Industrial RFID systems are composed of a read head, controller, and transponder. The transponders, or tags, are used to carry information from one process to another. The RFID tags could require a fixed pallet number that is programmed once, a fixed unique identifier that can’t be written such as a license plate, or a large database of information that describes everything about a specific part or manufacturing process. This post focuses on high-capacity tag databases.
High-capacity RFID tags have a memory capacity and block size. It is important to understand the limitations of a specific chip as well. The memory type may differ. Some manufacturers use EEPROM memory, and others use FRAM. Both types of memory have unlimited read capability, but FRAM memory can be written virtually an unlimited number of times (10,000,000,000). Also some tags have different data retention times — some are 10 years while others are 30 years. Data retention is the amount of time the data will remain unchanged on a tag. This can be a useful specification to look at if your tagged product will be in storage for a long period of time. Even the TID code, which is perma-locked [no longer writable] at the factory, will only remain unchanged for the defined data retention time. This retention time, of let’s say 10 years, is determined at a defined temperature (for example 131 °F). If your application operates only at room temperature, the data will be retained for a lot longer.
Let’s look closely at the memory size and block size and define what they are:
This is the total amount of read/write memory of the chip. Even though a tag may have a lot more memory on board, some of the memory is used to hold CRC checksums, read-only code data, application family identifiers, and other chip-specific information. Only a certain portion of the tag's memory can be used to read and write user data. For example, the Fujitsu chip MB89R118 has 16 kbit of memory; however, only 2000 bytes are available for user memory. These 2000 bytes equate to 15.625 kbit of memory or about 97 % of the total as stated on the data sheet. If you had thought that the 16 kbit was all user memory and purchased the tag because of it, you may have been disappointed.
This is the smallest memory size that can be written. Typically, but not always, the larger the tag size, the larger the block size. This is true for the high-frequency standard ISO15693, which requires that the block address be less than 255 (256 blocks addressed 0…255). You have to remember that every block has to be individually addressed. So if a block size is small, for example 1 byte, then it would take 2000 blocks or addresses to communicate to a 2000 byte tag. This would not be possible under ISO15693 where the max is 256 blocks. So in consequence, the 2000-byte chip MB89R118 has 8-byte blocks with a total number of 250 blocks (8 bytes/block x 250 blocks = 2000 bytes).
Here is a table of just some of the high-capacity tags on the market.
The block size has another consequence. If the smallest addressable size is 32 bytes, like the MB89R112 from Fujitsu, then this is the smallest amount of data you can read or write at one time. Let’s look at a situation where you have five pieces of data each 32 bits long. Let’s call this information Model Number, Serial Number, Born on Date, Expiration Date, and QC Check.
If you want to access each one of these five fields individually and write them individually, then you have to put them each into separate blocks. This may leave a lot of extra unused data if you use a chip that has a very large, 32-byte block size.
If the data is written all at one time, then the addressability of each piece of information isn’t important, and the five pieces of information can be put in the first five 32-bit words on the tag, filling all of the first block and a little of the second.
Remember that the high-capacity chips are using different ISO standards, which means they need different read heads. The tags to ISO 15693 should use all IQH1-… read heads. Tags to ISO 14443A have to use read heads starting with IQH2-…. And lastly, all UHF tags to EPC Class 1 Gen 2 must use read head models starting with IUH-…. The UHF read heads are also country specific, so make sure you pick the country-specific read head designed and tested for that country.