Capacitive Touch with Gloves, Water and Thick Cover Glass: How to Specify It
A projected capacitive (PCAP) touch panel works by measuring very small changes in capacitance through the cover lens. Gloves and thick cover glass make the change smaller. Water and electrical noise create changes that look like a finger. Whether a PCAP display works with gloves, water and a thick lens depends on the whole stack and on how the touch controller is tuned for it, not on the sensor alone. So write the worst case into the specification at the start: which gloves, how much water, what cover material and thickness, and what power supply. The panel can then be tuned on the stack that will actually ship.
Mutual and self capacitance: two ways to see a finger
A PCAP sensor has rows and columns of transparent electrodes. Controllers measure them in two ways:
- Mutual capacitance measures the coupling at each row-column crossing. A finger takes some of that coupling away. Every crossing is measured on its own, so mutual sensing gives true multi-touch and accurate positions.
- Self capacitance measures each electrode against ground. The signal is stronger, which helps with gloves and thick covers, and it reacts differently to water than to a finger. It cannot separate several touches reliably on its own.
Modern controllers use both. Microchip’s maXTouch mXT1066TD datasheet, for example, says mutual data is used to classify touches wherever possible and that multiple touches can only be resolved from mutual data. Self capacitance is used to detect single touches when moisture is present and to catch thick-glove touches that mutual sensing alone may miss. In practice, many products drop to single-touch while the screen is wet. Allow for that in the user interface.
Cover thickness: count it in glass equivalents
A finger couples to the sensor through every layer in front of it. As a first approximation, the signal falls with each layer’s thickness divided by its dielectric constant. Soda-lime glass has a dielectric constant of roughly 7, PMMA and polycarbonate roughly 3 to 3.5, and air 1. That gives a useful rule: 1 mm of plastic weighs about as much as 2 mm of glass, and 0.5 mm of air about as much as 3.5 mm of glass. The mXT1066TD datasheet makes the same comparison, rating it for 10 mm of glass or 5 mm of PMMA with a bare finger.
Worked example: why the sensor is laminated to the cover. A 1.1 mm glass cover laminated to the sensor counts as about 1.1 mm of glass, plus a little for the adhesive. Leave a 0.5 mm air gap between the same cover and the sensor and the total rises to about 1.1 + 3.5 = 4.6 mm glass equivalent, four times worse, without the product looking any thicker. That is why PCAP sensors are laminated to the cover (G+G or G+F constructions) rather than mounted behind it with a gap. Air gaps between the sensor and the LCD below it are a different matter; they affect noise and optics, not the finger signal.
How far you can go depends on the controller, the screen size and the tuning. Some published figures:
| Source | Bare finger | With gloves |
|---|---|---|
| Microchip mXT1066TD datasheet | Up to 10 mm glass or 5 mm PMMA | Multi-finger with a 5 mm glove through 6 mm glass or 3 mm PMMA |
| Riverdi controller comparison | Max. cover glass 10 mm (ILI2511, ILI2132A), 15 mm (EXC80W32) | Max. glove 2 mm (ILI2511), 5 mm (ILI2132A, EXC80W32) |
| Elo touch-through guidance | Glass up to 6 mm, plastic up to 3 mm | — |
These are controller limits under good conditions, not guarantees for your stack. Riverdi’s standard industrial PCAPs use 1.1 mm glass. Our S35001D uses a 0.70 mm chemically strengthened lens in a G+G construction with an FT5446 controller on I²C. Thicker is possible, but it moves the tuning closer to its limits, and gloves and water then have less margin.
Gloves
Thin latex and nitrile gloves usually work with standard tuning. Thick work, winter or leather gloves do not. They need a glove mode in the controller firmware, which raises sensitivity and lowers the touch threshold. That makes the panel more sensitive to everything, including noise, hovering fingers and water. Good practice:
- Name the actual gloves: material and thickness, single or double gloves, dry or wet.
- Tune and test with those gloves on the final cover lens, not on a bare sensor.
- Check that a bare finger still works normally when glove mode is on.
- Make buttons large: a gloved fingertip is wider and less precise.
