

See

Table 4

for the typical requirements for the touch-over-metal use cases.

Underwater/waterproof touch

Flyback inductive sensing is inherently immune to water/liquid which make it ideal for underwater and waterproof touch interface/button applications such as:

• Underwater cameras

• Personal care products (shaver)

• Intercom, EV charging, and other systems that are exposed to harsh and wet environments

• Washing machines

Figure 15.

Example application use cases of waterproof buttons

See

Table 4

for typical requirements for the waterproof sensing use cases.

Force touch buttons

Inductive sensing-based buttons need deflection of (or movement of metallic) in the vicinity of the sensor coil, which means that pressure is required to turn the button ON. This feature is ideal for applications where buttons need to sense the amount of force being applied to their surfaces and shall not be activated accidentally. Different thresholds can be set to enable detection of different levels of force.

The use cases and applications are:

• Personal computer products (computers, tablets)

• Outdoor keypads (smart locks, security products)

• Headphones wear detection

• Power tools, buttons, and grip detection

Figure 16.

Example application use cases of force touch buttons/sensing

See

Table 4

for typical requirements of the force touch application use cases.

Proximity sensing and reed switch replacement

Inductive sensing exploits the fact that the effective inductance of a current-carrying coil decreases when a metal is in its vicinity. This property of the inductive sensing is ideal for detection of nearby metal which enables metal proximity sensing and reed switch replacement applications such as:

• Proximity sensors and reed switch replacements

  • Smart home (door locks, security systems, etc.)

  • Home appliances (washing machines, ovens, refrigerators, etc.)

  • Industrial applications

• Position detection/counting and rotary/shaft encoders

  • Motor applications

  • Rotary encoders/knobs

  • Product counting

Figure 17.

Example use cases of metal-proximity/reed switch replacement

See

Table 4

for typical requirements of the force touch application use cases.

Typical parameters for different use cases

The following table provides the typical requirements of the buttons for the different use cases.

Table 4.

Typical parameters for different use cases

Parameters	Touchovermetal (ToM)	Waterproof sensing	Force touch	Reed switch replacement
Power sources	Mains powered/Battery powered	Battery powered/Mains powered	Battery-powered	Mains-powered
Overlay	0.5 mm to 0.8 mm thick stainless steel, seamless	0.3 mm thick metallic (SS or Al), rubber, or plastic	0.3 mm thick metallic (SS or Al), rubber, or plastic	Target metal sensing through rubber or plastic (Al target is preferred)
Button size	Typically, 25 mm	Typically, 3.5 mm to 8 mm	Typically, 5 mm to 8 mm	Typically, 10 mm to 20 mm
Number of buttons/sensors	Typically, 4	Typically, 1 or 2	Typically, 12 for keypads and 2 for other applications	Typically, 1 or 2
Sensor - overlay air gap	3 mm to 4 mm	0.2 mm to 1 mm	0.2 mm to 1 mm	N/A
Button to button spacing	10 mm	5 mm	3 mm	>10 mm
PCB Layers	1 or 2	2 or 4	2 or 4	1 or 2
Activation force	3 Nm to 8 Nm	4 Nm to 5 Nm	5 Nm to 8 Nm for buttons	N/A
Other	Consumer grade EMI/EMC	Consumer grade EMI/EMC	Consumer grade EMI/EMC	Consumer grade EMI/EMC
Major challenges	Reliability, noise immunity, EMC	Low power, performance with small sensors, noise immunity (e.g., small motor, Bluetooth® LE), EMC	Low power, performance with small sensors, noise immunity (e.g., motor, Bluetooth® LE), EMC	Reliability, robustness, noise immunity (e.g., motor, Bluetooth® LE) EMC

Schematics and components selection

Schematics

Inductive sensing

Figure 18

illustrates typical schematics for flyback inductive button sensor.

Figure 18.

Schematic for flyback inductive sensor connection

Component selection

C_Vref selection

Figure 19.

Flyback inductive sensing schematics

C

Vref

represents the voltage reference capacitor for the sensor. It needs to provide a stable Vref value during the sensing operation. The required Vref is set by the PSOC™ device during initialization.

The recommended C

Vref

value is 220 nF.

Common C

Vref

configuration

A single common C

Vref

can essentially be connected multiple sensors in case the sensor properties are same/similar, for example, a four-button system, if all four sensor coils of the buttons are same inductance, a common C

Vref

can be used.

Dedicated C

Vref

configuration

If the sensor coils have different inductance values, because of sensor size, number of PCB layers etc., a dedicated C

Vref

is required for each sensor.

Sensor series resistance selection

Tx series resistance RS

TX

limits the sensor current and Rx series resistance RS

RX

limits the CVref (Cs) recharging current. The default recommended values are 560 Ω for both resistances.

Figure 20

shows the SNR for selection of sensor series resistance values.

Figure 20.

SNR vs LxClkDiv for different RS

RX

and RS

TX

resistors

External capacitors

MSCLP devices require an external capacitor for self-capacitance sensing. These external capacitors are connected between a dedicated GPIO pin and the ground. Recommended values of the external capacitors are as follows:

• C_MOD1 and C_MOD2 = 2.2 nF

External capacitors pin selection

The following table lists the recommended pins for C

MOD

capacitors for an ISX-based design.

Table 5.

Recommended pins for external capacitors

Device	C MOD1	C MOD2
PSOC™ 4000T	P4[2]	P4[3]
PSOC™ 4100T Plus	Channel0: P4[0]	Channel0: P4[1]
	Channel1: P7[0]	Channel1: P7[1]

To know more about pins that support external capacitors in PSOC™ devices, see the respective device datasheets.

Pin assignment

An effective method to reduce the interaction between sensor traces and communication or non-sensor traces is to isolate each by proper port/pin assignment.

Figure 21

shows an ideal pin assignment that allows such isolation.

As much as possible, the PSOC™ 4 controller should be oriented such that there is no crossing of switching signals such as communication and LED lines with the sensing traces.

Figure 21.

Recommended: Port isolation for communication, CAPSENSE™, and LEDs

ModusToolbox™ CAPSENSE™ Configurator can be used for the pin assignment in the ISX design as shown in

Figure 22

.

Figure 22.

PSOC™ pin assignment with ModusToolbox™ CAPSENSE™ Configurator

Tolerances/quality of discrete components

As ISX is a precision sensing technology, good tolerance components are recommended for the sensing circuit. Typically, 1% for resistors and 10% for capacitors are recommended.

Using connectors

In some use cases, use of the connector becomes necessary to connect sensors to PCB, for example, sensors can be designed on a flexible PCB and the PSOC™ device placed on the main PCB, in such scenarios a good quality sensor with low impedance and low contact resistance is recommended.

Note that the total trace length from PSOC™ device to sensor coil, including connectors should not exceed 100 mm.

Schematics and components checklist

This section lists the checks which should be done before finalizing the ISX design schematic. These recommendations are strongly recommended to optimize the performance.

The following table provides the checklist to verify your ISX schematic.

Table 6.

Schematics and components checklist

Category	Recommendations
C MOD	2.2 nF
C Vref	220 nF
Sensor series resistance on input lines	560 Ω resistance is recommended on Rx and Tx pins of the ISX sensor
Tolerances/quality of discrete components	1% tolerance for resistors. 10% tolerance for capacitors
Using connectors	If needed, use low impedance and low contact resistance connectors Trace length from the PSOC™ device to the sensor coil, including the connector should not exceed 100 mm
Sensor pin selection	If possible, avoid pins that are close to the GPIOs carrying switching/communication signals. Physically separate DC loads such as LEDs and I2C pins from the ISX pins by a full port wherever possible

Sensor design and layout guidelines

The sensor layout plays a crucial role in achieving the required sensitivity and reliability of the detection. An inductive proximity/touch sensor can be constructed in various shapes, sizes, and using different materials depending on the application requirement. The basic objective during sensor layout design is to:

• Achieve high sensitivity

• Reliable detection

• Maximum sensing distance

• Optimum sensor inductance and minimum parasitic capacitance (C_P)

Major types of sensors for the most common applications are:

• Button sensor

• Proximity sensor

• Slider

• Rotary encoder

Supported sensor parameter range

A sensor coil, with a target present, should be designed to be within the range listed as given in the following table.

Table 7.

Sensor design parameter supported range

Parameter	Minimum	Maximum
Effective sensor inductance	0.1 µH	200 µH
Series resistance	0 Ω	1000 Ω
Sensor parasitic capacitance	Check the following text	Check the following text

• The maximum acceptable parasitic capacitance depends on the sensor inductance.

Figure 23.

Maximum parasitic capacitance based on inductance value

$C_{p\_\max }\left(pF\right)=\frac{L\left(uH\right)}{4R_{{tx}_{total}}^2}\mathrm{.}{10}^6$

• For instance, R_txtotal= Rext + Rswitch = 560 Ω + 100 Ω and L(µH) = 20 µH

• Then in this case, the maximum acceptable parasitic capacitance is 10.81 pF for optimum performance.

• Higher parasitic capacitance (than the value given by the equation) can be supported at the cost of a reduction in the operating frequency. Note that in the R_txtotal calculation, include the series resistance of the inductor for larger coil sizes

• Note that the bare sensor, without the target (metal), can have much larger inductance than the sensor with the target

• In general, larger coils with higher inductance (or higher coupling of magnetic field lines) provide higher sensitivity

Inductive sensor design

A long PCB trace on an FR4 or a flexible printed circuit (FPC) board can form an inductive button/proximity sensor. The trace can be a spiral as shown in the following figure. This is the simplest form of inductive sensor.

Figure 24.

Simple inductive button sensor

Sensor parameters

For a given shape, the sensor coil is specified by the following parameters:

• n = The number of turns in a layer.

• nl = The number of layers

• w = The turn width.

• s = The turn spacing.

• Din = The inner diameter.

• Dout = The outer diameter. Usually, either Din or Dout needs to be specified; the other can be derived from other parameters.

• BtnSize = The size of the button, which is normally ≥ Dout

The sensor coil can span across two or four layers of the PCB to increase the value of inductance and sensitivity. For a simple button following parameters are recommended:

Table 8.

Layout recommendation for a simple ISX ToM button

Parameters	Recommendation
Button size	15 mm to 30 mm
Sensor coil size	12 mm to 23 mm (75% of button size)
Shape	Circular, rectangular, octagonal
PCB layers	2, 4
Trace width and spacing	5 mils
Number of turns	25-60
Overlay material	Stainless steel (preferred for longer lifespan and good material strength) Aluminum (slightly higher signal)
Overlay thickness	0.5 mm to 0.8 mm
Overlay – sensor coil spacing	0.3 mm to 4 mm
Recommended minimum SNR	10

Sensor shapes

The shape of the inductive sensor is important because it determines the shape of the generated magnetic field and therefore the change in inductance in the presence of a target metal object. The following are common shapes of PCB and flex coils:

• Circular coil: Circular coils are generally used when sensing a target object that is moving orthogonal to the sensor plane, as Figure 24 shows. The illustration also highlights the optimal plane of movement of the target for a circular sensor.

