Advertisement

Article

Designing capacitive sensing for a specific application (Part 2 of 3)

(Editor's note : We are pleased to present this lengthy, multipart, hands-on article about capacitive sensing, which address solid engineering and design issues. Touch sensing and touchscreens are extremely popular topics with our audience; for your convenience, you can see a linked list of all articles we have published on this topic here .)

(To read Part 1 of this article, click here.)


Design flow – capacitive-sensing user interface

Figure 7 provides the typical design flow for implementing capacitive sensing. Firmware (F/W) Development, Tuning, and Production Fine Tuning are the critical phases in the life cycle of a capacitive-sensing User Interface design.

 

Figure 7: Capacitive-sensing interface design flow

  1. F/W Development:

In a broader sense, firmware implements the functionality required for the specific application; i.e., number of buttons, additional features like PWM, ADC, DAC, etc. From a capacitive sensing standpoint, the firmware does the job of scanning sensors (i.e. measuring the sensor capacitance) as well as other associated functions like processing feedback based on the sensor ON/OFF status.

For systems implementing only capacitive sensing, devices with configurable options are available. Registers are configured through serial communication protocols (like I2 C) for specific sensor functions and no firmware development is required. Implementing capacitive sensing using a programmable device provides flexibility to meet varying user needs as well as perform sensor scanning and processing.

  1. Tuning:

Tuning is the process of determining the optimum values for set of capacitive-sensing parameters for robust and reliable performance under various environmental conditions and for different mechanical constructions of the interface. This demands a thorough understanding of how a capacitive sensing system behaves under various conditions.

The key things to be considered while tuning are

  • SNR(signal-to-noise ratio) of sensor
  • Sensor scan time
  • Finger threshold settings

SNR of sensor :

 

Figure 8:Signal and noise

One of the main goals of tuning a capacitive sensing system is to reliably discriminate between TOUCH and NO TOUCH sensor states. In an SNR calculation, the signal is the change in the sensor response when a finger is placed on the sensor.


Noise is the peak-to-peak variation in sensor response when a finger is not present. For reliable capacitive sensing performance, signal strength needs to be significantly larger than noise; the general recommendation is that the signal should be at least five times the noise, for a minimum recommended SNR of 5:1.

Sensor scan time :

Sensor xcan time is the amount of time the counter counts, as described in the capacitance measuring system section above. Shorter sensor scan times lead to lower SNR. Higher scan times lead to delayed response time and higher power consumption. Thus, based on the sensor parasitic capacitance (CP ), the sensor scan time needs to be optimized for SNR, response time, and power consumption.

Finger threshold setting:

The Finger Threshold is set to indicate a finger touch. This Finger threshold should be set carefully to avoid false triggering because of noise and atmospheric changes. The general recommendation is that the finger threshold should be set to 75% of the signal strength for reliable touch detection as was shown in Figure 5.


Figure 9:Typical tuning flow for a capacitive sensing design

(To enlarge,click here )

Figure 9 shows that tuning is a time consuming, laborious, and repetitive process that has to be repeated whenever the PCB or overlay is changed during development.

3. Production Fine Tuning:

Capacitive sensing performance depends on the physical properties of the capacitive sensors and environmental conditions. The parasitic capacitance for a sensor varies when there is vendor change, process variation, or variation in environment such as humidity or temperature. This requires fine tuning through statistical analysis on samples during production in order to minimize yield loss due to failure. As we can see, there are many steps and issues which one needs to address before releasing a design for mass production.

While designing a capacitive-sensing system for a particular application like a TV or monitor, we encounter typical challenges such as PCB vendor change and noise affecting the capacitive-sensing performance, which often lead to retuning. Some methods to deal with such challenges and reduce the tuning effort are:

  1. Auto-tuning
  2. Layout considerations
Auto-tuning

An innovative method is now available in the industry where the device dynamically tunes itself (monitors and sets parameters automatically) by monitoring changes in noise and the environmental conditions of the system. This method also enables the device to initialize all the capacitive-sensing-related parameters at power up, based on the environmental conditions and the mechanical design of the system.


a. Baselining:

Common environmental changes that affect the capacitive sensing measurement are temperature and humidity drift. Temperature/humidity drift causes changes in capacitive measuring circuit components/parameters and also affects CP and CF , causing the raw counts to change. A typical variation of raw count with temperature is shown in Figure 10 .

Figure 10: Raw-count variation with temperature

If a constant reference is used to detect a button touch, the temperature/humidity drift may result in a false button press or a missed button press.

Baseline compensation as part of an auto-tuning sequence adjusts the reference level (baseline) and the noise thresholds automatically so that low-frequency noise is kept below the threshold levels to avoid false triggers.


b. Optimal parameter settings:

Based on a sensor’s physical properties and environmental noise, sensor parameters should be assigned with optimal values at power up to make the capacitive sensing system work reliably.

