In a perfect world a high reliability system should be designed to avoid single point failures and provide a means of isolating faults in such a way that operation may continue perhaps at a reduced performance level. It should also be able to contain faults to avoid propagation to downstream or upstream electronics.
Built-in redundancy, either in the form of parallel circuits that share the load actively or that wait in a standby until a failure occurs, is one solution.
In each case, fault detection and management requires additional overhead circuitry contributing to the overall complexity and cost.
Some systems also create dissimilar parallel circuits to add diversity and avoid the risk of a common failure mechanism; this is the case for some aircraft flight control systems.
High complexity systems increase power supply performance requirements and high conversion efficiency and good thermal management are critical as for every 10°C rise in junction temperature the IC lifetime is approximately halved.
Output Current Limiting
This is not a new feature but its implementation has become more accurate and sophisticated and additional flexibility is provided as user programmable features are added. Also it is possible to use an IC (Figure 1) with both a switch current limit and catch diode current limit such that the output current is controlled during fault conditions such as a shorted output.
These measures not only protect the device itself, but also the downstream electronics should a fault develop.
Input Current Limiting
This is commonly found in circuits such as those performing energy harvesting from photovoltaic cells where the high impedance source requires that the current be carefully controlled to prevent the source voltage collapsing.
In addition to protecting the upstream electronics from overload, it can also be employed as a safety feature for a backup supply where large capacitors must be protected and safely charged.
Thermal protection is implemented in the majority of power regulator ICs with internal power transistors. Some ICs also contain a thermal regulator to prevent thermal shutdown when charging very large capacitors at high current.
Controlling Multiple Input Sources
Power supply systems that contain a main supply and a redundant backup with perhaps an external auxiliary supply need a system to arbitrate which supply has priority and to monitor their status. Furthermore, it must protect the system from cross-conducting and back-feeding during source switching.
It is possible to use an IC which automatically selects the source based on validation of user defined supply thresholds for each input.
An alternative approach is to share the load between two input sources that operate simultaneously, increasing reliability by reducing the burden on each source and at the same time providing protection against failure of one source if they are each suitably sized to support the full load requirement.
In the past, a simple but inefficient diode-OR arrangement might have been adopted but that required each supply to have active control to balance the loading.
This can now be accomplished with a current sharing controller with reverse blocking that prevents a fault in one supply, bringing down the power system.
Military and aircraft electronics must confirm to transient protection specifications such as MIL-STD-1275 (vehicles) and MIL-STD-704 / DO-160 (aircraft). However, protection from voltage surges, spikes and ripple is desirable in any high reliability system and there are products that are dedicated to that function.
While advances in silicon process technology now allow regulator ICs to operate with input voltages of 100V or more, the dedicated transient protection ICs provide more functionality and control.
Digital Power System Management
New products are combining the advantages of analogue power regulation with digital control over a 2-wire PMBus I2C-based digital interface protocol to enable remote management of power supply systems.
Telemetry and diagnostics data can be used to monitor load conditions, read fault logs and provide access for trimming and margining to ±0.25% accuracy, maximizing system efficiency and reliability. Such systems offer the opportunity to move from time based maintenance schedules to condition based maintenance and can potentially highlight performance degradation prior to system fault conditions taking hold.
High reliability aircraft power supply systems include an isolation barrier to protect the aircraft power buses from faults in downstream line replaceable units, typically rated at hundreds or thousands of watts.
Increasing numbers of sensors and actuators are also driving demand for smaller, locally isolated power supplies and data interfaces to reduce noise induced problems from ground loops and common mode interference. There are now complete galvanic isolated BGA module solutions to simplify design and increase reliability.
Most of this article has been dedicated to new functions that simplify designing high reliability power supplies or product features that protect the device from fault conditions or mistreatment. However it is critical not to overlook the importance of component quality and of selecting the correct grade of component for the anticipated environmental conditions.
Design of high reliability power supplies have been simplified by user programmable features and more sophisticated on-chip protection mechanisms.
Digital power system management provides the means to remotely monitor and control power systems and to further improve efficiency and reliability. Finally, selecting the correct grade of component from a reputable supplier will reduce the chance of quality and reliability issues.