• Hexagonal or Octagonal coil: Polygonal coils are designed to approximate circular coils in cases where a circular coil is difficult to manufacture.

• Square coil: Square coil provides optimal performance with respect to sensitivity in both horizontal and vertical directions.

See

Figure 25

for an example of a hexagonal and square coil structure.

Figure 25.

Hexagonal coil and square coil examples

Rectangular coil:

Rectangular coils can be used to detect movement along a preferred axis.

Figure 26

shows an example of a rectangular coil with the axis for optimal detection of movement for this coil indicated in the diagram.

Figure 26.

Rectangular coil illustration showing the optimal plane of movement

Note:

Non-circular coils have a higher series AC resistance (R

S

) for the same inductance as of circular coil sensor.

For touch-over-metal type of applications, the circular coil sensor performs better than a rounded square coil for the same button size as shown with the data in

Figure 27

. Rounded square coils are only recommended if there are constraints on the PCB design that do not allow circular coils.

Figure 27.

Signal/delta L (nH) for two different coil shapes for 23 mm x 23 mm button

Sensor modeling

The inductance values L and series AC resistance, R

S

of the sensor need to be estimated for designing and the tuning of the sensors.

Sensor design – approximate equations

The following equations can be used to estimate the parameters of the inductor from the coil dimensions.

Note:

These expressions are not as accurate as a field solver and are used for approximations only.

The inductance of the coil can be estimated using the following expression derived from Simple Accurate Expressions for Planar Spiral Inductances

1

.

Figure 28

and is based on a Current Sheet Approximation.

Figure 28.

Inductance of the coil equation

${\mathit{L}}_{\mathit{g}\mathit{m}\mathit{d}}\mathit{=}\frac{{\mathit{\mu }}_{\mathit{0}}{\mathit{n}}^{\mathit{2}}{\mathit{D}}_{\mathit{a}\mathit{v}\mathit{g}}{\mathit{C}}_{\mathit{1}}}{\mathit{2}}\mathit{(}\mathit{ln}\mathit{}\mathit{\left(}\frac{{\mathit{C}}_{\mathit{2}}}{\mathit{\rho }}\mathit{\right)}\mathit{+}{\mathit{C}}_{\mathit{3}}\mathit{\rho }\mathit{+}{\mathit{C}}_{\mathit{4}}{\mathit{\rho }}^{\mathit{2}}\mathit{)}$

• C1, C2, C3, C4 = Layout constant dependent on the shape of the coil. See Table 9.

• $D_{avg}=0.5\cdot (D_{out}+D_{in})$

• ρ = The fill factor =

  $(D_{out}-D_{in})/(D_{out}+D_{in})$

• μ₀ = The permeability of free space, 4π x 10-7

• n = The number of turns

Table 9.

Coefficients for current sheet method

Coil shape	C1	C2	C3	C4
Square	1.27	2.07	0.18	0.13
Hexagonal	1.09	2.23	0.00	0.17
Circle	1.00	2.46	0.00	0.20

An expression to estimate the RS of a conductor due to the skin effect has been derived from High-Speed Digital Design: A Handbook of Black Magic

2

. This equation provides an estimate of the AC resistance of the inductor in ohms/inch.

Table 10

shows some values for ρR (the relative resistivity of the coil material, compared to copper).

Note that the value of RS calculated in the following equation is the unit value. To get the total value of RS, the unit value needs to be multiplied by the length of the trace which can be measured from the PCB Design tool by making a sample coil of desired shape.

Figure 29.

R

S

estimation equation

$R_s=\frac{2.16x{10}^{-7}\cdot \sqrt{f{\rho }_R}}{2(w+d)}$

Where,

• w = trace width, inches

• d = trace height, inches

• f0 = frequency, Hz

• ρ_R= relative resistivity, compared to copper = 1.00

• RS = AC resistance, ohms/inch

Table 10.

Conductivity of different materials compared to copper

Metal	Resistivity (× 10-7 ohms inch)	Resistivity relative to copper
Copper	6.7879 10	1.000
Aluminum	10.3606	1.526
304 SS	302.847	44.616
Titanium Alloy	667.29	98.306

Sensor design – 2D/3D EM solver

An electromagnetic (EM) field solver tool such as from COMSOL or Maxwell can be used to get a complete electrical model of the inductor and understand its characteristics.

Figure 30

shows an example of the coil modeling tool graphical outputs.

Figure 30.

Examples of output from an electromagnetic field solver (COMSOL)

Sensor size vs. sensing distance

The extent of the magnetic field is decided by the sensor physical dimensions; the outer diameter of the sensor, Dout, is the critical parameter. The ISX solution aims for sensing distances up to a coil diameter from the coil while maintaining the SNR at greater than or equal to 10:1. See Section

Signal-to-noise ratio (SNR)

for details.

For a button sensor with a metal overlay, the geometry of the inductive coil (circular or rectangular, for example) is one of the most important factors that affects the performance. For example, a circular button with 14-mm diameter will have more deflection of overlay than a button with 10-mm diameter.

Figure 31

shows the comparison of a normalized signal for two different button diameters.

Figure 31.

Impact of coil diameter on normalized signal

If the metal target is near the coil, a small ratio of Din/Dout is optimal for the coil to give maximum change in inductance with distance. When the metal target is further away from the coil (Dout), the outer windings contribute most to the electromagnetic field and the windings near the center can be removed.

Figure 32

shows the normalized change in inductance (ΔL/L) as the metal target is moved away from the coil at the distances Dout and Dout/4 with different coil Din/Dout ratios.

Figure 32.

Normalized inductance Change (ΔL/L) vs. Din/Dout ratio for a 13-mm diameter coil for proximity applications

Note:

This graph indicates ∆L/L for various loop sizes under lab conditions. The actual ∆L/L varies depending on the end-system environment.

Number of turns and PCB layers in a sensor

In general, the higher the number of turns results in coupling with the metal, more the sensitivity. As shown in

Figure 28

, the inductance is dependent on the number of turns. The additional number of turns that can be incorporated with an increase in coil diameter can be found as follows:

Figure 33.

Additional turns obtained with an increase in coil diameter

$Additionalturns=\frac{\mathrm{\Delta }Coildiameter}{Tracewidth+Tracethickness}$

The number of turns in an equivalent size coil on PCB can be increased by reducing the trace width and spacing between the traces, for example, a 3.5 mil trace with 3.5 mil spacing can accommodate a greater number of turns than 5 mil trace with 5 mil spacing within same coil area.

Similarly, more the number of PCB layers, higher the inductance and resulting in higher sensitivity given the size of the sensor coil is <= 10 mm. For big sensors, increasing the turns in the upper layers of the PCB, increases sensitivity, as it results in more coupling with the metal. Increasing turns in the lower layers cannot have the same impact as the layers are further away from the metal. The size of the sensor coil, i.e., the diameter of the sensor has bigger impact on the sensitivity of the sensor than the number of layers as shown in

Figure 34

and

Figure 35

.

In

Figure 34

, both scenarios can be considered as having an equivalent number of turns (20 turns in 1 layer, 10 turns each in 2 layers). The figure shows the higher sensitivity of the 16 mm coil on movement (deflection) of a metal target by around 0.3 mm at a 1 mm and 4 mm air gap.

Figure 34.

Absolute inductance response for 1-layer and 2-layers sensor coils

Figure 35

follows from

Figure 34

. It shows the change in inductance versus the deflection (movement) of the metal target. The solid line corresponds to the 16 mm coil which is more sensitive than the 8 mm coil.

Figure 35.

Relative inductance change for 1-layer and 2-layers sensor coils

Figure 36

shows the impact of coil stack-up for a 20 mm coil (23 mm button). The more the coupling with the metal overlay, the higher the sensitivity of the coil. The coils exists in different layers of a four-layer PCB. The coil in the top two layers performs better than the coil in one and four layers. The coil in all four layers of the PCB has the highest change in inductance (highest sensitivity) for similar deflection of metal.

Figure 36.

Signal (in terms of dL) for three different coil stack-ups

The data is captured for a 20 mm coil (23 mm x 23 mm button) with a 0.93 mm spacer/air gap and 0.5 mm thick stainless steel overlay.

To get higher sensitivity, design the PCB to have the coil in all four layers of a four-layer PCB rather than the standard two-layer PCB design.

Layout considerations

This section highlights the layout considerations to be considered made during the physical design of an inductive sensing solution.

In general, consider the following guidelines while routing signals in inductive sensing solution PCB:

• Ensure that the current always flows in the same direction when making multilayer coils, so that the magnetic flux always adds up. Figure 42 illustrates this

• When the coil is designed to the minimal trace and space rules and then connected to the via ring, be careful with possible violations (over-etching or under-etching process) here. It is recommended to expand the trace width at connection to the via and avoid sharp angles, see Figure 43

• Place the Tx and Rx lines parallel to each other

• Make sure that the Rx lines are not near any digital toggling lines like communication traces and switching LED traces. See Figure 40

• Make sure that the Rx lines do not use the communications line port, for instance, the I² C or SWD ports as they may interfere with the signal. See Figure 21

• Minimize the trace length whenever possible. Long traces can pick up more noise than short traces. Long traces also add to parasitic capacitance (C_P)

• For etched or milled overlays, the Rx and Tx traces should not overlap with metal foot of the target. It is good to hide these traces in the inner layers of PCB

• Coil to coil spacing must be maintained to ensure that there is no crosstalk. Around 0.5 * Dout for ToM use cases and greater for proximity kind of applications

• LEDs and discreet components must be kept at least 0.5 * Dout from the coil edge

Figure 37

shows examples of good and bad sensor design layout.

Figure 37.

Good sensor layout design patterns

Figure 38

shows the layout considerations for placement of bottom side of the sensor.

Figure 38.

Layout considerations for bottom side of the sensor

Figure 39.

Coil-to-coil spacing for a four-button use case

Route the sensor traces on the bottom layer of the PCB, so that the finger does not interact with the traces. Do not route traces directly under any sensor pad unless the trace is connected to that sensor.

Do not run traces too close to the switching signals or communication lines. Increasing the distance between the sensing traces and other signals increases the noise immunity. If it is necessary to cross communication lines with sensor pins, ensure that the intersection is at right angles as shown in the following figure.

Figure 40.

Routing of the sensor and communication lines

Sensors with multiple PCB layers

For a given diameter of the coil, there is a maximum limit to the number of turns due to the minimum allowable space between traces. If additional inductance is needed on a PCB, a multi-layer inductor can be designed using multiple layers of the PCB.

