Design Article
Enabling next gen smart utility meters
Milen Stefanov, Texas Instruments Inc.
10/2/2012 11:24 AM EDT
3. HW architecture of a wM-Bus RF subsystem
A wM-Bus subsystem can be used for both home (HAN) or neighbor area network (NAN) communication, depending on the frequency band or the transmit power used. An example of a HAN implementation is a smart meter with a 868MHz RF link, using either S-, T- or the new C-mode of the wM-Bus protocol. In a real world application, the RF chip (transceiver for bi-directional or transmitter-only device for uni-directional communication) with transmit output power of +10 to +12dBm (measured at the antenna port) and the antenna gain (up to +2dBi) itself will provide the optimum solution within the ETSI 300220 limit of +14dBm EIRP.
The NAN solution for a smart meter will typically be at 169MHz, where the +27dBm EIRP limit enables superior range coverage. Also, the C2-mode (only allowed from data-collector-to-meter) at 869.525 MHz uses a +27dBm limit sub-band and is suitable for NAN applications. For the NAN system solutions, the addition of an external power amplifier is needed, as no integrated RF transceiver chip today can offer +27dBm transmit output power.
The two variants for the hardware of a wM-Bus subsystem, based on the EIRP power limit, are:
1. Without external power amplifier (blue block in Figure 5 is dropped)
2. With external power amplifier and optional LNA (e.g. TI’s CC1190 @ 868MHz or dedicated RF front end @169MHz)
In fact, the RF sub-system in Figure 5 represents a complete RF module, which is typically connected over UART to the main MCU. In this case, the dedicated MCU (orange block) will run the wM-Bus stack as well as a serial protocol application to connect to the main application MCU.
The second option is to have the wM-Bus stack running on the application MCU and connecting the RF device over SPI interface, avoiding the serial protocol application altogether (orange block in Figure 5 is omitted). It is very important to clarify the hardware and software partitioning inside the smart meter upfront, as there are advantages and disadvantages for both architectures, the main considerations being:
1. Certification of the metrology part and the wM-Bus stack
2. Real-time requirements: both metering and RF communication are time-critical tasks and could sometimes be an issue when the MCU should simultaneously run metrology and communication tasks.
3. Field upgrades of the Firmware: for RF communication and/or the complete Meter.
For those reasons, manufacturers often prefer to keep their systems modular by separating the metrology and the communications functions (a two MCU approach). It can achieve much higher flexibility in the device selection of MCU and RF devices thru independent price and/or performance optimization. Typically, there are multiple pin-compatible MCUs or RF derivatives that offer higher performance or additional features.
The single MCU solution could save some cost, but will usually make the metrology part certification more complex, as the metrology firmware code needs to be secured against tampering or other sources of manipulation or failure.
A wM-Bus subsystem can be used for both home (HAN) or neighbor area network (NAN) communication, depending on the frequency band or the transmit power used. An example of a HAN implementation is a smart meter with a 868MHz RF link, using either S-, T- or the new C-mode of the wM-Bus protocol. In a real world application, the RF chip (transceiver for bi-directional or transmitter-only device for uni-directional communication) with transmit output power of +10 to +12dBm (measured at the antenna port) and the antenna gain (up to +2dBi) itself will provide the optimum solution within the ETSI 300220 limit of +14dBm EIRP.
The NAN solution for a smart meter will typically be at 169MHz, where the +27dBm EIRP limit enables superior range coverage. Also, the C2-mode (only allowed from data-collector-to-meter) at 869.525 MHz uses a +27dBm limit sub-band and is suitable for NAN applications. For the NAN system solutions, the addition of an external power amplifier is needed, as no integrated RF transceiver chip today can offer +27dBm transmit output power.
