- Optical imaging solutions like CIS and LED emitting/receiving may replace capacitance as mainstream technologies for display-based fingerprint sensing.
- Under-display CIS could come faster, as early as 2018, but in-display solutions will be preferred as long as the AMOLED display yield is more mature.
Building the fingerprint sensor into the display helps drive the 18:9 display trend and is a strong selling point. Due to technological limitations, only under-display solutions will use CIS (CMOS Image Sensor) in the second half of 2017 and first half of 2018. CIS is not exactly built into the display, but fingerprint sensing can be achieved within the display’s active area by sacrificing some thickness. Display-based fingerprint sensing has not completely matured, but smartphone brands and panel makers are looking forward to offering it to convince end users to replace old models.
Fingerprint sensing within the display’s active area will influence demand for 8-inch and 12-inch semiconductor fabs, and it will be a technical milestone as well reshape the supply chain. Currently, silicon-based fingerprint IC makers handle the design of the IC (sensor and controller), but display-based fingerprint sensing is forcing IC makers to co-work with panel makers because the sensors must be combined or integrated with the display. The under-display CIS solution already has a release schedule. Other solutions are still under development and being optimized.
Fingerprint sensor structure design
The major approaches to structure include under-display, on-display, and in-display. Due to the display’s emitting and presenting functions, the fingerprint sensor cannot hinder transmittance or lower the aperture ratio. This is similar to what has been done with touch panels. In addition, AMOLED and LCD have quite different principles and structures, which can affect the design and adoption of fingerprint sensors. IC makers and panel makers have prioritized development for AMOLED because it is targeted at the premium or flagship segment, which has a higher cost tolerance.
An under-display structure with CIS has a smaller impact on the display but it requires modifications on the display end. A top-emission AMOLED panel with a reflective (opaque) film beneath the anode cannot let light pass through, so it is necessary to punch pinholes. With an LCD, holes in optical films can damage their function. Replacing the CIS with LEDs may not be feasible due to the light path designed for Tx to reach Rx. Tx can be under-display with a specific light spectrum, but an under-display Rx could have a weak SNR (signal-noise ratio) for recognition. The CIS is simpler because it takes pictures instead of differentiating the Tx and Rx signals to determine the SNR. Another under-display approach uses ultrasonic technology, but it is not in practice, yet. Qualcomm’s SenseID has penetration at 700-800 µm, and it can be used with an under-glass design, but it is far from being used in an under-display structure. Apple’s ultrasonic patent (US20170053151) describes an on-display structure, and its sensor design is around the display and beneath the cover glass.
An on-display Rx is on the display, and it can be on the encapsulation layer (AMOLED) or color filter (LCD). It can also be patterned on an additional transparent layer. The locations of Tx depends on the technology. For example, with LEDs, the Tx can be near the IR located in the backlight. In addition to an ultrasound-based technology, Apple also patented a capacitive on-display design (US9582102), but these two concepts are not in practical development and are only in the patent stage. Capacitance has a quite weak SNR, so signal enhancement is necessary, and it should be infeasible for under-display and in-display designs. Crutialtec’s technology could be faster. The company demonstrated a transparent sensor layer in the first quarter.
An in-display structure should be panel makers’ destination. It integrates the Tx and Rx into the TFT circuitry. So far, only the optical imaging method with LED emitting and receiving has been adopted. An in-display design is fully integrated with the display, so panel makers can take advantage of their internal manufacturing processes without an additional assembly. It is similar to the transition from add-on touch to embedded touch. Some panel makers are developing in-display solutions but the modifications to the light path, Rx, SNR, and display are still questionable and without clear results.
Optimizing display-based fingerprint sensing
Patents describe more possibilities, but they are not necessarily practical. Makers must consider and optimize critical factors such as the display’s internal structure, manufacture, principle, Tx/Rx matching, light path for an effective SNR, Rx circuitry, and so on.
The Tx light source can be discrete (e.g., near infrared LEDs) or use the display’s light. An OLED panel’s light is more divergent (for view angle) and a concentrating mechanism is helpful. If the Rx is a photodiode, different material types will likely have a respective sensitivity to the Tx light’s wavelength in the spectrum. As for Rx circuitry, IC makers would need to provide controllers and an algorithm to panel makers. With a CIS, a conventional semiconductor process and package will work. A TFT is another option.
A top-emission AMOLED panel has a semi-transparent cathode (metallic) and reflective anode (opaque), and panel makers must drill holes (µm level) to facilitate light’s penetration, especially for under-display solutions. In-display solutions usually have infrared LEDs and photodiodes on the TFT backplane, which affects the pattern of the existing RGB OLED materials and lowers the aperture or resolution.
With an LCD, the color filter glass influences the Tx and Rx and photoresist. That influence is more remarkable with an in-display design. An on-display design affects the display’s transmittance due to the additional layer of fingerprint sensing and its assembly. If it a discrete light source is adopted, it uncertain where the Tx will be located (on the TFT, color filter, or backlight). Some makers propose using FTIR (Frustrated Total Internal Reflection) using a specific light guide. This would be an on-display design and result in the same transmittance issue.
Some makers suggest light filtering or a collimator for more effective light sensing and an improved SNR. Besides the calculation of light angle and receiving, it would likely require additional steps during the display manufacturing process and lower the compound yield rate. Most panel makers have not found an optimal design.
After examining these structures, IHS estimates that the mainstream structure will be under-display CIS. FPC, Synaptics, and Goodix have proposed similar under-display solutions to panel makers. The CIS solution will be a good temporary solution in 2018-2019 until an in-display design is achieved. As AMOLED capacity surges with good yields, panel makers will develop in-display structures as a new add-on value. 18:9 AMOLED displays with display-based fingerprint sensing has become a clear trend. Compared to LCD, AMOLED displays are positioned for higher ASPs and simpler internal display structures. It is no wonder that IC makers and panel makers have prioritized designing fingerprint sensors for AMOLED displays. However, smartphone LCDs will account for more than 50% of the market in a few years, and appropriate solutions for them will also be considered.