The key consideration for multi-layer inductors is that the current flow on different layers is in the same direction so that it constructively adds to the EM field and does not destructively take from it. Spiral coils in each layer are usually connected in series magnetically and electrically.

Figure 41

shows the connections for a coil in four layers in the PCB.

Figure 41. Different layers of the coil arranged to allow current flow to be in the same direction in each layer

Figure 42

shows two-layer inductors in series with a flexible metal target above then coils. The rainbow colors indicate the voltage distribution on the coil turns.

Figure 42.

Two-layer inductors in a series configuration with a flexible metal target

In addition to magnetic field interaction, the coil and the target have parasitic capacitive coupling. In case the target (metallic overlay) is grounded, the coil-to-target capacitance adds to the total parasitic coupling of the coil to the ground.

In case the target (metallic overlay) is not connected to the PSOC™ 4 device ground, the parasitic capacitance creates a path to the noise from the environment (for instance, a finger-injected noise)

When the target (metallic overlay) moves, it changes the parasitic capacitance, which may contribute to the signal measured by PSOC™ 4. This capacitive signal may add up or partially cancel the pure indictive signal. Parasitic capacitive signal should be reduced for confident measurement of inductive signal. It is done by connecting the Tx feed to the bottom layers and the top layer to the C

V

REF

(reference) point.

In this case the top layer will be at a fraction of the Tx potential. The outer turns of the coil will be at about zero potential, consequently the charge generated by parasitic capacitance variation will be minimized.

Such a design would also reduce the electromagnetic emission.

Figure 43.

Via between spiral coils in two layers

Figure 43

shows the recommended via placement in multi-PCB layer sensor implementation.

Sensor series resistor placement

As shown in schematics, series resistors are required on Rx and Tx signal lines of each sensor. These sensors are recommended to be placed as close to the PSOC™ device as possible since the traces after the resistor (towards PSOC™) are very sensitive.

Clearance between sensors and ground

Solid grounds near the sensors reduce the inductance. If needed, it is recommended to use broken hatched or comb ground planes surrounding the sensor, that exclude formation of the eddy currents as shown in

Figure 44

.

Figure 44.

Clearance between sensors and ground top side of PCB

Solid or hatch fill are strongly not recommended at the button area of sensor coil. If GND is needed for ESD protection add ground shape which does not affect the magnetic fields of the sensor coil as shown in

Figure 37

.

Trace clearances

Route the sensor traces as far from each other as possible, add ground trace in between if required.

Power and ground

• VDDIO noise could couple with the sensor through the IO, in case of excessive supply noise in the system, consider a regulator or a filter on the VDDD supply

• Make sure a low resistance ground plane to prevent any ground noise coupling into the sensor

LED placement and routing

When the LEDs are switched ON and OFF, the voltage transitions on the LED traces can create crosstalk in the inductive sensor lines, creating noisy sensor data. To prevent this crosstalk, the sensor lines and the LED traces must be isolated from each other as shown in

Figure 40

. LEDs and discreet components must be kept at least 0.5 * Dout from the coil edge for best performance.

LEDs can be mounted on the PCB and covered by the overlay with holes in it, or they can be separately exposed from the metal overlay. LED diffusers can be used for better electrical isolation by keeping the LEDs away from the coils.

The following figure shows the Keypad-4 and Keypad-2 expansion boards showcasing these two cases.

Figure 45.

Different LED placements for ISX solutions

Placement of LED at the center of the button is possible, but at the expense of inductive sensitivity. Button performance may degrade in such case.

Layout rule checklist

This section lists the checks which shall be done before finalizing the ISX design layout. These are strongly recommended to optimize the performance.

Table 11

provides the checklist to help verify your layout design.

Table 11.

Layout checklist

Category	Parameter	Recommendation
Coil	Shape	Circle or rectangular with curved edges
	Size	25 mm
	PCB layers	2 or 4
	Sensor routing	Must ensure that target/finger can only approach from top of the sensor (where the overlay is present) and not the bottom of the PCB. The top is least sensitive to the finger where the bottom is sensitive to the finger. If you touch to the top surface (blue curve) sensitivity to the finger is <1% from proximity signal From the bottom (orange curve) sensitivity strongly depends on the scanning frequency: The smaller the frequency the smaller the sensitivity Connect Tx feed to the bottom layers and the top layer to the C VREF (reference) point
Overlay	Type	Aluminum, Stainless steel, or other conductive material as recommended in Table 12 with spacer
	Thickness	0.25 mm– 0.5 mm (max.) thick, lower the thickness, better the deflection. Recommended deflection is 200 µm
Spacer	Type	Touch-over-metal use case overlay shall consist of nonmetallic spacer between coil and metal plate Non-conductive non-magnetic material with zero magnetic permittivity
	Thickness	0.1 mm – 5 mm (max.) thickness (based on sensor coil size) with cavities/openings over buttons to provide top metal overlay bending of 200 µm
Button	Cavity size	133% of the coil size
Series resistors	Placement	Place the resistor as close to the PSOC™ device pin, within 10 mm
C MOD	Placement	Place the resistor as close to the PSOC™ device pin
Sensor traces	Width	Max. 7 mil. Use the minimum width possible with the PCB technology that you use
	Length	Max. 300 mm for standard (FR4) PCB, 50 mm for flex PCB
Sensor traces	Clearance to ground and other traces	Maximize
	Routing	Route on the opposite side of the sensor layer. Isolate from other traces. If any non-ISX trace crosses the ISX trace, ensure that intersection is orthogonal. Do not use sharp turns
Trace vias	Number of vias	Recommended maximum 2 since at least one via is required to route the traces on the opposite side of the sensor layer
	Hole size	10 mils
Ground	–	No solid or hatch ground fill in the button area of the sensor coil. In case GND is at all needed for ESD protection, add ground shape which does not affect the magnetic field As shown in the schematics, series resistors are required on Rx and Tx signal lines of each sensor. These sensors are recommended to be placed as close to the PSOC™ device as possible, as the traces after the resistor (towards PSOC™) are very sensitive. For clearance between sensors and ground details, see Sensor series resistor placement
Other components	Placement	Do not place metal/conductive elements on the back side of the button close to coil. This decreases inductivity. See Layout considerations for more details

Mechanical construction guidelines

The mechanical construction of an inductive sensor involves various aspects that need to be considered together that include:

• Sensor coil geometry (size, shape, layers)

• Overlay material properties (elasticity, conductivity)

• Overlay thickness

• Button cavity size

• Air gap (sensor to overlay spacing)

These guidelines are recommended to improve the performance of the inductive sensing design. This section helps to understand how changing a few mechanical configurations will affect the overall performance of the ISX button.

Figure 46.

Mechanical construction of a touch-over-metal button

When designing an end application, it is recommended to use the Sensor Designer tool, which simplifies the process of designing flyback inductive sensors and the mechanical stack-ups. For details on the tool, see Section

Sensor Designer

.

Touch-over-metal involves detecting the deflection of a metal overlay upon a touch. It uses a metal overlay separated from the sensor using a thin spacer or a metal overlay with an etched cavity as shown in

Figure 47

. The encasing of the product could also act as an overlay, providing a seamless design. When the metal is pressed, it deflects, and this deflection is detected by the sensor.

The deflection of the metal overlay obtained must be relative to the PCB that has the coil. The PCB must have appropriate support such that it does not deflect when the metal overlay is being pressed.

Figure 47.

Arrangement of metal overlay for a typical touch button

The etched metal overlay has metal surrounding the cavity and can produce lesser deflection when pressed. This can result in reduced sensitivity compared to an overlay on top of a non-conductive spacer.

Air gap/spacing between sensor and overlay

The distance of separation between the sensor coil and the target is a major parameter that affects the performance of the ISX button.

As the air gap increases, the inductance of the coil also increases. This relative increase in inductance starts to flatten as the air gap approaches the diameter/diagonal of the sensor. For a given deflection, the measured change in inductance (dL signal) will decrease with increased distance.

Figure 48

compares L (absolute inductance) and dL (relative inductance) for 16 mm coils at air gaps of 1 mm and 4 mm.

Observe

Figure 48

, the sensitivity of a 16 mm coil is ~5x lower because of the larger air gap and lower diameter/air gap ratio. The dL is obtained by bringing the metal target closer to the coil (change in air gap).

Figure 48.

Air gap vs. absolute inductance (L) and relative inductance (dL)

The sizes of the button considered are 3 mm wider than the coil, i.e., 19 mm overlay for a 16 mm coil.

As shown in

Figure 49

, for the same deflection, the SNR drops as the air gap/spacer height increases while keeping other parameters constant. If SNR needs to be increased, one option is to decrease the air gap while keeping the rest of the configuration as is.

Figure 49. SNR vs. air gap/spacer height for a 20 mm coil (23 mm × 23 mm button) for two different stainless steel overlay thickness values

For example, in the configuration shown in

Figure 50

, a stainless steel overlay with 0.8 mm air gap works well with air gaps up to 2 mm. With an air gap of 3 mm, the SNR drops under 10.

Button cavity size

For optimum performance, the button cavity (overlay) size must be 133% of the coil/sensor size.

Figure 50.

Button (overlay) vs. sensor coil size

When the cavity size is larger, the deflection obtained in the metal for the same force is larger. Also, for thinner metals, the deflection is higher.

Figure 51.

Deflection values ( in µm) for different overlay thickness values and button (coil) sizes

Figure 51

shows the deflection data for different coil (and button) sizes. In

Figure 51

, as the cavity size increases, the deflection increases. Also, thinner overlay materials have higher deflection. The deflection required for a particular configuration is dependent on other parameters such as air gap, coil size, etc.

Overlay selection

In an ISX design, the overlay material and thickness play a key role in the deflection and inductance change produced. This determines the overall sensitivity of the button.

Overlay material selection

The overlay material is selected based on the application requirements by analyzing the mechanical, electrical, and magnetic properties and finding the right balance.

Materials such as stainless steel and aluminum can be used.

Stainless steel is more durable and has a longer life span. It can withstand more force than aluminum before deforming permanently due to its higher yield strength. This makes it ideal for touch-over-metal (ToM) kind of applications.

Aluminum has more conductivity and less Young’s modulus compared to other commonly used metals which makes it suitable for inductive sensing applications (ideal in case of proximity sensing/reed switch applications using Al as a metal target).

Overlay material mechanical properties

The amount of deflection for the given force at a distance depends on the elasticity of the target material. The Young’s modulus of the material is the measure of elasticity of the material. Materials with a lower Young’s modulus are more flexible. Copper, aluminum (AL6061-T6), and stainless steel (SS304) are the commonly available materials that can be used as the target materials. The following table shows the mechanical properties of different overlay materials. These are generic values; vendors can provide the exact specifications for the material used.