As mentioned before, it is critical to maintain the SNR above 5:1 in order for the sensors to work consistently. The first thing that an auto-tuning algorithm needs to do is compute and optimize values of specific device parameters, such that the SNR is maintained above 5:1. Optimal scan speed is set in order not to increase power consumption and response time. Threshold levels for noise and finger response are optimized and set at power up.

Let us understand by taking most common scenarios how auto-tuning is beneficial.


I. PCB process variation:

A single board manufacturer may have different manufacturing sites, which leads to slight variations in the manufacturing process that changes the parasitic capacitance of sensors from board to board. This leads to retuning if the yield loss because of this change is too high. For example, in Figure 11 , one of the sensors of Board #4 does not respond to finger press since the sensor has a high Cp due to a printed circuit board (PCB) process variation.

 


Figure 11:Performanceofboards subjected to process variations

II. PCB Vendor change:

OEMs typically have many PCB/FPC manufacturing sources qualified in order to protect against manufacturing cost increases and capacity shortage. Each board manufacturer may use different PCB materials which, in turn, may have different parasitic capacitance on sensors, resulting in lower yield. For example, if tuning is finalized for a board manufactured from a vendor X, the same tuning parameters may not be fully applicable to a board manufactured from vendor Y, resulting in lower yield.

 

Figure 12: Performance of boards from different vendors

In Figure 12 , the sensors on Board #3 and Board #4 do not respond to a finger press, since the sensor has high CP due to the PCB manufacturer being different.

An auto-tuning algorithm at power up which automatically determines the parasitic capacitance of the sensors eliminates both of the above-mentioned problems.

III. Overlay dielectric / thickness change:

Sensitivity is directly proportional to the dielectric constant of the overlay material and inversely proportional to overlay thickness. This means that if a design tuned for 2mm glass (dielectric constant ? of 8.0) is changed to a 2mm plastic (? of 2.8), the design needs to be retuned to work effectively. Similarly, if a 3mm plastic overlay is changed to a 3.5mm plastic overlay, the tuning has to be done all over again, Figure 13 .

 

Figure 13: Performance of boards with different overlay thickness

(To enlarge,click here )

An auto-tuning algorithm adjusts the sensitivity when there is a change in overlay thickness, such that the sensors work as expected even when there is overlay-thickness variation. The capability to do all of the auto-tuning functions at run time as well as set the parameters automatically to provide a stable SNR of greater than 5:1 is the key to a robust auto tuning technique.

Design portability is another common scenario in TV/monitor applications which requires considerable reworking, in terms of tuning when porting a design across different applications having similar requirements. In TV/monitor applications, we usually come across requirements where a single design needs to be used for different models of TV/monitors with varying screen size, Figure 14 . Hence, the placement of the touch control panel may vary, which in turn may lead to change in button position.

When there is change in button position, there is a possibility that the trace length will vary as well. Two different models can also have two different button-size requirements. In addition, different models of TV/monitors may have different requirements for overlay thickness. All of these differences lead to variations in parasitic capacitance.

With the capability of auto-tuning to handle changes in parasitic capacitance and overlay thickness, it is possible to achieve seamless porting of the design.


Figure 14: Different models of TV/monitor

c. Noise determination: noise–immunity level improvement:

Noise is random fluctuation in an electrical signal. Noise generated by electronic devices varies greatly, as it can be produced by several different effects and sources. Noise can also be described as a summation of unwanted or disturbing energy from natural and man-made sources.

Multiple tuning efforts to compensate for noise are needed, especially if noise levels are in-between noise threshold and finger thresholds or if the noise level keeps varying in a system. Auto-tuning must determine noise levels in the system dynamically and adjust the noise and finger threshold accordingly.

(End of Part 2 ; Part 3 looks at layout issues, and noise sources and reduction.)

About Authors

Subbarao Lanka is a Senior Applications Engineer working at Cypress Semiconductor Corp. on Capacitive Touch Sensing applications since 2007. His responsibilities include defining technical requirements for new capacitive sensing controllers, developing new capacitive sensing controllers, conducting system analysis, debugging technical issues for customers, and technical writing. He can be reached at slan@cypress.com.

Shruti H is an Applications Engineer working at Cypress Semiconductor on Capacitive Touch Sensing applications since 2009. She works on defining debugging technical issues for customers, technical requirements for new capacitive sensing controllers, developing new capacitive sensing controllers, conducting system analysis, and technical writing. She can be reached at sshh@cypress.com.

Editor's note : Liked this? Want more?

If you are interested in “analog” issues such as signal input/output (sensors and transducer, real-world I/O); interfacing (level shifting, drivers/receivers); the signal chain; signal processing (op amps, filters, ADCs and DACs); and signal integrity, then go to the Planet Analog home page here for the latest in design, technology, trends, products, and news. Also, sign up for our weekly Planet Analog Newsletter here.

0 comments on “Designing capacitive sensing for a specific application (Part 2 of 3)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.