Adding Touch to a 1.18-Inch Round 128x128 Monochrome OLED in Lab Instruments

Key takeaways

1. A circular 1.18-inch aperture leaves less usable area than the diagonal suggests, so the sensor border and the size of the smallest target both matter.
2. Resistive sensing registers pressure, so it tolerates gloves, styli and damp fingers, which is common on shared laboratory equipment. ^[2]
3. Capacitive sensing gives a flat front and a light touch, but the controller has to be tuned around the noise produced by the display and its driver. ^[3]
4. The module already speaks SPI and I2C, so an external touch controller can normally share the existing bus instead of adding a second connector family. ^[4]
5. Fix the front-panel sealing concept first: it constrains the sensor technology, the cover thickness and the calibration routine all at once.

Laboratory instruments are touched by people wearing gloves, holding probes, or working with one hand while the other is occupied. That single fact rules out some otherwise attractive input designs before any electrical question is asked. A bench pH meter, a small centrifuge controller and a portable analyser all share the same pattern: the display is a status and parameter readout, and the input is occasional rather than continuous.

A 1.18-inch round 128x128 monochrome OLED fits that pattern. The panel is small, the interface is simple, and the 3.0V supply requirement is easy to meet in an instrument that already has a regulated logic rail. Because the display is monochrome and passive-matrix, the visual language is line art, numbers and simple glyphs, which is exactly what a parameter readout needs.

The rest of this article deals with the part that usually costs schedule: deciding whether to add touch, choosing between resistive, capacitive and bonded on-cell options, and checking what each one does to wiring, firmware and the enclosure drawing.

What a circular 1.18-inch aperture does to touch geometry

A 1.18-inch diagonal with 128x128 pixels gives a comfortable text size at reading distance but leaves very little room for interface chrome. A round active area sharpens the problem: a circular touch overlay has to cover the full circle while the bezel stays narrow, and any icon placed near the rim risks falling partly outside the reliable sensing region.

In practice that pushes the interface toward a small number of large targets: a central value with two or three radial controls, a swipe to change pages, or a rotating interaction around the dial. Size the smallest target from the finger or stylus the user will actually bring, then check that it fits inside the circle with margin for the sensor edge dead zone.

The monochrome, passive-matrix nature of the panel helps here. There is no colour rendering to preserve and no backlight diffuser to keep flat, so the front stack can stay simple: a sensor layer, an optically clear adhesive and a cover, with the display mounted behind.

Resistive, capacitive and bonded on-cell sensing compared

The three families are often presented as a quality ladder, which is misleading. Each solves a different problem, and on a small round instrument panel the deciding factor is usually the environment rather than the feel of the interface.

Resistive layers

Resistive sensing registers pressure, so it works with a gloved finger, a plastic stylus or a probe tip, and it does not care whether the finger is dry. That tolerance is why the technology remains common on equipment used in wet or gloved conditions. ^[2]

The cost is mechanical. Two conductive layers separated by spacers mean a flexible front surface, slightly reduced optical clarity and a stack that must not be clamped unevenly. On a round window the clamping geometry needs care, because pressure concentrated at one edge can bias the readings.

Capacitive layers

Capacitive sensing measures the change a conductive object makes to an electric field, which allows a flat front, no moving layers and a very light touch. ^[3] For an instrument operated with a bare finger it is usually the more pleasant option.

It is less forgiving in other ways. A thick glove breaks the coupling, water films on the cover can be read as touches, and the controller has to be tuned against the noise floor of the display and the switching supply. Those are solvable problems, but they are bench work rather than datasheet work.

On-cell and directly bonded stacks

Bonding the sensor to the display or to the cover removes an air gap, which improves clarity and reduces internal reflections. It also reduces the number of parts the assembly house has to align.

The trade is flexibility. A bonded stack is harder to rework, and changing the sensor later means changing the display assembly too. On a low-volume instrument that is usually acceptable; on a product expected to iterate quickly it is a real commitment.

Touch stack comparison for a round 1.18-inch OLED instrument

Wiring, bus sharing and firmware habits that save time

The display end of this design is unusually simple. SPI is a synchronous serial bus with a separate clock, so the host controls timing exactly and there is no arbitration to write. ^[4] I2C needs only two wires and can carry both the display and a touch controller on the same pair, which keeps the harness small, at the cost of shared bandwidth and a need for a sensible interrupt strategy.

If a capacitive controller joins an existing I2C bus, plan the interrupt line first. Polling a touch controller continuously wastes processor time and can interfere with the analogue measurements a laboratory instrument is usually making. A dedicated interrupt pin that wakes the firmware only when something changes is worth the extra trace.

Calibration belongs in production test, not in the field. Record a baseline with the panel powered inside its final enclosure, store it, and provide a service routine that lets a technician re-baseline the sensor without opening the instrument. Equipment cleaned with solvent or used with wet gloves will drift, and a hidden re-calibration path saves a service call.

Adding a sensor also adds a path for electrostatic discharge and for conducted noise. A series resistor and a small capacitor on each sensor line, placed close to the connector, is cheap insurance on a front panel that a user touches with a charged hand.

Decisions to settle before the enclosure drawing is released

The list of interdependent choices is now clear: sensor family, cover thickness and material, bonding method, bus sharing and calibration ownership. They cannot be settled one at a time, because each one changes the constraints on the others.

A useful way to break the loop is to fix the display first and treat everything else as a layer added around it. The 128x128 OLED display with SPI interface gives a known 3.0V supply requirement, a known interface set and a known 1.18-inch round footprint, so the touch stack can be designed as an addition rather than as a co-design.

Once the display is fixed, the remaining questions are mechanical and procedural: who owns the sensor firmware, where calibration data is stored, and how the front panel is sealed against the liquids the instrument will meet. Writing those answers into the specification before tooling starts is what keeps the schedule intact.

Frequently asked questions