COMMUNICATION TECHNOLOGIES MODERN ENERGY-METER NETWORKS CORRECT CROSS-WIRE FAULTS AUTOMATICALLY By Thomas Kugelstadt, Texas Instruments Worldwide energy metering networks use differential data transmission based on RS-485 technology to span long distance data links. To overcome large ground potential differences between remote bus nodes, transceivers are galvanically isolated from their node circuitry. An e-meter network is a master-slave system with a host processor (master) that is located in a control center. The master addresses multiple slave nodes which are located in the end user premises along the bus. A single network typically comprises up to 60 nodes, making the potential for cross-wire faults of the twisted-pair bus cable during installation rather high. To assure reliable data transmission, modern transceivers apply automatic polarity correction (POLCOR) of the bus signal polarity. Figure 2: Polarity correction timing after power-up Figure 1 shows a typical e-meter network with POLCOR transceivers. Each slave transceiver contains an internal timer reversing the polarity of transmit and receive data, if the receiver output state is logic low for a specified time span. Due to the wide operating temperature range of POLCOR transceivers, this time varies between a minimum of 44 ms (t FS-min ) and a maximum of 78 ms (t FS-MAX ). The master node determines the signal polarity on the bus via a failsafe-biasing resistor network (R FS and R T ). The transceiver in the master node requires no polarity correction. Polarity correction might, therefore, be initiated by a constant bus voltage present for as short as 44 ms duration. Hence, data strings of consecutive 0-bits must be shorter than 44 ms to prevent false polarity correction. The slave transceivers, however, do require integrated polarity correction to sense and, if necessary, correct the bus signal polarity during bus idling. That is when no node is actively driving the bus. On the other hand, polarity correction automatically initiated after power-up requires the bus-idle voltage to be present for at least 78 ms to ensure the polarity correction process is complete. After the initial correction the bus polarity status is latched within the transceiver and maintained for all following data transmissions. This means that switching between drive and receive modes will not lose the bus polarity status. Figure 2 shows polarity correction after a power-up sequence. During power-up the receiver output, R, is undetermined. Once the slave node supply, VSS, is stable, the bus must idle for at least t FS-MAX to ensure the completion of a polarity correction. Because of the cross-wire fault, the positive bus voltage at the master’s failsafe network, V AB-M , appears negative at the transceiver input. Thus, after completion of t FS-MAX the transceiver’s internal polarity is switched to invert receive and transmit data. Hence, the negative input voltage, VAB-S, is converted into a positive output voltage. Long data strings of ones and zeros exceeding the minimum polarity correction time t FS-MIN also can trigger a polarity correction. To prevent such incidents, e-metering standards such as DL/T645 add a fixed-bit pattern to the transmit data within the driving node and subtract it from the output data in the receiving node. Figure 1: Typical e-meter bus with polarity-correcting transceivers 22 Bus loading Because an e-meter bus uses galvanically isolated transceivers, ground potential differences between bus nodes do not METERING INTERNATIONAL ISSUE - 5 | 2014