The rfid tags how to write the device, initialize the coding and resolve the issuer process
Last year, at an apparel warehouse project, more than 3,000 blank RFID tags were waiting to be programmed. The customer's purchasing manager held up the reader and asked, “Can't we just plug it into the computer and start writing?” I shook my head. Without setup, the tags probably would not accept the data or would be unreadable afterward. We spent half a day getting initialization and issuance working end to end, then encoded and associated all 3,000 tags in under two hours. It made me realize that many people still think a reader and tags are all they need. In reality, every step from connecting the device to writing the data needs attention.
We'll start by matching reading and writing devices, then we'll talk about data.
The RFID tags write in two parts: a reader (also called a reading and writing module or card-maker) and an associated antenna. The reader is responsible for the generation of radio frequency signals and data processing, and the antenna is responsible for their transmission and reception, which are connected by SMA interface or internal PCB lines.
Common desktop carders use USB interfaces for power and communication, or plug-in. This type of device is suitable for door stores, offices, small bulk card scenes, a device that writes a label at a time, with speed between 50 and 200 milliseconds. If it is a mass card, like a warehouse, which sends thousands of tags in one time, using a fixed reader to match the antenna, the antenna gain is generally 6-8dBi, covering about half to one metre, and the label needs to be placed on top of the antenna when you write it.
Once the hardware is installed, the software pit begins. Windows requires a corresponding USB trunchean drive or a software development toolbox provided by the manufacturer to drive the computer to see the corresponding COM port in the device manager, which counts as a layer of hardware connection. We've run into customers with a laptop that is still unrecognized, and finally we find the USB interface inadequate for power supply, and then replace it with a back-up interface -- it's not uncommon to leave the USB belt intact.
Initialization: the labels have changed from "Blank" to "identity."
RFID tags are not entirely blank when they leave the factory. UHF tags generally have four memory banks:
- EPC Sector(electronic product code): the only global identifier used to store the label, and then written as the "ID number" for the item;
- TID sector(label identification code): The label was solidified at the time of the plant and is unique to the world and cannot be modified;
- User district(User data area): free-to-use writing, with production dates, batch numbers, custom codes, etc.;
- Reserved district(Reserve area): Store access passwords and destroy them for use in rights control.
Initialization involves writing an identifier to the EPC bank and setting an access password to prevent accidental or malicious changes. EPC identifiers must follow a consistent scheme. A common approach is to allocate a vendor code prefix to the project and append a serial number or business code, such as "EA.001.00001" through "EA.001.03000." The first part identifies the project and the last part is the serial number. Keep the code within 96 bits (12 bytes). UHF tags generally have 96-128 bits of EPC memory. A code that is too short wastes space, and one that is too long reduces read/write efficiency.
A common practice is to set a four-byte access password and a four-byte kill password. Passwords can be shared across tags or unique. Bulk issuance generally uses shared passwords for efficiency. High-security uses such as financial services and ticketing need separate passwords for each tag, recorded in a database.
Use software to write in the data.
Once both the hardware and code rules are ready, you will be able to write to the actual content.
- A self-carrying tool for the reader, which is simple and suitable for testing;
- (a) Universal RFID intermediates, which support multiple reading and writing protocols (commonly ISO 18000-6C, EPC Class1 Gen2);
- Clients develop their own up-to-date systems (i.e., the issuer ' s own business management software) and embed them into business processes.
In practice, the software interface usually has several core parameters to set: Reader power, antenna port, target storage area, initial address, write length, access password. Power generally starts from 20dBm (100 mW), reads and writes in close proximity and adds up; access passwords can be left empty when writing the EPC area (many labels are defaulted without password), but writing for User or Reserved area must first verify the password, otherwise it will be rejected.
The post is part of our special coverage Global Voices.We have to do a check-up., confirming that the written data exactly matches the original. Skip this step and troubleshooting unreadable data after tags are applied becomes very time-consuming. In warehouse projects, we randomly read back 5%–10% of the tags after encoding each batch. The batch is complete only if every sample passes.
Card process: from single-shot testing to bulkwork
The complete card process is roughly four steps:
- Single Test(a) A test to confirm that the reader, antenna, software parameters are correct and that the data is correctly read back;
- Rule Configuration(a) The inclusion of EPC code rules, password policy, writing areas and other parameters in the issuer software;
- Batch writing: Put the labels one by one in the antenna, triggers writing. The design of the bulk delivery card is critical -- not too fast when the label goes past the antenna, otherwise the reader cannot finish handshakeing and writing. The label is usually required to stay at more than 100 milliseconds in the reading and writing area, with a speed control of 2-3 sheets per second;
- Bind business systems: The entries are bound to the EPC's info on the business system, such as the dress code, the warehouse's position. This is usually done by Excel import or interface, the issue card software exports the coded and the business system receives it into the library.
After the whole process is running smoothly, the efficiency of the bulk cards depends on the bottlenecks. Manual tags can distribute about 200-400 copies per person per hour; and with automated card devices (like a label structure), they can refer to thousands of them per hour.
A couple of easy-to-touch pits.
The issuer project has experienced several typical "failure" cases over the years, and is listed here for information:
- Label on the metal surface and write it directly. Metals reflect frequency signals, which make it impossible for readers to stabilize communication. Such scenarios require metallic labels or a suction material between the tags and the metal.
- Multiple labels into the reading and writing areas simultaneously. Readers read and write responses to multiple labels, but only one of them is written, causing data confusion. The solution is to open the tag impact-proof algorithm in a single writing mode or control the antenna direction so that only one can be read at a time.
- EPC prefix duplicateThe standard is that each project apply to the relevant agency for an independent manufacturer code or use the first two as a project distinction within the project.
- Writing successfully but not read. The common reason is that access passwords are set but not entered when readbacking, or the reader is too low. It is suggested to cross-check with another device after writing to avoid a single device being covered up.
It's at the end.
The RFID tags write and issuer, simply put data in the label, say complex but involves hardware connections, protocol understanding, code code, lotwork, business binding multiple links. Avoiding each pit ahead of time, the whole process is stable.
If you are going to do your haircard, it is recommended that you start with a single test and then go through the lot; if you have been unable to read at the mass writing stage, you can run power, you can direct the antenna, you can type the label three variables, most of which can be traced to the reason. Shanghai-based Select Technologies has made a lot of storage in the RFID reading and writing equipment and the haircard system, retailing, asset management landing projects, you can break down when you have problems with the relatively complete choice of readers and you can design the haircard.

Ms. Lei
Mr. Pang