Since the iPhone 5 was equipped with fingerprint recognition, fingerprint recognition has become a market trend. According to industry insiders, a price war for fingerprint recognition chips had already begun as early as the beginning of 2016.
Among the many chips, Feinger, a rising star from Chengdu, has become a vital force in the market. With its advanced structural design and the support of Chengdu's cost-effective microelectronics industry, its high-performance and reasonably priced products have been shipped in large quantities to the market.
The price of fingerprint chip modules has plummeted from over ten US dollars to the $4 range, where Feinger's mass-produced modules are shipped. What is the current state of fingerprint chips? What factors determine their price, and what is the bottom line for fingerprint chip and module prices?
At our invitation, Mr. Huang Hao, CEO of Feinger, provided an in-depth analysis of the patent structure and pricing secrets of fingerprint chips and modules, covering three key aspects: technical routes, algorithms, and packaging processes. This analysis combines technological and market insights. The following content was provided with the assistance of Feinger.
I. Core Technology Roadmap and Patents
When it comes to fingerprint recognition, the most crucial technology is actually the semiconductor capacitive sensor technology that detects the capacitance of the finger. Whether active or passive, it's all capacitive sensor technology. Just like the saying "all martial arts originate from Shaolin," all current fingerprint recognition semiconductor sensor technology stems from a patent application filed by Harris on May 16, 1997, with the U.S. Patent and Trademark Office, entitled "Electric Field Fingerprint Sensor Having Enhanced Feathers and Related Methods," which was granted on August 17, 1999, with patent number US9540526. In the patent specification, the inventors described previous capacitive fingerprint sensors (previously called passive by Harris, and now active by Harris) and proposed a new design based on electric field theory, improving the sensitivity of fingerprint detection and enhancing reliability by increasing the thickness tolerance of the protective medium on the sensor surface.
The company was later acquired by Authentec, which further developed radio frequency sensor technology based on this. Authentec's patents and technology were one of the reasons Apple acquired it.
The validity period for US invention and design patents is as follows: For patents filed before June 8, 1995, the protection period is 17 years from the date of grant; for patents filed after June 8, 1995, the protection period is 20 years from the earliest filing date. Note that this is calculated from the earliest filing date. Furthermore, if the scope of protection involves a composition or a method of using that composition, the validity period can be extended, up to a maximum of 5 years. For example, Harris's US9540526 patent is valid until at least May 16, 2017. This means that during this period, any solution using active capacitive fingerprint recognition will be incompatible with this technological barrier.
Although capacitive fingerprint sensors and patents have existed for over 20 years, their penetration ability is relatively weak and cannot meet the needs of today's mobile phone applications. To improve the penetration ability of fingerprint sensors, fingerprint sensor design companies have been improving upon existing capacitive sensor patents and applying for patents. An effective way to improve the penetration ability of fingerprint sensors is to eliminate the parasitic capacitance and noise of the pixel. Based on the methods used to eliminate pixel parasitic capacitance, two main schools of thought have emerged.
1. Boost and Modulation Approach
Simply put, this technology uses a floating ground design and requires the use of external driver or modulation chips to eliminate the parasitic capacitance of the pixel.
The first method is voltage boosting. By adding a driver chip, the voltage is increased from the usual 2.8~3.3V to 12~16V. While this method greatly improves the chip's penetration capability, it also brings the disadvantage of higher power consumption. In addition, some people are sensitive to voltage and may experience a stinging sensation during use. Currently, chip manufacturers that use voltage boosting to improve penetration capability mainly include Silead and Sunway.
The boost method had drawbacks in user experience, so a second method was adopted: modulation. This method eliminates the need for a high voltage on the metal ring, instead using a frequency (b). A modulation chip is then added to modulate the power supply to a negative voltage. This improves penetration and solves the stinging sensation problem, but it still doesn't address the high power consumption issue. Currently, chip manufacturers using the modulation method include Apple, FPC, and Goodix.
The following diagram is an illustration of the FPC patent application. The driving signal 32 shown in the diagram is generated by the peripheral driver chip. Its advantage is that most of the parasitic capacitance of the PIXEL can be eliminated by using the floating ground design of the PIXEL and cooperating with the peripheral driver chip. Its disadvantage is that the floating ground design means that the peripheral driver chip cannot be integrated into the fingerprint chip, the design structure is complex, the peripheral driver chip is required, the yield is low, and the cost is high.
Based on Harris's capacitive sensor technology, Apple is currently leading the way. Goodix, on the other hand, has increased the pixel depth, upgrading Apple's 8-bit pixels to 16-bit. Both boost and modulation methods require an additional chip to address the penetration issue. These two methods share the characteristics of high penetration, high power consumption, and high cost (multiple chips).
2. Pure capacitor type
The origins of pure capacitor technology lie in the principle of parallel-plate capacitors from the 1980s. Since 20 years have passed, the patent for this technology has expired and become a public domain patent, therefore there is no risk of patent infringement.
Pure capacitive technology uses a single-chip design with a charge pump structure, eliminating the need for external driver chips to eliminate parasitic capacitance and also eliminating the need for metal rings. Currently, chip manufacturers using pure capacitive technology include Feinger, Mairui Microelectronics, and Aegis, while Goodix has also recently launched a single-chip product.
The image below is an illustration of the patent application filed by Feinger. The PIXEL design requires only a few switching transistors, along with a feedback circuit and corresponding processing timing, to eliminate parasitic capacitance, FPNs between PIXELs, and power supply interference. Its advantages include a simple circuit structure, no need for external driver chips, low power consumption, a yield rate exceeding 99%, and low cost.
In summary, sensors using pure capacitive technology have advantages such as low cost (single chip) and low power consumption (one-quarter that of RF sensors). However, this technology places higher demands on semiconductor process technology and sensor design teams.
A comparison of the two main technologies for fingerprint recognition sensors reveals a price difference stemming from the technology itself: fingerprint sensors using boost or modulation technologies require multiple chips, resulting in higher costs; while fingerprint sensors using pure capacitive technology require only a single chip, leading to lower costs. As semiconductor processes advance and sensor design teams improve their skills, the cost advantage of pure capacitive technology will become increasingly apparent.