Table 12.

Properties of commonly used materials as overlay

Overlay material	Young’s modulus [Gpa] (Lower is better)	Elastic limit (yield strength) [MPa]
Copper	130	33–70
Aluminum (AL6061-T6)	68.9	20
Stainless steel (SS304)	200	520
Regular steel	200	100–1650
Phosphor bronze	110	200
Titanium	105	225–830
Ferrite	Not flexible	Not flexible

Overlay thickness is one of the factors that affect the deflection of the target for the applied force. The target with lesser thickness deflects more for the given force.

Beyond the elastic limit (yield strength), the permanent deformation of the material occurs. Therefore, the selected material should withstand the maximum pressure/force from expected finger size without deforming permanently.

The center-point deflection for thin membrane with fixed edges as explained in

3

and

4

can be defined by

Figure 52

as:

Figure 52.

Empirical formula for thin membrane with fixed edges

For metal sheet

$w_0=\frac{\alpha \cdot q\cdot b^4}{E\cdot t^3},whereq=\frac{F}{A}$

Where,

• w₀= center-point deflection (mm)

• t = Target sheet thickness (mm)

• b = Rectangle short side length

• E = Young’s modulus

• F = Force applied (N)

• A = Area of the button (m2)

• α = Coefficient equal to 0.0138 for a = b and fixed edges

Overlay material electrical properties

The change in the inductance of the sensor for a given target deflection depends on the conductivity of the target material as well. The material with higher conductivity has more induced eddy current on the surface of the target and it causes more change in the sensor inductance for the given deflection of the target.

The following table shows the electrical properties of different overlay materials. These are generic values and vendors can provide the exact specifications for the material used.

Table 13.

Properties of commonly used materials as overlay

Overlay material	Electrical conductivity (S/m)(Higher the better)
Copper	59.77x10 6
Aluminum (AL6061-T6)	36.9x10 6
Stainless steel (SS304)	1.67x10 6
Regular steel	2x10 6
Phosphor bronze	8.97x10 6
Titanium	2.4 x10 6
Ferrite	10 -5

Figure 53.

Electrical conductivity of metal as function of signal level and distance/diameter

Overlay material magnetic properties

Opposing effect of ferromagnetic and conductive materials

Magnetic flux at 1 MHz is rejected by conductive objects. The higher the conductivity and size of the object, the more the eddy current generated to reject the primary magnetic flux. Materials with magnetic permeability (mu) >1, especially ferromagnetic materials are known to concentrate the magnetic field and therefore, increase the inductance of the coil.

Ferromagnetic materials (with high magnetic permeability) will have an additive effect on the sensor coil inductance, which is not desirable in the current scheme. The effective signal from eddy current and the effect of high magnetic permeability will cancel each other at the operating frequency of MSCLP. This is why the response of regular steel will be worse than SS304. Materials with lower permeability are desirable.

When the target is ferromagnetic and conductive, such as nickel or steel, there are two opposing phenomena. The magnetic field is concentrated due to ferromagnetic properties and is being rejected because of the eddy currents.

Figure 54.

Properties of ferromagnetic materials as function of signal level and distance

Table 14.

Properties of commonly used materials as overlay

Overlay material	Relative magnetic permeability(Lower the better)
Copper	0.999994
Aluminum (AL6061-T6)	1.000022
Stainless steel (SS304)	1.01–1.05
Regular steel	200–1000
Phosphor bronze	1.005
Titanium	1.00005
Ferrite	10–20 000

Table 15

shows the magnetic properties of different overlay materials. These are generic values, vendors can provide the exact specifications for the material used.

Overlay thickness

The deflection of the metal used as an overlay depends on the material properties and metal thickness. Thinner metals are deflected more easily. The more the deflection, the higher the change in the effective inductance and the higher the signal obtained. Therefore, the sensitivity drops with an increase in the overlay thickness. SNR can be increased by decreasing the metal thickness while keeping other parameters constant. However, the metal should not be so thin as it gets completely deformed when pressed.

Figure 55

shows the data for different overlay thickness values of stainless steel. The graphs are for a 20 mm coil (23 mm × 23 mm button) in 1, 2 layers of a four-layer PCB with a 0.93 mm spacer/air gap between the metal and the PCB. The three curves represent different overlay thickness values. The X-axis for all three curves show the force with which the button is being pressed.

Note:

The setup used for capturing this data has been optimally tuned for the 0.5 mm overlay thickness.

Figure 55. SNR, deflection (µm), and signal (nH) data for different overlay thicknesses of stainless steel

Observe the SNR curve in

Figure 55

, for a 0.93 mm air gap and 20 mm coil, the SNR is sufficiently high (>10) for all three overlay thickness values: 0.3 mm, 0.5 mm, and 0.8 mm. Additionally, the deflection and signal curves show clearly that thinner materials are deflected more for the same pressing force which in turn results in a larger signal (or delta L).

Overlay warpage

In touch over metal buttons, consistent usage of the buttons can cause overlay warpage over time which in turn affects the button performance. The following strategies can help mitigate this behavior.

• Having larger buttons

• Choosing the overlay material such that its maximum elasticity is not exceeded when subjected to button presses.

• Construction of cavity in such a way, that it prevents the excessive deformation of the overlay when high force is applied.

Grounded or floating overlay

In inductive sensing applications, the metal overlay can be connected to ground or kept floating based on the system conditions. A grounded overlay can provide ESD immunity and will have lesser parasitic capacitive coupling with a hand/finger nearby as compared to a floating overlay.

Flyback-based inductive sensing works with both floating and grounded metal overlays. In the LC resonance-based method the capacitive coupling can cause false triggers.

Isolation of buttons

In a multi-button system, the mechanical disturbance from the adjacent button can cause a false trigger on the given button. To minimize the mechanical disturbance caused by the adjacent buttons, it is particularly important to provide proper isolation between the buttons.

Also, see the section

Mechanical crosstalk

.

The cross sensitivity is dependent on the factors as given in the following table.

Table 15.

Effect of materials used in system on overlay lift-off/deflection crosstalk

Parameter	Relationship/effect	Recommendation
Overlay thickness	Higher the thickness, more the movement/lift-off of overlay on nearby buttons	Thin overlay is recommended. However, ensure that the overlay is not so thin as it gets deformed permanently
Spacer material	Elastic spacer material will allow the overlay to move in a wider region than the button being pressed.	Use rigid overlay which will limit the movement of overlay on adjacent buttons
Adhesive material/thickness	Elastic adhesive material will allow the overlay to move in a wider region than the button being pressed, weak adhesive will have similar effect	Use rigid and firm adhesive which will limit the movement of overlay on adjacent buttons. 3M acrylic adhesives provide strong and consistent bonding. Thinner adhesive layers can be used when there is minimum surface irregularity
Grooves on overlay	Grooves on overlay/corrugated overlay help add flexibility to the overlay making it possible to deflect the required area with smaller pressure and avoid impact to adjacent buttons/area	Use grooves on the button borderline as deep as possible
Coil to coil spacing	The coil spacing determines the crosstalk between adjacent coils	Coil to coil spacing must be 0.5 * Dout or more. For spacing between two coils of different sizes, consider the Dout of the larger coil

An example 4-button system is shown in

Figure 56

.

Figure 56.

4-button system

The cross-sensitivity matrix of buttons for the given multi-button system can be used to tune the finger threshold of a given button in firmware to avoid the false trigger caused by adjacent buttons. The cross-sensitivity matrix as shown in

Table 16

provides the probability of the given button to get false triggered by the adjacent buttons.

Figure 57.

Probability of a button false triggered by an adjacent button

$\mathrm{Pr}\mathrm{o}\mathrm{b}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}\mathrm{}\mathrm{o}\mathrm{f}\mathrm{}\mathrm{a}\mathrm{}\mathrm{b}\mathrm{u}\mathrm{t}\mathrm{t}\mathrm{o}\mathrm{n}\mathrm{}\mathrm{f}\mathrm{a}\mathrm{l}\mathrm{s}\mathrm{e}\mathrm{}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{g}\mathrm{g}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{d}\mathrm{}\mathrm{b}\mathrm{y}\mathrm{}\mathrm{a}\mathrm{d}\mathrm{j}\mathrm{a}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{}\mathrm{b}\mathrm{u}\mathrm{t}\mathrm{t}\mathrm{o}\mathrm{n}=\frac{Interferencesignalfromadjacentbutton}{Fingerthresholdofgivenbutton}$

Table 16. Cross-sensitivity matrix for a 4-button system with 0.5 mm overlay thickness, 0.25 mm spacer thickness, and 14 mm sensor outer diameter and ~7 mm spacing between buttons

Button	By BTN1	By BTN2	By BTN3	By BTN4
False trigger probability of BTN1	–	0.33	0.18	0
False trigger probability of BTN2	0.58	–	0	0.4
False trigger probability of BTN3	0.19	0	–	0.36
False trigger probability of BTN4	0	0.36	0.58	–

The following techniques can be used to provide proper mechanical isolation between the buttons:

• A groove around the button.

• Use of a thin metal target/overlay.

• Separate adjacent coils at least by a distance more than half the coil diameter.

• Proper tuning (see the Tuning the inductive-sensing solution section for more information).

System design considerations

When designing inductive sensing technology in an application, it is important to remember that the ISX system exists within a larger framework. Careful attention to every level of detail from PCB layout to user interface to end-use operating environment will enable robust and reliable system performance.

EMI/EMC considerations

Series resistors on digital communications lines

Communication lines, such as I2C and SPI, can have long traces that act as antennae. They benefit from the addition of series resistance; 330 Ω is the recommended value as shown in

Figure 58

. The recommended pull-up resistor value for I2C communication lines is 4.7 kΩ. If a series resistor of value greater than 330 Ω is placed in series with these lines, the VIL and VIH voltage levels may fall out of specifications, while 330 Ω will not affect I

2

C operation as the VIL level remains within the I2C specification limit of 0.3 VDD when PSOC™ outputs a “LOW”.

Figure 58.

Series resistors on communication lines

Current loops

Another important layout consideration is to minimize the return path for currents. This is important as the current flows in loops. Unless there is a proper return path for high-speed signals, the return current will flow through a longer return path forming a larger loop, thus leading to increased emissions and interference.

If you isolate any ground hatch and ground fill around the device, the sensor-switching current may take a longer return path. The sensors are switched at a high frequency, so the long return current may cause EMC issues. It is recommended to use a single ground fill to minimize the length of such a return path and any subsequent emissions.

Use of spread spectrum sense clock

Since the ISX technique does not use resonance frequency operation, a spread spectrum excitation frequency can be used for the sensor. This reduces the effective emission compared to resonance frequency inductive sensing techniques. The following figure shows a comparison of emissions between resonance frequency inductive sensing techniques and ISX technique.

