SLVAFN0 august 2023 MSPM0L1306 , MSPM0L1343 , MSPM0L1346 , TPS8802
Currently available, threshold based, smoke alarms do not have the capability to distinguish between certain types of smoke particles (for example, flaming polyurethane) and nuisance sources like from cooking, steam from a shower, and so on. This application report describes a dual chip design, using the TPS8802 analog front end (AFE) ASIC, which is a UL recognized UTGT2.S36499 smoke alarm component, and the Arm Cortex-M0+ MSPM0L1306 microcontroller. This design enables making the distinction between real sources of smoke and nuisance sources to satisfy new UL217 requirements for smoke alarms. This document also includes test data showing measurements under simulated smoke and water vapor conditions.
Demo source code for this design is available in the latest MSPM0 SDK. An optional graphical user interface (GUI) allows developers to configure the AFE and observe the system response in real-time. The GUI is published in the TI Cloud Tools gallery.
LaunchPad™ is a trademark of Texas Instruments.
Arm® and Cortex® are registered trademarks of Texas Instruments.
All trademarks are the property of their respective owners.
The latest changes to the UL217 standard (eighth and ninth editions) mandate that smoke alarms have the ability to distinguish between smoke from real fire and smoke from nuisance sources such as cooking, steam from showers, and so on. Smoke from these nuisance sources tend to contain particles sizes much smaller than those found in sources from real fires. However, flaming polyurethane is an exception, where the particle sizes in this type of smoke consists of sizes in the upper range as those found in nuisance sources. This causes a change to the typical single wavelength, threshold-based detection algorithms for reliable smoke detection and reduced false alarms.
Two prevalent photoelectric sensing techniques that are used for particle size estimates and smoke type determination are LED wavelength and scattering angles. The signal response of light scattering configurations in smoke detection follow Mie scattering physics. It follows that LED’s with low wavelength, such as blue, allow for increased signal response and therefore higher signal-to-noise ratio (SNR) for sensing smaller particle sizes typically found in nuisance sources. In order to detect particle sizes in the range of 50nm average diameter to approximately 1µm or larger average diameter, it is advantageous to use LED’s with different wavelengths to cover the range of particles sizes expected, typically IR and Blue. Measurement of the scattering response at different scattering angles, such as back scatter (scattering angle <90°) and forward scatter (scattering angle >100°), allows the estimation of particle size. These two techniques together allow for a robust, multi-criteria approach to distinguish between real sources of smoke and nuisance sources.
A solution that includes the necessary hardware for distinguishing between real sources of smoke and nuisance sources is shown in Figure 1-1. The two LED drivers allow for use of LED’s with different wavelengths as well as different scattering angles depending on the optical chamber design.
The optical chamber used in this design is configured for forward scattering measurements of both LED’s with a single broad-spectrum photo diode (PD). However, the TPS8802 PCB accommodates different optical chamber designs so that different combinations scattering angles and wavelengths can be used with the same hardware.
While set up for demo purposes, the hardware is intended to show the capabilities of TPS8802 together with the MSPM0 micro controllers. As such, the firmware and GUI for this demo does not include the use of the discrete back scatter photo diode channel shown in Figure 1-1. However, these can both be customized for more advanced sensing and algorithm implementations.
In addition to smoke detection, the TPS8802 AFE includes circuitry for sensing Carbon Monoxide (CO) which is necessary for combination detectors required in residential installations or can be used as part of a multi-criteria smoke sensing algorithm together with the photoelectric sensing.