This technique enables detection of when a power stage fault occurs, and the type of fault.
A vehicle electronic control unit consists of various high-side power stages for driving different loads. Common faults that these power stages experience are Short Circuit to Battery (SCB), Short Circuit to Ground (SCG), and Open Load (OL). These faults can occur during either on-state or off-state of the power stage output. It is essential to diagnose a fault such as SCB during switch off-state, SCG during switch on-state, and OL during both on- and off-state of the switch in order to avoid system malfunction or power stage damage.
A high-side power stage mainly consists of a switching power MOSFET and its driver section. Semiconductor manufacturers often provide diagnostic information through a digital status (ST) output. A logic low level on the ST pin indicates a fault. The limitation of the ST signal is its inability to distinguish among faults SCG, SCB, and OL. For high-side switches (HSS) to overcome these drawbacks, there is a need for proper diagnostics. The diagnosis technique should detect a fault of the power stage, and clearly distinguish the kind of fault. The improvised diagnostic strategy provides a simplified solution for fulfilling these requirements by utilizing ST and reading output via ADC.
The concept developed in this work provides improved diagnostic features making use of discrete HSS having the digital status signal ST. Commonly available discrete high-side switches have a single digital status output for diagnosis. Though a status signal of such discrete HSSs helps in fault diagnosis, the common drawback is the inability of such power stages to distinguish the kind of fault.
Other failures in high-side switches are overload current, under-voltage, over-voltage, and over-temperature (OT). Under-voltage and over-voltage are generally protected by the high-side switch. For overload current fault, these high-side switches are protected by an internal current limit technique.
In this technique, diagnosis is done with a two-step detection process. The first step checks for the occurrence of a fault by monitoring the ST signal from the high-side switch device, wand the second step identifies the kind of fault by reading the voltage at the high-side switch output. Ideally, the ST signal is fed to an interrupt input of a microcontroller unit (MCU), and the state of the highside switch output is measured by the ADC of the MCU via a suitable resistor divider network. Hence, ST pin digital status signal is used, in combination with the ADC feedback (one after the other) for proper diagnosis.
With this approach, the software will respond appropriately, completing the diagnosis and fulfilling the application requirement. A pulse width modulation (PWM)-controlled high-side switch needs an improvised diagnostic scheme. In this scheme, both the digital status signal and the analog feedback signal from the power stage output are used for diagnostics. For the PWM-controlled high-side switch, the digital status signal is used in combination with the state of the power stage output signal via ADC.
The diagnostic concept consists of a high-side switch with a digital status pin, analog feedback network, and a MCU that runs diagnostics software. The improvised diagnostic technique has two main diagnostic steps. In step 1, it checks for the fault through a status signal (ST) pin. Whenever a fault is found, the software turns the power stage to off. The power stage then remains off. In step 2, it specifies the kind of fault by reading an analog feedback signal that gives the state of power stage output via ADC.
In a PWM-controlled switch, diagnostics are carried out during both the on-period and off-period of a PWM. With this technique, diagnosis of SCG faults can be detected during the on-period of PWM. Diagnosis of SCB and OL faults can be detected during the off-period of PWM.
A microcontroller that processes the digital status signal should turn off the high-side switch (which was PWM-controlled) immediately after it sees a logic low fault signal at the digital status pin at any state of PWM. Once the switch goes to steady-state off, then ADC feedback will be useful to distinguish the fault type. Though the final detection of fault type is done during steady off-state of the switch, this technique can be effectively used for a PWM-controlled high-side switch.
This work was done by Rajesha Kundrukote of Robert Bosch Engineering & Business Solutions. The full technical paper on this technology is available for purchase through SAE International at http://papers.sae.org/2013-01-2883/.