Figure 59.

EMI for typical inductive sensor with constant and SS frequency

Firmware design guidelines

After the proximity sensor layout is ready, the next step is to implement the firmware. This section describes the guidelines for the firmware and configuration of the ISX system parameters.

To learn more about using ModusToolbox™ to create a CAPSENSE™ project, see the

PSOC™ 4: MSCLP CAPSENSE™ low-power proximity tuning

code example.

Configuring the type of widgets

PSOC™ devices with fifth-generation MSCLP offer two types of widgets configurations: active and low-power widgets.

Active widgets

Active widgets are the regular widgets scanned during the active mode of the CPU and cannot be scanned during the sleep mode. All the widget types available in CAPSENSE™ Configurator are by default active widget types, except for the low-power widgets offered by the fifth-generation CAPSENSE™ devices.

Low-power widgets

PSOC™ devices offer multiple power modes such as Active, Sleep, Deep Sleep, etc. Devices with fifth-generation CAPSENSE™ offer the capability to keep the CAPSENSE™ peripheral active even in Deep Sleep mode. This capability can be implemented using a widget type called “Low-Power Widget”. These widgets are scanned and processed without any CPU intervention in Deep Sleep mode; they can act as a wake-up source for the device.

Widgets such as buttons, ganged sensors, and proximity sensors can be configured as low-power widgets (see

Figure 60

) to enable wake-on-touch/wake-on-approach functionality.

Note:

If a widget is configured as a low-power widget for normal operation in active mode, they need to be additionally configured as an active widget as well.

Figure 60.

Adding a widget using CAPSENSE™ Configurator in ModusToolbox™

Adding ISX widgets to the design

Adding the ISX button

The ISX widget can be added to the design as shown in the following figure.

Figure 61.

Adding ISX button to the design

Scan configuration

Scan configuration defines the distribution of the sensors, make ganged connection, pin assignments, and scan slots assignments for each sensor channel.

Figure 62.

CAPSENSE™ Configurator - Scan Configuration

Channel and slots for the sensors can be assigned using the

Scan Configuration

tab in CAPSENSE™ Configurator. See the

CAPSENSE™ Configurator user guide

for more details.

Force-touch (detecting different levels of force)

In some applications, it is desirable to have detection of different levels of force. In such cases, the sensor can be tuned to detect the minimum amount of force that needs to be detected. After tuning, thresholds can be set to detect different force levels by observing the signal levels corresponding to each force level (

Figure 64

shows the different signals corresponding to force levels). Custom thresholds can be set to detect 5 N and 6 N forces in the firmware. The finger threshold parameter in the CAPSENSE™ configuration is set according to the signal obtained by the 3 N force (the minimum force expected to be detected in this case).

Figure 63.

Signal levels corresponding to different force levels

Filters

Proximity sensors are susceptible to noise because of their large sensor area and high sensitivity setting. Filters help to greatly reduce the noise in the raw count, therefore, improving the SNR and the response time. A higher SNR implies a large proximity-sensing distance and reliable sensing.

The CAPSENSE™ ecosystem features the following types of filters:

• Hardware filters

• Software filters

Based on what these filters are applied for. They are further classified as:

• Raw count filters

• Baseline filters

As the baseline is used as a reference for calculating the signal, it needs to be more stable than the raw count. Therefore, a separate filter needs to be applied for the baseline than that of the raw count.

Additionally, as the CAPSENSE™ widgets can be configured as active or low power, the ecosystem offers separate filters for these types of widgets.

Table 17

shows an overview of the filters offered by the CAPSENSE™ ecosystem.

Table 17.

Filters offered by CAPSENSE™ ecosystem

Implementation	Filter application	Filter type	Active widget	Low-power widget
Hardware	Raw count filter	CIC2	Yes	Yes
		HW-IIR	Yes	Yes
Software	Raw count filter	SW-IIR	Yes	–
		Average	Yes	–
		Median	Yes	–
	Baseline filter	SW-IIR	Yes	–
		SW-IIR fast	–	Yes
		SW-IIR slow	–	Yes

The following sections describe how to use and configure these filters effectively for proximity-sensing applications.

Hardware filters

The PSOC™ devices with fifth-generation MSCLP architecture feature two types of hardware raw count filters:

• Cascaded integrator-comb 2 (CIC2) filter

• First-order hardware IIR filter

This section describes these filters and their configuration in more detail.

Cascaded integrator-comb 2 (CIC2) filter

It is a second-order digital low-pass (decimation) filter for delta-sigma converters. It provides a higher resolution and then the SNR for a given scan period.

The CIC2 filter receives the output of the CAPSENSE™ analog front-end, which is a delta-sigma convertor. This filter is a combination of a moving average low-pass filter and a down-sampler. This filter is also known as a decimator. It provides a significant improvement in SNR when used along with an IIR filter.

Enabling CIC2 filter

The following figure shows how the CIC2 filter can be enabled in the

Advanced

tab of CAPSENSE™ Configurator in ModusToolbox™.

Figure 64.

Enabling CIC2 filter

Table 18

describes the parameter that can be tuned for CIC2 filter.

Table 18.

CIC2 Parameters to be tuned

Parameter	Description
CIC2 accumulator shift	The CIC 2 accumulator shift sets the right shift value applied to the CIC2 accumulator to form raw counts when the CIC2 filter is enabled.
Decimation rate	Sets the configurable decimation rate when the CIC2 filter is enabled.

See the

AN85951 – PSOC™ 4 and PSOC™ 6 MCU CAPSENSE™ design guide

for more details.

First-order hardware IIR filter

This filter is used to filter raw counts of sensors (both

Active widgets

and

Low-power widgets

). The input to this filter is:

• Output of CIC2 filter when CIC2 is enabled

• Output of the CAPSENSE™ analog front-end when the CIC2 filter is disabled

Figure 65

represents the instantaneous output raw count of the filter:

Figure 65.

Hardware IIR filter raw count equation

$\mathit{R}\mathit{a}\mathit{w}\mathit{C}\mathit{o}\mathit{u}\mathit{n}\mathit{t}\mathit{=}\frac{\mathit{1}}{{\mathit{2}}^{\mathit{i}\mathit{i}\mathit{r}\mathit{R}\mathit{C}\mathit{c}\mathit{o}\mathit{e}\mathit{f}}}{\mathit{R}\mathit{a}\mathit{w}\mathit{C}\mathit{o}\mathit{u}\mathit{n}\mathit{t}}_{\mathit{C}\mathit{u}\mathit{r}\mathit{r}\mathit{e}\mathit{n}\mathit{t}}\mathit{+}\mathit{(}\mathit{1}\mathit{-}\frac{\mathit{1}}{{\mathit{2}}^{\mathit{i}\mathit{i}\mathit{r}\mathit{R}\mathit{C}\mathit{c}\mathit{o}\mathit{e}\mathit{f}}}\mathit{)}{\mathit{R}\mathit{a}\mathit{w}\mathit{C}\mathit{o}\mathit{u}\mathit{n}\mathit{t}}_{\mathit{Pr}\mathit{e}\mathit{v}\mathit{i}\mathit{o}\mathit{u}\mathit{s}}$

Where,

iirRCcoef = IIR filter raw count coefficient.

The valid range is 1 to 8; a low coefficient means lower filtering, but a faster response time

Enabling hardware IIR filter

The following figure shows how the hardware IIR filter can be enabled in the

Advanced

tab of CAPSENSE™ Configurator in ModusToolbox™.

Figure 66.

Enabling the hardware IIR filter

Software filters

The CAPSENSE™ ecosystem’s middleware library offers the following types of software filters:

• Raw count filters

  • Average

  • First-order IIR

  • Median

• Baseline filters

  • First-order IIR for active widgets

  • Slow baseline IIR filter for Low power widgets

  • Fast baseline IIR filter for Low power widgets

For more information on the details of software filters, see the “Software filtering” section of the

Getting started with CAPSENSE™

.

Software IIR filter

All the software IIR filters offered by CAPSENSE™ middleware are implemented with the following equation:

Figure 67.

Software IIR filter equation

$Output=(\frac{Coefficient}{256}\times input)+(\frac{256-Coefficient}{256}\times previousoutput)$

Adding software filters to the project

The required software filters can be added to the ModusToolbox™ project from the

Advanced

tab of CAPSENSE™ Configurator as shown in the following figure.

Figure 68.

Adding software filters to the project

Software raw count filters

Three types of raw count filters are offered by CAPSENSE™ middleware which can be configured using the CAPSENSE™ Configurator tool.

Table 19

describes these filters, their applications, and respective tunable parameters.

Table 19.

Software raw count filters offered by CAPSENSE™

Type	Description	Application	Tuning parameter
Average	Simple moving average filter of 4 samples length. Consumes 6 bytes of RAM per each sensor to store a previous raw count (filter history)	Periodic noise (e.g., noise from AC mains)	None
IIR	Similar to a simple RC filter, a software IIR filter with tunable IIR co-efficient. Consumes 2 bytes of RAM per each sensor to store a previous raw count (filter history)	High-frequency white noise (1/f noise)	IIR co-efficient – Lower the co-efficient, higher filtering effect; The valid range: 1‑128
Median	Similar to a moving average filter, a nonlinear filter that computes the median input value from a buffer of size 3 Consumes 4 bytes of RAM per each sensor to store a previous raw count (filter history)	Noise spikes from motors and switching power supplies	None

If multiple filters are enabled, the execution order is as follows:

1. Median filter

2. IIR filter

3. Average filter

Software baseline filters

As the baseline is used as a reference for calculating the signal, it needs to be more stable than the raw count. Therefore, separate filters are offered in CAPSENSE™ middleware, which can be configured using the CAPSENSE™ Configurator tool.

Table 20.

Software baseline filters offered by CAPSENSE™ ecosystem

Type	Description	Application	Tuning parameter
First-order IIR for active widgets	A software IIR filter with tunable IIR co-efficient	High-frequency noise	IIR co-efficient – Lower the co-efficient, higher filtering effect: The valid range: 1–255
Slow baseline filter	IIR filter coefficient for slow changing baseline (Low power widget only)	High-frequency white noise	IIR co-efficient – Lower the co-efficient, higher filtering effect; The valid range: 1–15
Fast baseline filter	IIR filter coefficient for fast changing baseline (Low power widget only)	High-frequency white noise

Middleware library and APIs

Infineon provides a middleware library for CAPSENSE™ products and solutions which can be used to implement the features mentioned in the

Firmware design guidelines

section.

See the CAPSENSE™ middleware library help document available in the ModusToolbox™ Quick panel as shown in the following figure.

Figure 69. CAPSENSE™ middleware library help document in the ModusToolbox™

The PSOC™ 4: MSCLP ISX keypad-4 demo code example available on GitHub can serve as an excellent starting point to learn how to use Infineon’s CAPSENSE™ middleware library and respective APIs.