If users wear thick gloves all the time or need a hard stylus, a resistive panel may be the simpler answer. See capacitive vs resistive touch screens.
Water, condensation and cleaning
Water is conductive enough to couple to the electrodes. A droplet that spans a row and a column can look like a light touch. A water film connected to a gloved or bare hand can spread a touch over a wide area. Salt water, sweat and cleaning fluids are worse than tap water because they conduct better. Controllers use self-capacitance data to tell water from fingers. The mXT1066TD datasheet, for example, specifies no false touches from condensation or water drops up to 22 mm in diameter. Your own tests should cover:
- Droplets of several sizes, a spray, running water and, if relevant, salt water or the customer’s cleaning agent.
- Touch while wet: does it fall back to single-touch, and is that acceptable?
- Condensation after moving the product from cold to warm.
- Wiping the screen. A software “cleaning lock” that ignores touch for a few seconds avoids wiping turning into button presses.
Noise: the problem that shows up last
PCAP signals are small, so electrical noise matters. The usual sources are the LCD itself, switching power supplies, chargers with poor isolation, motors and fluorescent or LED lighting drivers. A ground shield layer in the touch stack reduces noise from the LCD, and controller firmware filters the rest. Specify immunity to conducted RF disturbance (IEC 61000-4-6) if your market requires it. Riverdi, for example, lists its industrial PCAPs as passing level 3, 10 Vrms. Test with the real power supply and cabling, not with a lab bench supply. Ghost touches that appear only when a charger is plugged in or a motor runs are almost always noise.
ESD is the other half of the electrical picture. The cover lens is the part people touch. See the LCD ESD protection guide.
What to put in the specification
| Item | Why it matters |
|---|---|
| Cover material, thickness and printing | Sets the signal level; the tuning is tied to it |
| Bonded or air gap to the LCD | Affects noise from the LCD and readability |
| Gloves: type and thickness | Decides whether glove mode is needed |
| Water: droplets, spray, running, salt, cleaning agents | Decides the moisture settings and tests |
| Number of touches and gestures needed | Single-touch tuning has more margin than multi-touch |
| Interface and software: I²C or USB, operating system and driver | USB panels often appear as standard HID devices; I²C needs a driver |
| Enclosure: metal frame, grounding, bezel gap | Metal near the active area changes the edge response |
| EMC and ESD standards to pass | Decides the filtering and the shield layer |
Video: Capacitive Touch with Gloves, Water and Thick Glass
This SuccessLCD explainer covers mutual and self capacitance, how to count a cover in glass equivalents, published controller limits, and what to test for gloves, water and noise, with real SuccessLCD capacitive touch samples.
Capacitive touch FAQs
Can a capacitive touch screen work with gloves?
Thin latex or nitrile gloves usually work with standard tuning. Thick work, winter or leather gloves need a controller with glove mode, tuned on the final cover lens with the actual gloves. Published controller limits run from about 2 mm to 5 mm of glove, depending on the controller and the cover.
How thick can the cover glass on a PCAP touch screen be?
Standard panels typically use about 0.7 to 1.8 mm of glass. With suitable controllers and tuning, bare-finger operation through 6 to 10 mm of glass is possible, and some controllers quote up to 15 mm. Plastic covers count roughly twice their thickness, and any air gap between the cover and the sensor counts several times its thickness.
Why does water cause ghost touches on a capacitive touch screen?
Water is conductive enough to couple to the sensor electrodes, so a droplet across a row and a column can look like a light touch. Controllers use self-capacitance measurements to recognise water and often fall back to single-touch while the screen is wet. Salt water and cleaning fluids are harder than tap water.
Do I need to retune the touch panel if I change the cover glass?
Yes. The tuning depends on the cover’s material, thickness and printing and on whether it is bonded. A change to any of them should be followed by retuning and a repeat of the glove, water and noise tests.
Need touch that works with gloves or water? Send the display size, cover material and thickness, the gloves and water conditions, the number of touches and your host interface. We will suggest a touch construction and the tests to run. See touch panels and cover lenses, or ask for a touch review. For a complete touch display, see custom TFT display engineering.