BY JANE ADAMS
Traditionally, contactless technology has been used primarily for mass transit. However, during 2002 both Visa and MasterCard announced contactless payments specifications based on standardized contactless technology. Its use has also been growing in the access control and network security, and authentication markets.
While contact cards need physical contact to communicate, contactless cards receive their power from– and exchange signals with–the reader over an inductive or radio frequency link. Signals–which include power, the clock signal, and data to transfer to and from the card–are received and transmitted via an antenna embedded in the card. Information is stored and processed via the card’s embedded integrated circuit or chip.
Manufacturing a contactless card: From silicon to finished product
Just like contact cards, contactless cards contain chips–memory chips or microprocessor chips–that store, protect and process information. The chip vendors that sell contactless smart card chips include Philips, Infineon, Inside Contactless, STMicroelectronics, Atmel, Texas Instruments, Hitachi, NEC, Fujitsu, Samsung, Sharp and Toshiba (for Sony).
As with contact cards, contactless chips are enclosed in modules. The chip is attached to a lead frame and either encapsulated in resin or via injection moulding techniques. But unlike contact chips, Contactless chips have just two connectors. These are attached to the frame and the resulting module is then connected to a copper, aluminium, or printed antenna on a foil. The foil is also known as the substrate. Together, the encapsulated chip and the antenna on the substrate are known as the inlay.
Different companies are involved at different stages of the manufacturing process. Some provide the raw silicon or chips, others provide antennas and foils, while still others create the modules, inlays, and finished cards. Some companies are involved in multiple stages in the process but it is rare for any one company to take the process from start to finish.
Contactless chips can also be used in hybrid and dual interface cards. A hybrid contains two chips – one contact and one contactless, with separate means of communication with the outside world and no direct connectivity between the two. A dual interface card, on the other hand, contains just one specially designed chip with both contact and contactless interfaces. These chips have the standard eight contact pads on the front of the module and the two contactless contact areas on the back of the module. Dual interface chips are manufactured by Philips, Infineon, and ST Microelectronics.
Types of memory used in contactless cards: EEPROM, FRAM, and FLASH
The most widely used memory type for data storage in both contact and contactless smart cards is EEPROM. Philips chips for example offer EEPROM in sizes ranging from 512 bits to 16 kilobytes with a 64 kilobyte version planned for Spring 2003.
Contactless microcontrollers also contain two other types of memory. Nonvolatile Read Only Memory (ROM) stores the card operating system and application code. ROM is set during the manufacturing process and cannot be changed. Additionally, Random Access Memory (RAM) is a volatile memory that is used for temporary storage of data during processing.
For some years, chip manufacturers have been investigating alternatives to EEPROM. The main two are FLASH and FeRAM or FRAM. Both of these are slightly more dense than EEPROM, which means that they require a smaller area of silicon. This makes the chip cheaper and less fragile. (Note: in many cases EEPROM pricing remains lower than other technologies simply because the production volumes are significantly higher.)
FRAM takes up 40% less space than EEPROM for the same amount of memory. It works faster (in less than 200 nanoseconds rather than milliseconds like EEPROM), and uses less power. All these qualities make it suitable for contactless technology. On the downside, FRAM is difficult to manufacture and it features a ‘destructive write’ process, meaning that data stored is deleted before updating it.
Silicon vendors specializing in FRAM products in the past have included Racom, Ramtron and Rohm. However STMicroelectronics and Fujitsu have recently announced that they have teamed up to develop the ST19ZR01, the first contactless smart card IC with FRAM and ROM, but no standard RAM (see ContactlessNews December 2002).
FLASH is programmable with large chunks of data in one operation, making it faster but less flexible than EEPROM, and 50% denser than EEPROM. However it is limited in the number of write cycles it allows (lower than the number possible with EEPROM) thus reducing the lifetime of the card. It can however provide an alternative to mask programmed ROM, thus considerably shortening the card manufacturing cycle. Atmel has been active in developing smart card chips using FLASH.
STMicroelectronics has also recently announced Page FLASH, a derivative of Standard FLASH technology. Page FLASH offers the same fast programming time as Standard Flash but also allows individual data words (32-bits) to be erased in a few milliseconds and rewritten more than 100K times, thereby providing all the functional benefits of EEPROM. The two types of FLASH memory can be mixed on the same silicon chip.
Many issuers of contactless card systems are not familiar with the chip vendors discussed in this review. That is because we most often deal with the vendor providing us the completed card. We all are familiar with Schlumberger, HID, Gemplus, Oberthur, and the other finished card suppliers. The key lesson is to understand that there are a variety of companies in the supply chain and be aware of the different components in the RFID products we are using or evaluating.