Tuning the inductive-sensing solution

After the sensor layout and firmware are ready, the next step is to tune the parameters for the ISX sensor to achieve optimum performance.

The change in inductance caused by a target object at a distance from the sensor is in tens of nH. To detect such a small change, the ISX circuitry must be tuned for high sensitivity, and the threshold parameters should be set to the optimum values. The process of setting these parameters for an optimum sensor performance is called “tuning”. This section describes how to tune the parameters in ModusToolbox™ for a sensor to achieve optimum performance.

This tuning guide is created based on existing knowledge and can be revised after acquiring new information on the flyback inductive sensing method.

Signal-to-noise ratio (SNR)

Before tuning the capacitive sensor for optimal sensing, it is important to understand the concept of SNR and how to calculate it.

As a thumb rule, an SNR of ≥ 10:1 is required for reliable sensing in inductive-sensing use cases. Higher the SNR, the higher the sensitivity that can be achieved.

Calculating SNR

The first step in measuring SNR is to monitor the raw count for each sensor with OFF and ON scenarios.

• OFF scenario is where the buttons is not pressed, or the metal target object is not present in the proximity of the sensor.

• ON scenario is where the button is pressed, or the metal target object is present, and this is detected by the sensor. At least 3000 samples each are recommended to be logged to measure the SNR.

Another factor to consider is how the signal is produced. The worst-case ON and OFF scenarios should be used when measuring SNR.

For a touch-over-metal (ToM) button, SNR can be calculated for the least amount of force (deflecting the metal overlay) that is expected to trigger the button. If the system is designed to sense the presence of a metallic target in proximity, then measure the SNR with the presence of the target at the farthest required distance from the sensor.

Figure 70.

Use a force gauge (also called force meter) to deflect the metal target with the minimum required force while measuring SNR

Figure 71.

ISX sensor signal and noise

As an example of measuring SNR, consider the raw count waveform in

Figure 71

, with the calculations mentioned in the following table:

Table 21.

Parameters to calculate SNR

Parameters	Value
Minimum raw count without a target in the proximity	2850 counts
Maximum raw count without a target in the proximity	2880 counts
Average raw count without a target in the proximity (baseline)	2865 counts
Average raw count with a target in the proximity	3110 counts
Diff Count (Signal)	3110 – 2865 = 245 counts
Noise count	2880 – 2850 = 30 counts
SNR	245:30 = 8.2:1

Using ModusToolbox™ for tuning and SNR measurement

See the PSOC™ 4: MSCLP inductive sensing touch-over-metal keypad-4 demo and PSOC™ 4: MSCLP inductive sensing touch-over-metal keypad-2 code example for details on tuning and measuring SNR using the CAPSENSE™ Configurator and CAPSENSE™ Tuner tools.

ISX SmartSense (Beta version)

ISX SmartSense helps you to get the basic tuning parameters and acts as a starting point for tuning a particular configuration.

SmartSense has two modes:

Full auto-tune:

In this mode, the middleware sets both the hardware and the threshold parameters. This mode is not available for low-power widgets.

Hardware parameters:

In this mode, the middleware sets the hardware parameters only and you must set the threshold parameters.

To enable SmartSense, change the tuning method in the CAPSENSE™ Configurator as shown in the figure below. Touch sensitivity is the minimum sensitivity for which the configuration is tuned.

Figure 72.

Enabling SmartSense in the CAPSENSE™ configurator

For more information regarding SmartSense, see the

PSOC™ 4: MSCLP inductive sensing touch-over-metal keypad-4 demo

code example.

Manual tuning flow

The tuning procedure for the inductive sensors can be explained in the following stages:

1. Set the initial parameters.

2. Measure the sensor inductance and set the sense clock frequency (Lx).

3. Set CDAC and dither parameters.

4. Fine-tune for the required SNR, power, and refresh rate.

5. Tune the threshold parameters.

Figure 73

shows the high-level steps for tuning an ISX sensor.

Figure 73.

High-level steps for tuning an inductive sensor

Parameters to be tuned

CAPSENSE™ ecosystem provides great flexibility to support different types of CAPSENSE™ touch and proximity sensing applications. The following sections list proximity sensing parameters available in ModusToolbox™ along with their recommended values.

Set initial parameters

• Lx clock divider: 200

• Clock source: Direct (default)

• Number of sub-conversions: 300

• Other parameters leave as default

• ISX raw count calibration percentage to 40%

Setting the Lx clock divider

The Lx clock divider should be configured such that the pulse width of the sense clock is long enough to allow the sensor inductance to accumulate its energy completely while preventing prolonged charging which will waste scan time.

1. Find the expected or approximate value of the Lx clock divider using the following equation.

   Figure 74. Lx clock divider approximation

   $$\begin{gathered} LxClockD\mathit{i}\mathit{v}\mathit{=}\frac{\mathit{N}\mathit{t}\mathit{a}\mathit{u}\mathit{}\mathit{\times }\mathit{N}\mathit{p}\mathit{h}\mathit{a}\mathit{s}\mathit{e}\mathit{s}\mathit{\times }\mathit{F}\mathit{}\mathit{mod}\mathit{}\mathit{\left(}\mathit{M}\mathit{h}\mathit{z}\mathit{\right)}\mathit{\times }\mathit{L}\mathit{\left(}\mathit{u}\mathit{H}\mathit{\right)}}{\mathit{Re}\mathit{x}\mathit{t}\mathit{+}\mathit{R}\mathit{s}\mathit{w}\mathit{i}\mathit{t}\mathit{c}\mathit{h}\mathit{+}\mathit{R}\mathit{i}\mathit{n}\mathit{d}\mathit{u}\mathit{c}\mathit{t}\mathit{o}\mathit{r}}\mathit{+}\mathit{S}\mathit{e}\mathit{t}\mathit{t}\mathit{l}\mathit{i}\mathit{n}\mathit{g}\mathit{}\mathit{S}\mathit{h}\mathit{i}\mathit{f}\mathit{t} \\ LxClockDiv=\frac{3\times 4\times 46\times L\left(uH\right)}{(560+100)}+8 \\ LxClockDiv=4\times round\left(\frac{27.6}{4}\right) \\ LxClockDiv=28 \\ (ifthisvalueislessthan8,settheLxClkDivto8) \end{gathered}$$

   Where,

   • Lx clock divider: Approximate value is obtained (Should be divisible by 4 to ensure all four scan phases have equal durations)

   • Ntau: Settling constant. Set to 3 by default

   • Nphases: Number of scan phases. Set to 4

   • Fmod (MHz): Modulator clock frequency in MHz. In this example, the frequency is 46 MHz

   • L(µH): Obtain the inductance value using an LCR meter. For the buttons in this example, the value is around 34 µH

   • Rext: Tx resistor in series. 560 ohms in the kit

   • Rswitch: Internal switch resistance. The default value can be set as 100 ohms

   • Rinductor: This value can be obtained using the LCR meter while obtaining the inductance value. Include when the sensor is larger and can contribute to the total resistance value significantly

   • Settling shift: Provide a shift to consider actual settling time. The default value is 8. Change this based on oscilloscope observations

   Note:

   This equation serves solely as the starting point for finding the actual Lx clock divider value using the oscilloscope as explained in further sections. It is not recommended to set the Lx clock divider exactly as given by this equation.

   Set this Lx clock divider value in the CAPSENSE™ configurator and program the device

2. Probe the Rstx resistor and observe the waveforms at both CH4 and CH3 as shown in the following figure

   Figure 75.

   Oscilloscope probe setup to measure Lx clock divider

   
   
   

3. Perform a math function on the oscilloscope to calculate the approximate current passing through the resistor

   Figure 76.

   M1 waveform equation

   $M1=\frac{Ch4-Ch3}{Rstx}\sim currentthroughtheRstxresistor$

   • If the charging/discharging phases are too short, increase the Lx clock divider in steps of 4 until the sensor is charging/discharging completely in both phases. See the following figure for more information

   Figure 77.

   Improper charge cycle of a sensor with incomplete charging/discharging

   
   
   

   • If the charging/discharging phases are prolonged, start decreasing the Lx clock divider value by 4 such that current through Rstx becomes constant (

     Figure 78

     ). It means that the inductor has accumulated its energy (transient processes are finished)

   Figure 78.

   Improper charge cycle of a sensor showing prolonged charging/discharging

   
   
   

   The following figure shows the cases where both conditions are met such that the inductor charges/discharges completely however, it does not have prolonged cycles.

   Figure 79.

   Proper charge cycle of a sensor

   
   
   

   As the Tx phase width selection step is limited to the Fmod period, it might not be possible to select the optimal value of LxClkDiv. In this case, choose a divider value closer to the start of constant current.

   When there are multiple sensors in a widget, choose the maximum LxClkDiv value obtained among these sensors.

   When unclear, it is better to increase the LxClkDiv and have a slightly prolonged charging/discharging cycle as compared to incomplete charging/discharging.

Setting RefCDAC and CompCDAC

The reference and compensation CAPDAC values can be set to auto mode in the configurator as shown in the following figure. The middleware will calculate the values based on the raw count calibration percentage set in the CAPSENSE™ Configurator. It is recommended that the compensation CDAC is always enabled.

Figure 80.

Setting CAPDAC values to Auto in the CAPSENSE™ configurator

Note:

To increase sensitivity, RefCDAC boost can be enabled.

Number of sub-conversions

Calculate the number of sub-conversions and resolution

Use the following equation to find the Nsub based on the scan time and the LxClkDiv obtained previously.

Figure 81.

N

SUB

equation

$NumOfSubConv=floor\left(\frac{ScanTime\left(us\right)\times F\mathrm{mod}\left(MHz\right)}{LxClkDiv}\right)$

i.e., LxClkDiv = 24 , Fmod = 46 MHz and Scan Time = 800 µs : N

SUB

= 1533

Alternatively, Nsub can be set based on the SNR requirements. Increase the Nsub to obtain higher SNR, which will come at the cost of an increase in scan time.

Calculate resolution (max raw count) with the following equation

Figure 82.

Resolution (max raw count) equation

$NumOfSubConv=floor\left(\frac{ScanTime\left(us\right)\times F\mathrm{mod}\left(MHz\right)}{LxClkDiv}\right)$

i.e., for scan time equal to 800 µs and Fmod equal to 46 MHz: 36800 raw counts

Note:

The resolution should not be greater than 65535 (maximum raw count value supported (16 bits, *CIC2 disabled)). Therefore, maximum scan time is limited. i.e., for Fmod = 46 MHz, maximum scan time is 1424 µs. If longer scan time is needed – sample averaging can be used.

Set the obtained Lx clock divider and number of sub-conversions in the CAPSENSE™ Tuner.

