SLVAET3 October   2021 TPS8802

 

  1.   Trademarks
  2. 1Introduction
  3. 2System Architecture
    1. 2.1 Battery Voltage
    2. 2.2 VCC Supply
      1. 2.2.1 Connecting VCC to VBST
      2. 2.2.2 Connecting VCC to VBAT Through a Switch
    3. 2.3 MCU Supply
      1. 2.3.1 MCU Connected to VBAT
      2. 2.3.2 MCU Connected to MCU LDO
      3. 2.3.3 MCU with VCC Connected to VBAT Through a Switch
    4. 2.4 Photoelectric Smoke Sensor LED Supply
      1. 2.4.1 LED Connected to VBAT
      2. 2.4.2 LED Connected to PLDO
      3. 2.4.3 LED Connected to LEDLDO
    5. 2.5 Example Schematics
      1. 2.5.1 Smoke and CO Schematics
      2. 2.5.2 Smoke-Only Schematics
  4. 3Current Consumption
    1. 3.1 Standby Current
      1. 3.1.1 TPS8802 Standby Current
      2. 3.1.2 Microcontroller Standby Current
    2. 3.2 Measurement Current
      1. 3.2.1 Smoke Measurement Current
      2. 3.2.2 CO Measurement Current
      3. 3.2.3 Battery Test Current
      4. 3.2.4 User Alarm Test Current
    3. 3.3 Other Current Consumption
      1. 3.3.1 Boost Charge Current
      2. 3.3.2 Initialization Current
  5. 4System Power Calculation and Measurements
    1. 4.1 Power Calculation Spreadsheet
      1. 4.1.1 Power Consumption Overview Page
      2. 4.1.2 Detailed Calculation Pages
    2. 4.2 Power Consumption Measurements
      1. 4.2.1 Power Measurement Method
      2. 4.2.2 Smoke and CO System Measurements
      3. 4.2.3 Smoke-Only System Measurements
  6. 5Summary
  7. 6References

4.2.1 Power Measurement Method

The power consumption measurements are taken using a Keithley 2400 to supply power and measure the output current. A 1 kΩ resistor is installed between the Keithley 2400 output and VBAT on the EVM, and a 2200 µF capacitor is installed between VBAT and GND. The resistor and capacitor filter the current draw from the Keithley 2400. This resistor and capacitor is required for two reasons: to lower the maximum current draw out of the Keithley 2400 for more accurate measurements, and to anti-alias the current waveform. The TPS8802 current draw can have large 500-mA spikes from the boost converter that would greatly diminish the accuracy of the microamp average current measurement if the Keithley 2400 was set to measure up to 500 mA.

The Keithley 2400 has an adjustable integration time for each measurement that is best set to an integer multiple of a power line cycle (PLC) to reject power line noise. However, due to the nature of the instrumentation, there is time between each measurement sample where the current draw is not included in the measurement. This means that the large 500-mA current spikes may be missed if no anti-aliasing filter is used. One approach is to set the integration time to the maximum value supported by the instrument for the best accuracy, but this would require a very large anti-aliasing filter due to the low sample rate. Therefore, the optimal integration time is 1-PLC.

The last step is to take multiple measurements and average them together. This is performed using the Keithley 2400 repeating filter function. This digital filter averages a programmed number of measurements and outputs the value when a new set of measurements is taken. Ideally, the total time taking the set of measurements should equal an integer multiple of the smoke alarm measurement interval (the amount of time between each smoke measurement in the smoke alarm).

It was experimentally determined that the Keithley 2400 sample rate for 1-PLC integration time is 3-PLC, and the sample rate for 2-PLC integration is 6-PLC. Therefore, it was chosen to set the smoke alarm measurement interval to 3.333 seconds (200-PLC), the integration time to 2-PLC, and the number of averages to 100. A set of 100 current measurements takes 10 seconds (600-PLC), equal to three smoke alarm measurements.