One article tells you how to estimate the power supply time of wireless sensor batteries

著者：Iflowpower – ポータブル電源サプライヤー

Wireless sensors serve excellent vision for monitoring environmental conditions or industrial plants and machinery equipment. Because they are easy to install, they can deploy in a variety of environments. In the past few years, with the extensive deployment of "IoTs", the use of wireless sensors will present explosive rise.

But a most important factor that limits the use of wireless sensors is that their battery life is very limited. When the wireless sensor works completely on the battery, once the battery is exhausted, the sensor becomes a waste. If you are designing wireless sensors powered by the battery, you must ensure that it works for a certain time, to deal with countless puzzles.

The usual method is to use electricity when making the necessary activities, and other time the sensor will enter low power usage mode. Wireless sensors can be divided into a series of activities, each activity is required to consume certain energy within a period of time. The most common activities include: ● Wake up, measure, and add data to information ● Power supply to RF power amplifiers, send information, and turn off the RF power amplifier ● In bidirectional sensors (send and receive): Wake up, to receive Power-on, receive signals, solve data, according to information action, and disconnect it again, there are many operations that consume battery power.

Extend battery power time, the simplest practice is to use larger batteries, with larger capacity batteries. However, your customers may want the sensor to be small, high performance (so they can send large amounts of data, and have smart / data resolution locally). Obviously, the customer's expectations and the simplest way to handle this problem in the battery supply time is completely laid.

Figure 1: Current level of wireless sensor in three feet. How do engineers estimate the battery power supply time? As a design engineer, you have to consider the balance between the battery volume and the wireless sensor function, so that the small battery can be used to play the best performance, and continue to have a long time. The optimization process must first understand energy demand.

Collecting data about energy is the first step in characterizing device performance. The battery has a pre-embankment of energy (Watt (WH)) and capacity (MAH)). If you understand how much the device is working, you can calculate the power supply time of the battery.

Battery Power Supply Time (Hour) = Battery Capacity (WH) / Uniform Discharge Power (W) The energy of the battery is also the product of its rated voltage (V) and capacity (AH). The rated voltage is the midpoint value of the battery discharge curve determined by related experience, which can be correctly related to the energy and capacity of the battery. Based on this meaning, the power supply time of the battery can also be determined by this formula: battery power supply time (hour) = battery capacity (AH) / uniform discharge current (a) Figure 2: Keysightn6781ASMU can accurately measure across the current current level.

However, when the device is actually working, the battery power supply time is usually shorter than your calculation. The opinions we have often heard are: "The quality of this battery is too bad!" Some big battery brands usually supply specific technical indicators and explain that between the same type of batteries, the capacity will usually be 5%. 10% difference.

But even in accordance with the conservative estimated battery capacity, the battery supply time is often not reached. The length of the device work is shorter than what we expected. Why is this this? Do we be estimated that the power usage is correct? May not correct.

Let us explore this problem. Measuring the complexity of dynamic current consumption In battery power supply devices such as wireless sensors, in order to save power, the sub-circuit of the device is only activated. Engineers design the device to the smallest sleep mode in most of the time.

In sleep mode, only real-time clock operation. Devices will wake up regularly to perform measurement. Then send the collected data to the receiving node.

Different working modes will result in current consumption to a wide range of dynamic ranges from sub-ua to 100 mA, with a ratio of 1: 1,000,000. Traditional measurement technology and its limitations should measure current, well-known approach is the use of a meter meter function. Using a modern digital multimeter is like a good current measurement accuracy, its technical indicators are determined according to the fixed range and relatively static signal level, because the wireless sensor has dynamic current consumption, is not very suitable for use of a multimeter.

Measurement. Figure 3: Data Recorder: All samples are uniform to the continued sampling cycle. No sample loss.

For each sampling cycle, you can also supply minimum values ​​and maximum values. The digital multimeter is connected in series between the battery and the device to measure the current. Due to the influence of the sensor activation cycle or send mode, we will see unstable reading from time to time.

We understand that digital multimeters have multiple ranges, using automatic range to choose the most appropriate range, and supply best precision. However, the digital multimeter is not all advantages. Automatic range Time to change range and stable measurement results.

Automatic range time is usually 10ms to 100ms, longer than transmission or activation mode. Therefore, the user wants to disable the automatic quotation function, manually select the most appropriate range. The measurement principle of the digital multimeter is to insert a splitter in the circuit and then measure the pressure drop on this splitter.

Typically, you want to measure small currents, you can use high-resistance junction and select low range; to measure large currents, use low-resistance jidopters and select high-grade range. This pressure drop is also called a load voltage. Due to this pressure drop, not all battery voltages can reach wireless sensors.