Setting the CDAC dither scale

As the input inductance changes, the raw counts should change linearly. However, there are regions where the inductance does not change linearly, which are called flat-spots. Dithering helps reduce these flat-spots by adding white noise that moves the conversion point around the flat region. The CDAC dither value can be set based on the following table.

Table 22.

CDAC dither scale recommendation

Sensor inductance range	Dither scale value
0 nH to 140 nH	6
140 nH to 232 nH	5
232 nH to 465 nH	4
465 nH to 1.4 µH	3
1.4 µH to 2.3 µH	2
2.3 µH to 4.6 µH	1
>4.6 µH	0

Calculate init sub-conversions

To calculate the number of init sub-conversions, set this value in the CAPSENSE™ Configurator. Build and reprogram the chip if this value is different from the initial value (3):

Figure 83.

Recalculating N_SUB

$Numberofinitsubconversions=ceiling\left(\frac{C_{\mathrm{mod}}\times V_{OS}}{VDDA\times LxClkDiv\times C_{ref}\times (1-Bal\%)}\right)+1$

Where,

CMOD = Modulator capacitor

VOS = Comparator offset voltage (3 mV)

Bal% = Raw count calibration percentage

LxClkDiv = Lx clock divider

Cref = Reference capacitance

Cref = RefCDACCode * Clsb

RefCDACCode = Reference CDAC value

Clsb = 8.86 fF

Figure 84.

Example of N_SUB calculation

$e\mathrm{.}g\mathrm{.},Numberofinitsubconversions=ceiling\left(\frac{2.2\times {10}^{-9}\times 3\times {10}^{-3}}{3.3\times 24\times 60\times 8.86\times {10}^{-15}\times (1-0.1)}\right)+1=2$

Post tuning (filters)

1. Enable filters (CAPSENSE™ Configurator, a reprogram is required):

   1. Enable hardware IIR filter to eliminate high-frequency noise. Set the filter coefficient value to 1. Higher values of coefficient provide better noise suppression but also slow down the touch response and need to be adjusted based on the requirements

   2. Enable average filter (moving average) to eliminate periodic noise (i.e., noise from AC mains), if present

   3. Enable median filter to eliminate spike noise (i.e., motors and switching power supplies), if present

   4. Enable CIC2 filter to increase resolution (recommended to enable this by default). Set decimation rate to the maximal allowable value and leave CIC2 accumulator shift value to Auto (widget hardware parameters)

2. Measure SNR using the SNR Measurement tool in CAPSENSE™Tuner to verify filters efficiency

Parameters: Descriptions and recommendations table

General tab parameters

Figure 85.

General tab

Table 23.

Scan settings parameters

Parameter	Description	Recommended values
IMO clock frequency	Frequency of clock used as source for the CAPSENSE™ peripheral	Keep it default i.e., 46 MHz (PSOC™ 4000T,PSOC™ 4100T Plus)
Modulator clock divider	Divider value used to divide system clock; the resulting clock then used as clock for Inductance to digital modulation	Use a slower modulator clock to reduce peak-to-peak noise in raw counts if required
Number of init sub-conversions	Selects the number of initialization sub-conversions at the start of the scan. This part of scan is intended to ensure proper initialization of MSCLP hardware and does not perform the raw count measurement	This parameter should be set after complete tuning. Initial value can be set to 3
Enable CIC2 hardware filter	The cascaded integrator-comb 2 (CIC2) filter is a second-order digital low-pass (decimation) filter for delta-sigma converters. See CIC2 filter section for details	Should be turned on to scale the resolution and for the best performance
Enable sensor auto-reset	When enabled, the baseline is always updated and when disabled, the baseline is updated only when the difference between the baseline and raw count is less than the noise threshold. When enabled, the feature prevents the sensors from permanently turning on when the raw count accidentally rises due to a large power supply voltage fluctuation or other spurious conditions	For most cases turned off. Can be turned on in specific conditions (harsh environment)
Enable self-test library	Not used	–
Enable multi-frequency scan (MFS)	Not used	–
Enable external frame start	Not used	–
Enable (software) IIR filter	Enables the software IIR filter (see Filters section for details)	Can be turned on depending on the noise level after tuning. The average filter (moving window) is recommended to be turned on. Hardware IIR filter is preferable over the software IIR filter
Enable median filter (3-sample)	Enables the non-linear median filter (see Filters section for details)	(See Filters section for details)
Enable average filter (4-sample)	Enables the finite-impulse response filter (no feedback) with equally weighted coefficients (see Filters section for details)	(See Filters section for details)
Enable (Hardware) IIR filter	Enables the hardware IIR filter (see Filters section for details)	(See Filters section for details)

Figure 86.

ISX Settings tab

Table 24.

ISX Settings tab parameters

Parameter	Description	Recommended values
Raw count calibration level (%)	The target raw count calibration % to which the raw counts are calibrated on enabling auto-calibration. It should be set such that the raw counts do not saturate during ISX sensor operation.	40% by default
Inactive sensor connection	The state of the ISX sensor when it is not being scanned	High-Z by default

Widget Details tab parameters

Figure 87.

Widget Details tab

Table 25.

Widget Details parameters

Parameter	Description	Recommended values
Enable multi-frequency scan	Not used	–
Clock source	Source of the clock to be used for ISX There are three sources available: Direct SSC PRS	Keep it default i.e., Direct. (SSC/PRS – not implemented/not tested).
Lx clock divider	Divider to the system clock (IMO clock frequency); the resulting clock is used to drive the sensor $F_s=\frac{F_{\mathrm{mod}}}{LxClkDiv}$ Value must be multiple of 4 The range of valid values: 8-4096	Lx clock divider can be set as per guidelines in Section Setting the Lx clock divider .
LSFR range	Not used.	–
Number of sub conversions (NSUB)	See Section Number of sub-conversions for details	See Section Number of sub-conversions for details
Decimation rate (CIC2) (5 th Gen CAPSENSE™)	See Section Cascaded integrator-comb 2 (CIC2) filter for details	See Section Cascaded integrator-comb 2 (CIC2) filter for details
CIC2 accumulator shift (5 th Gen CAPSENSE™)	See Section Cascaded integrator-comb 2 (CIC2) filter for details	See Section Cascaded integrator-comb 2 (CIC2) filter for details
Reference CDAC	Cref is used as gain, the lower it is the higher the response/sensitivity we will observe for the same changes in charge The range of valid values: 0-255	Set to Auto
Reference CDAC Boost	Divides the actual Reference CDAC value to increase sensitivity	Recommended to “Disable”. Use it to achieve higher sensitivity only after trying other available options
Fine CDAC value	Not used	–
Row fine CDAC value	Not used	–
Enabling compensation CDAC	The compensation constant amount of inductance from the sensor to increase the sensitivity, which allows for higher gain setting for ∆L conversion (lower reference CDAC value) The range of valid values: 0–255	Set to Auto
Compensation CDAC value	The compensation constant amount of inductance from the sensor to increase the sensitivity, which allows for higher gain setting for ∆L conversion (lower reference CDAC value) The range of valid values: 0–255	Set to Auto
Compensation CDAC divider	Divider to ‘Lx clock divider’, which decides the number of times the compensation CDAC is used (K comp ) within a single sense clock period. $K_{comp}=\frac{SenseClockDivider}{CompCDACDivider}$ Auto calibrated when ‘CDAC auto calibration’ is enabled With increase in the inductance value for the sensor, higher charge is produced. To compensate that, larger value of Comp CDAC is needed, thus smaller value of Comp CDAC Clock divider The value range is [3 to 4095]	Set to Auto
CDAC dither scale	Sets the CDAC dither value. As the sensor inductance is swept, the raw count should increase linearly with inductance. There are regions where the raw count does not change linearly with input inductance these are called flat-spots. Dithering helps to reduce flat-spots using a dither CDAC. The dither CDAC adds white noise that moves the conversion point around the flat region The value range is [0–6]. Higher values mean lower dithering	Set the value depending on the sensor inductance as explained in Section Setting the CDAC dither scale
Course Initialization bypass	Enables skipping the coarse initialization and thus, the scan refresh rate increases	Disabled

Widget threshold parameters

Table 26.

Widget threshold parameters

Parameter	Description	Recommended values
Finger threshold	The finger threshold parameter is used along with the hysteresis parameter to determine the sensor state as follows: ON = Signal > (Finger Threshold + Hysteresis) OFF = Signal ≤ (Finger Threshold – Hysteresis) Note that “Signal” in the above equations refers to: Signal = Raw Count – Baseline	80 percent of signal when sensor is touched
Noise threshold	Sets a signal limit below which a signal is considered as noise. Raw count limit above which the baseline is not updated In other words, the baseline remains constant as long as the raw count is > baseline noise threshold, unless Enable sensor auto-reset is selected	30 percent of signal. Because most of the proximity solutions, speed of human hand movement is slower to control algorithm, keep this threshold as low as possible, except when liquid tolerance is required
Negative noise threshold	Raw count limit below which the baseline is not updated for the number of samples specified by the low baseline reset parameter The negative noise threshold ensures that the baseline does not fall low because of any high-amplitude repeated negative-noise spikes on a raw count caused by different noise sources such as ESD events The baseline is not updated for the number of samples specified by the Low baseline reset parameter	40 percent of signal
Low baseline reset	If a finger is placed on the sensor during a device startup, the baseline gets initialized to a high raw count value at a startup. When the finger is removed, the raw count falls to a lower value. In this case, the baseline should track low raw counts. The Low Baseline Reset parameter helps handle this event. It resets the baseline to a low raw count value when the number of low samples reaches the low baseline reset number	Keep it default, i.e., 30
Hysteresis	Value used in addition to thresholds as mentioned below, to prevent the sensor status output from toggling due to system noise Sensor state is reported: ON= Difference Count > Threshold + Hysteresis OFF= Difference Count < Threshold – Hysteresis Hysteresis is not available for the low power widget	10 percent of signal
ON debounce	This parameter indicates the number of consecutive scans during which a sensor must be active to generate an ON state from the system. Debounce ensures that high-frequency, high-amplitude noise does not cause false detection	3

The design tuning parameters can be tuned from CAPSENSE™ Configurator in ModusToolbox™.

Tuning verification

This section explains how to be confident that the system is tuned correctly and works as expected. In general, this post-tuning procedure is redundant if all steps in the tuning procedures are done correctly.

After the complete tuning, the last step is to verify that the system is working as expected. If tuning was done carefully as described in the previous sections, the system should work properly.

There are two steps to verify that the system is working as expected:

1. Verify that the raw counts are changing due to changes in the inductance and there is no saturation

2. Convert obtained raw counts to the inductance and compare it to one measured with an LCR meter

The first step includes testing the sensor with the target:

• In case of force application:

  • Press the button and check whether the raw count level increases

• In the case of proximity application:

  • Move the target towards the sensor and check whether the raw count level increases

• Check that saturation is not occurring in any case

The second step includes converting raw counts to the inductance, measuring sensor inductance with an LCR

meter, and comparing both values:

1. Convert the raw counts to the inductance using the following equation:

   Figure 88.

