A PCAP panel that tracks ten fingers cleanly on a bench can behave very differently on a plant floor, writes Mike Wagner, Chief Innovation Officer at Butler Technologies. Washdown spray, a heavy glove, or a shared drive ground can create touches nobody made, and one phantom second touch turns a tap into a gesture.
Multi-touch is not simply present or absent. It comes from a sensing method, a sensor pattern, a scan budget and the cover glass, and each has a point where it stops working.
Quick answer: PCAP multi-touch comes from mutual capacitance sensing, which measures every row and column intersection separately, so several fingers are located without ghost points. Sensor pattern, scan rate and cover glass thickness set its limits. It pays off for zooming dense HMI screens, but water, gloves and conducted noise must be engineered for, and critical controls need single-touch lockout.
Mutual Versus Self Capacitance
Self capacitance measures each row and each column as a whole electrode. Two touches activate two rows and two columns, which creates four candidate positions, two of them false.
Barrett and Omote call this ambiguity “ghosting.” Mutual capacitance avoids it by measuring coupling at every intersection individually, so each touch sits at its own node.
An Analog Devices technical article describes mutual capacitance as “true multitouch with two or more unambiguous touches.”
Sensor Pattern and Scan Budget
Node count equals transmit lines multiplied by receive lines, and the controller must measure every node each frame.
- More lines or longer integration improve resolution and noise rejection but lower the report rate
- Electrode pitch sets how close two fingertips can be before they merge
- Diamond patterns, which Barrett and Omote say typically run 4 to 8 mm, and bar patterns both work if the pitch is fine enough
- A datasheet touch count is a tracking limit, not a promise of fast gestures
- Heavier filtering for a noisy site lengthens each frame, so a smooth lab pinch can lag in the field
PCAP touchscreen layers with different glass thicknesses
Cover Glass Thickness
Thicker cover glass moves the finger farther from the electrodes, weakening the signal and spreading it across more nodes. Impact resistance improves, so thickness trades ruggedness against precision.
Barrett and Omote note that the thinner the glass and the higher its dielectric constant, “the better the performance.” A Texas Instruments design guide adds that air gaps between electrodes and overlay reduce sensitivity and let moisture in.
Manufacturers that build PCAP touch screens in several glass thicknesses, such as Butler Technologies, Inc., offer up to 10 touch points with standard or chemically strengthened glass from 0.55 to 3.0 mm. The practical rule: use the thinnest glass that passes the impact test, and try chemically strengthened glass before thicker glass.
Water, Gloves and Electrical Noise
Capacitive touch in industrial settings faces conditions that either add a false signal or weaken the real one, and multi-touch suffers first.
- Conductive fluids such as tap water or coolant create false touches, countered by a driven shield and combined mutual and self sensing
- Thick gloves act like extra glass, and glove modes that raise sensitivity also amplify water and noise
- Conducted noise from drives causes jitter and phantom touches, which IEC 61000-4-6 immunity testing from 150 kHz to 80 MHz covers
- Frequency hopping, filtering, and low-impedance grounding counter that noise
- Palm rejection discards contacts above a size threshold when operators brace a hand
Where Multi-Touch Belongs
Multi-touch earns its place on dense screens: pinch-zoom on P&ID and trend displays, two-finger page swipes, and two-point confirmation of irreversible software actions. It should be off on screens that jog, start or change setpoints.
Two-point confirmation reduces accidental taps but is not a safety function. ISO 13851:2019 requires “simultaneous actuation by the use of both hands,” with a 0.5 s synchronous limit for Type III devices. Two fingers of one hand defeat a gesture by definition, so safety-rated two-hand controls stay physical.
On critical screens, single-touch lockout reports only the first contact, a second contact cancels any pending action, and hold-to-run needs one unbroken contact.
The Stack Behind Every Touch
Every limit above comes from the same place. Multi-touch is not a datasheet number but the combined result of sensing method, pitch, scan budget, cover glass and environment. A screen tuned on the bench with dry fingers says little about a gloved operator beside a drive cabinet, so tune on the final glass stack and test for real conditions.
Key Takeaways
- Mutual capacitance measures every node separately, which keeps PCAP multi-touch free of ghost points
- Node count, filtering and scan rate decide gesture speed and how close two fingers can be
- Cover glass thickness trades ruggedness against multi-touch precision, so use the thinnest glass that passes impact testing
- Use multi-touch for navigation, lock critical controls to single touch, and keep safety-rated two-hand controls physical
Frequently Asked Questions (FAQs)
Why do self capacitance touch screens show ghost touches?
Self capacitance reads each row and column as one electrode. Two diagonal touches activate two rows and two columns, which intersect at four points, and the controller cannot tell which two are real. Mutual capacitance measures every intersection separately, so each touch has its own node.
How many touch points does an industrial PCAP screen need?
Most industrial HMIs need two for pinch-zoom and swipes, and some collaborative displays benefit from more. Scan speed and noise performance with the final glass matter more than the headline count, because excess filtering slows gestures and false points break them.
Is IEC 61000-4-6 testing relevant for touch screens?
Yes. It tests immunity to conducted RF disturbances from 150 kHz to 80 MHz on power and signal cables. Capacitive touch measures changes of a few picofarads, so conducted noise can cause jitter or phantom touches. Frequency hopping, filtering and good grounding are the usual countermeasures.
Can a two-finger gesture replace a two-hand safety control?
No. ISO 13851:2019 defines two-hand control devices by simultaneous actuation with both hands, with measures against defeat by one hand. A two-finger gesture can be made with one hand, so it can confirm software actions but cannot replace a safety-rated two-hand control device.
Find out more https://butlertechnologies.com/