The low range is the most accurate, you can measure sleep current, but you will withstand a certain voltage during current peak, and this may even cause device reset. In fact, we finally make a concession, use large currents to ensure that the device can work properly when the device is in the current peak. This concession allows us to solve the peak current, and you can measure sleep current, but the price is also very large.

Since the bias error is specified based on the scale of the range, it will seriously affect the measurement result of the low current level. This error may be 0.005% of the 100mA range, which is 5μA, but about 50% of 10μA, about 1μA current is 500%.

The device is in this current level in most of the time, so this error has a huge impact on the estimation of battery power supply time. After measuring the low current level in sleep mode, we will not measure activation pulses and transport pulses, including current level, and time of the sensor at this level. The oscilloscope is an excellent tool for measuring signals over time.

However, we have to measure the level of levels of mass, the current probe is unable to compete for this task due to limited smacks and drift problems. Good clamp probe has 2.5Marms noise, often repeatedly performing a zero compensation program.

The current probe measures the electric field of the line, so the secret of improving the weakness is to pass the same line many times, thereby adding a new magnetic field several times - thus adding a few times new, so that we can better measure the current. With this approach, we can capture current pulses for activation time and transmission time. Even in the activation and transmission time, the current will change the level: it is like a burst of bursts composed of high and low levels.

In order to correctly calculate the uniform current, this waveform is to be exported, all measurement points are integrated to obtain a uniform value. Figure 4: 200 multi-second runtime current consumption record is a new field of vision for the dynamic current consumption of the observation device. The oscilloscope can capture a single burst very well.

However, if you want to verify how many times in the sensor activate in a period of time, how long it will launch TX burst, and the measurement will be more complicated. The oscilloscope can complete measurements in a short period of time, but the sensor has a few minutes or hours of operating cycle, capture and measurement more complex. The keySightn6781a power / measurement unit (SMU) for the battery power consumption decomposition, overcomes the limitations of traditional measurement methods: seamless current and long-term intervals data records.

This SMU module can be used in combination with Keysightn6700 small modular power system or N6705 DC power decomposition. The seamless current is a patented technology that allows the SMU to change the measurement range, and the output voltage remains stable, and does not cause pressure drop due to the variation of the range. This feature allows you to measure peak using high current quantities, measure sleep current using 1MAFS range (with 100NA offset error).

This low bias error (100na bias error is associated with 1μA of 10%, about 1% 10μA), and the order is superior to the traditional digital multimeter. The seamless current is combined with two digitizers to measure voltage and current with 200KSA / S (5US time discrimination rate). With the full-time discrimination rate, you can capture and display digital measurement results above 2 seconds, the lower the discrimination rate, and the time proportional is added.

However, when performing long-term measurements, the built-in data logger of the Keysightn6705B modular DC power decomposition instrument will measure the result of 200KSA / s within the integration period (20 uz to 60 seconds) specified, and will not lose the integral cycle. Any sample. Since the data logger is unparalleled, all samples will be classified into an integration cycle or the next integration cycle ------ Will not lose any sample.

Through the data logger, engineers can measure the current and energy consumption performance of wireless sensors in running time for up to 1000 hours. Measuring sleep current, as long as the cursor is placed, you can read the measurement reading. Figure 4 is a single collection of long-term measurements; we can get a complete current consumption map, and accurately measure sleep current to 599NA.

With pan and zoom function, we can see the maintenance time of current level and each power level. Details that use traditional measuring tools can not be seen in this chance to list and measure. The rear edge pulse used in Figure 4 is determined is a distinctive example.

The software unveiled this unexpected secret: 3.3μA's uniform current, the device consumes pulse energy at about 90μA peak, the continuation time is 500 ms. The current consumption is added to 599NA sleep current, and the result reaches 730NA, which is 22% higher than our expected current.

This accident may be one of the fact that our low-estal energy requirements is caused to make battery power supply than expected. When optimizing wireless sensor power consumption, understand these specific information about engineers is very helpful. When pursuing the balance between the user experience and battery consumption, and answering, "I should send a message every 5 seconds, every 5 seconds or every 10 seconds?" What is the problem of sending a packet to consume how much energy is very important.

Engineers can accurately estimate the impact of battery consumption on any curing software, and verify by actual measurement within rational time. Easily carry the Joule measurement Joule is very useful when estimating battery power time, because each activity will consume a certain amount of energy. We can also use Joule / Send Bit to compare the performance of devices.

But engineers rarely use Joules because they should calculate them. Use Keysight14585A to control and decompose software, you can measure the energy of Jouo units. For example, you can measure how much energy is to be transmitted to the data packet captured in trigger measurements?.

This advantage comes from the use of two digitizers simultaneously sampling voltage and current to achieve point-by-point power measurement. Joules can be easily read as a value between the cursor, the designer can even further mean the Joule / emission bit. .