   Raw counts to the inductance equation

   $L\left(uH\right)=2\times R^2(CompCDACCode\cdot \frac{LxClkDiv}{CompCDACDiv}+\frac{MaxRawcounts-RawCounts}{MaxRawcounts}\cdot \frac{\mathrm{Re}fCDACCode}{2}\cdot LxClkDiv)\cdot 8.864\cdot {10}^{-9}$

   Where,

   • R ≈ Rext + Rint ≈ 560 + 120 ohm

     • For larger inductance values add Rinductance (coil resistance) for more accuracy. R ≈ Rext+Rint+Rinductance

   • Resolution = Max Raw Counts = LxClkDiv x NumOfSubConv. If CIC2 filter is enabled, CIC2 resolution (max. raw count) should be used

   • This equation works for inductances larger than 5 µH. For smaller inductances, the equation provides an overestimation of the actual value

   • Equation accuracy is limited to CDACs mismatch and resistance variation. Note that the LxClkDiv value will affect the obtained inductance due to transition processes not included in the equation

2. Compare this value with one obtained with an LCR meter

3. If

   $|Lcalc-Lmeasured|$

   is less than 2-5 µH, the system works as expected (only when Rinductance is included in the equation)

Performance and reliability

When designing robust and reliable real-world applications of inductive sensing at a mass scale, there are certain challenges which need to be considered. This application note attempts to address the most significant challenges as described in the following sections.

Power supply transient noise

It is preferred to have an LDO regulator in the system to reduce the impact of power supply transients at all frequencies.

In case significant noise (>20% of the actual signal) is observed on the signal due to power supply variations, the following decision tree shall be followed to select a mitigation:

Figure 89.

Power supply noise mitigation decision tree

Table 27.

Mitigation options based on the PST noise frequency

Frequency range of noise	Preferred mitigation option
0–30 Hz	CMF(Common mode filter) LDO (requires hardware) Multi-pin configuration (requires hardware configuration)
30 Hz – 1/Tscan	LDO (requires hardware) Multi-pin configuration (requires hardware configuration) (CMF is not recommended for this frequency range)
>1/Tscan	Less impact of power supply transients post the cut-off frequency.

Where,

$\mathrm{T}\mathrm{s}\mathrm{c}\mathrm{a}\mathrm{n}\mathrm{=}\frac{\mathrm{L}\mathrm{x}\mathrm{C}\mathrm{l}\mathrm{k}\mathrm{D}\mathrm{i}\mathrm{v}\mathrm{\times }\mathrm{N}\mathrm{s}\mathrm{u}\mathrm{b}}{\mathrm{F}\mathrm{}\mathrm{mod}\mathrm{}}$

Tscan = Scan time

LxClkDiv = Lx clock divider

Nsub= Number of sub-conversions

Fmod = Modulation clock frequency (46 MHz by default)

It is recommended to keep the IIR filter enabled by default. RC filters and samples averaging also help mitigate the noise.

Linear regulator

The easiest way to suppress the power supply noise is to use a good linear regulator with high PSSR. The following are a few options with 3.3 V as output which can be used.

Table 28.

Suggested linear regulators

MPN	Manufacturer	PSRR	Iq (µA)
TLV74033PDBVR	Texas Instruments	67 dB ~ 32 dB (100 Hz ~ 1 MHz)	80
TLV74133PDBVR	Texas Instruments	70 dB ~ 55 dB (100 Hz ~ 1 MHz)	50
TLV74333PDBVR	Texas Instruments	68 dB ~ 28 dB (100 Hz ~ 100 kHz)	60
AP2120N-3.3TRG1	Diodes Incorporated	65 dB (1 kHz)	50
TPS7A0233PDBVR	Texas Instruments	55 dB (1 kHz)	0.6
TPS7A0333PDQNR3	Texas Instruments	55 dB (1 kHz)	0.3
TCR3UF33A,LM(CT	Texas Instruments	70 dB (1 kHz)	0.68

Multi-pin configuration

If the noise in the sensor signal due to power supply variation reaches beyond 20% of the actual signal, a multi-pin sensor configuration can be used as shown in Figure 90.

Figure 90.

Multi-pin sensor configuration

To enable this feature, the sensor should be connected as shown in the figure above. Using three Tx pins and one Rx pin gives better results than using two Tx pins and two Rx pins.

When selecting the pins in the CAPSENSE™ Configurator, follow the

Figure 91

. In the scan configuration, click on the “Pin” tab corresponding to the respective sensor button. Select the three Tx pins (Tx1, Tx2, and Tx3) and the Rx pin(Rx1) in the dropdown.

Figure 91.

Multi-pin configuration in the CAPSENSE™ Configurator

In multi-pin configuration, as multiple pins are connected, the effective internal resistance will be lesser (the internal resistance of the pins in parallel). Use this lower resistance value in equations, such as the Lx clock divider calculation.

Multi-pin configuration also helps with temperature-related raw-count drift.

Note:

When using multi-pin configuration, it is not recommended to use the same pin for different sensors.

Common mode noise filter (CMF)

A common mode filter can be used to suppress common mode noises present in the system. The filter can only be applied when there are more than two sensors present in a widget

2

.

The common mode filter can be enabled in the CAPSENSE™ Configurator by going to the

Widget Details

tab under

Advanced

.

Figure 92.

Enabling the CMF (common mode filter) in the CAPSENSE™ Configurator

The

Common mode filter threshold

is the signal threshold up to which the signal obtained is considered as common mode noise and included in the common mode noise calculation. It helps distinguish between the common mode noise and the actual touch signal.

The recommended value if large common mode noise is expected: 75% of the touch signal (maximum possible value). The maximum common mode noise must be below this value otherwise the noise will not be removed.

The recommended value if large common mode noise is not expected: 40% of the touch signal (same as noise threshold)

The maximum value is 75% of touch signal considering that the finger threshold is at 80% of the signal. Going beyond this value will result in improper CMF operation.

It is recommended to enable the median filter along with the common mode filter.

Note:

The common mode filter removes/suppresses common mode noise that is caused by temperature changes. Due to this, the baseline algorithm will not track the temperature changes. Hence, the common mode filter threshold should be set such that it handles the maximum expected raw count shift caused by temperature changes.

2

Currently available only for widgets with more than two sensors.

2

Currently available only for widgets with more than two sensors.

Raw count drift

Temperature drift can lead to slow raw count drift. In case the raw count drift is significant, a multi-pin configuration, or common mode filter (CMF) can be used to mitigate the raw count drift.

Crosstalk

Two types of crosstalk can occur in systems with more than one inductive sensor as described in this section.

Electrical crosstalk

Capacitive coupling between adjacent traces/components and sensors can contribute to crosstalk, i.e., unwanted signal being introduced in the sensor. Any nearby floating electrodes can also contribute to crosstalk. Rx lines/traces are more sensitive to crosstalk and higher on the sensors tuned for higher sensitivity.

Mitigation

• Route the sensor traces as far from each other as possible, add ground trace in between if required.

• Must be handled through appropriate tuning.

• Sensors need to have dedicated C_Vref capacitors.

Mechanical crosstalk

The pressure on a button can have some impact on adjacent buttons as the overlay used is a single unit as shown in the following figure. For thinner overlays, the impact is so small that it is not noticeable, however for thicker overlays (e.g. stainless steel with thickness > 0.4 mm and button size =<25 mm), it can produce mechanical crosstalk resulting in signal on the adjacent sensor.

Figure 93.

Mechanical crosstalk in 2 button setup

To avoid the mechanical crosstalk in adjacent buttons, the following actions are recommended as shown in

Figure 94

:

• Strong and stiff spacer between the PCB and overlay

• Strong adhesive/epoxy

• Grooves on the button boundary or corrugated overlay surface

Figure 94.

Mechanical decoupling measures to avoid undesired motion

For more information, see Section

Overlay thickness

.

2

Currently available only for widgets with more than two sensors.

Troubleshooting tips and techniques

See the Tuning debug FAQs section of the

AN85951 – PSOC™ 4 and PSOC™ 6 MCU CAPSENSE™ design guide

for troubleshooting and debugging.

References

Application notes

1. AN85951

   • PSOC™ 4 and PSOC™ 6 MCU CAPSENSE™ design guide

2. AN79953

   • Getting started with PSOC™ 4 MCU

3. AN234231

   • PSOC™ 4 CAPSENSE™ ultra-low-power capacitive sensing techniques

4. AN86233

   • PSOC™ 4 MCU low-power modes and power reduction techniques

Code examples

5. PSOC™ 4: MSCLP CAPSENSE™ low-power proximity

6. PSOC™ 4: MSCLP CAPSENSE™ liquid-tolerant proximity-sensing

7. PSOC™ 4: MSCLP CAPSENSE™ low power

8. PSOC™ 4: MSCLP robust low-power liquid-tolerant CAPSENSE™

Other resources

5. CAPSENSE™ webpage

6. ModusToolbox™ webpage

7. ModusToolbox™ software help on GitHub

8. PSOC™ 4 CAPSENSE™ Component datasheet

External references

5. Mohan, S. “Simple Accurate Expressions for Planar Spiral Inductances” in IEEE Journal of Solid State Circuits, vol. 34, no. 10, (Oct. 1999): pp 1419-1424

6. Howard Johnson, Martin Graham. “High-Speed Digital Design: A Handbook of Black Magic”

7. Theory of Plates and Shells (2nd Edition) – Timoshenko & S. Woinowsky-Krieger

8. Simplified Equations: Loaded Flat Plates

Revision history

Document revision	Date	Description of changes
**	2024-12-10	Initial release
*A	2025-02-13	Updated Layout considerations and Overlay selection sections Updated Figure 37 , Figure 46 , and Figure 72
*B	2025-03-11	Added sections LED placement and routing and Overlay warpage . Updated section Sensor Designer
*C	2025-04-03	Template update Updated figures Added sections CY8CPROTO-041TP PSOC™ 4100T Plus CAPSENSE™ Prototyping kit and Capacitive and inductive sensing buttons example

Trademarks

The Bluetooth® word mark and logos are registered trademarks owned by Bluetooth SIG, Inc., and any use of such marks by Infineon is under license.

PSOC™, formerly known as PSoC™, is a trademark of Infineon Technologies. Any references to PSoC™ in this document or others shall be deemed to refer to PSOC™.

1

Multi Frequency Scan (MFS) is not supported with Common mode filter feature. See

Common mode noise filter (CMF)

.

2

Currently available only for widgets with more than two sensors.