The two variants for the hardware of a wM-Bus subsystem, based on the EIRP power limit, are:
1. Without external power amplifier (blue block in Figure 5 is dropped)
2. With external power amplifier and optional LNA (e.g. TI’s CC1190 @ 868MHz or dedicated RF front end @169MHz)
Figure 5: Block diagram of a wM-Bus RF subsystem based on TI MCU and RF devices
In fact, the RF sub-system in Figure 5 represents a complete RF module, which is typically connected over UART to the main MCU. In this case, the dedicated MCU (orange block) will run the wM-Bus stack as well as a serial protocol application to connect to the main application MCU.
The second option is to have the wM-Bus stack running on the application MCU and connecting the RF device over SPI interface, avoiding the serial protocol application altogether (orange block in Figure 5 is omitted). It is very important to clarify the hardware and software partitioning inside the smart meter upfront, as there are advantages and disadvantages for both architectures, the main considerations being:
1. Certification of the metrology part and the wM-Bus stack
2. Real-time requirements: both metering and RF communication are time-critical tasks and could sometimes be an issue when the MCU should simultaneously run metrology and communication tasks.
3. Field upgrades of the Firmware: for RF communication and/or the complete Meter.
For those reasons, manufacturers often prefer to keep their systems modular by separating the metrology and the communications functions (a two MCU approach). It can achieve much higher flexibility in the device selection of MCU and RF devices thru independent price and/or performance optimization. Typically, there are multiple pin-compatible MCUs or RF derivatives that offer higher performance or additional features.
The single MCU solution could save some cost, but will usually make the metrology part certification more complex, as the metrology firmware code needs to be secured against tampering or other sources of manipulation or failure.
|
|
|
|
|
Navigate to related information
WKetel
10/3/2012 9:18 PM EDT
I see a whole lot of discussion about the technology for the communication with the smart meters but only a small amount about how this is going to save enough energy to be worth the effort. Of course, the very first use will be to raise prices a lot just when everybody wants power the most. We all know that is the primary goal of the two way system. The other goal is to eliminate the expense of manual meter reading, which will reduce the utilities costs quite a bit, but will not reduce our price for power at all.
The way to eliminate all of the privacy concerns is to keep the specific minute-by-minute consumption data in the meters, and to just totalize the product of KWH and the rate at the time of consumption. Then the utility could poll the meters to determine the charge for that month's power. No customer information except for the total charge would ever need to be sent out, just as in the past when meters were manually read. The utility has no need of individual household consumption at any particular time, only block by block, or subdivision by subdivision. The assertion that they need to know the individual rate of use at any particular instant is not valid.
Sign in to Reply
GPBobby
1/18/2013 11:06 PM EST
Data is only useful to the extent that it can alter some outcome, in this case, a customer's use of electricity. Almost no electrical customer even knows that their new smart meter is capable of turning off their power. The above article states, "to enable utilities to ease peak power demand..." where ease should be read as: a) a total disconnect; b) a rolling brown out; c) by interfacing with your own personal in home network to turn off selective appliances such as A/C, dryers, heat pumps or any other device which has built in communication capability.
I can see a small business in modifying newer appliances (if possible) to make sure they cannot communicate. The polite way to modify power consumption is to let the consumer know (by checking an Internet site at their convenience) their current charge per kWh. Then let them decide what devices to turn off.
Sign in to Reply
pekon_
1/24/2013 11:42 PM EST
But if this was so lucrative, then why did google retired their power-meter service ?
http://www.google.com/powermeter/about/
Its just that there isn't enough eco-system built around to manage|optimize energy. And also its usage is expensive.
Many companies have smart-metering products from long, but lack of government urgency, and tough norms are also pulling this market low.
[pekon]
Sign in to Reply
GermanInTI
2/27/2013 4:46 PM EST
Not only MCU but also memory is a key factor for smart utility meters. TI is somehow leading the MCU part but not the memory, suprisingly even having a great technology called FRAM in Texas. FRAM makes the meter design extremely simple and flexible. Fujitsu and Ramtron leads the memory market for smart meters with FRAM. TI wake up...
Sign in to Reply