SLVAE30E February 2021 – March 2021 TPS1H000-Q1 , TPS1H100-Q1 , TPS1H200A-Q1 , TPS1HA08-Q1 , TPS25200-Q1 , TPS27S100 , TPS2H000-Q1 , TPS2H160-Q1 , TPS2HB16-Q1 , TPS2HB35-Q1 , TPS2HB50-Q1 , TPS4H000-Q1 , TPS4H160-Q1
Open load detection can determine if there is a broken wire or faulty module, however it cannot determine a partial failure. In many applications LEDs are configured as multiple strings in parallel as shown in Figure 5-2. In this case it is important to know if any of the LED strings in the array are non-functional even if the remainder of the strings continue to work. TI Smart High Side Switches can determine this by using an accurate load current measurement to sense an absolute change in output current that comes with a partial open that is caused by a string failure. This information can then be used to communicate a partial failure to a MCU.
For example, consider the case of driving a parallel array of six LED strings with each string drawing a current of 50 mA. In order to recognize the open failure of one of the six parallel strings, the load current has to be sensed within ±16% accuracy. For best practices this means an accuracy within ±10%, however the system can be assumed to have an additional ±5-6% variability in common LED string current draw, ADC digitization, and other parasitic elements. This means that the Smart High Side Switch current measurement must be accurate within ±3-4% in the load current range. The current sense accuracy of devices like the TPS2H160-Q1 and the TPS2H000-Q1 match this requirement and are capable of providing an absolute current measurement with enough accuracy to diagnose a single LED string failure. Table 5-1 shows the two parts and their current sense accuracy for the given load current.
|Device||Load Current||Current Sense Accuracy|
Load current sensing enables increased diagnostic abilities by giving a designer the ability to ensure that the entire LED array is functional rather than just parts of it. This in turn improves overall system reliability and helps for predictive maintenance.