Circuit protection is like insurance，at best, an afterthought and even when put in place, it often turns out to be insufficient. But while the cost of underinsurance can be enough to threaten the stability of a business, inadequate circuit protection could result in much worse: The loss of human lives.
Consider the sobering example of Swissair Flight 111 flying out of JFK International Airport, New York, on September 2, 1998. The plane, a seven-year-old McDonnell Douglas MD-11 carrying 215 passengers and 14 crew members, had recently been fitted with an upgraded inflight entertainment (IFE) system. Fifty-two minutes after take-off, the crew smelled smoke and reported an emergency. Attempting to divert to Halifax, the pilots eventually lost control of the plane when a fire in the ceiling of the cockpit burned through electrical control cables. The aircraft crashed into the ocean 8km off the coast of Nova Scotia, killing all on board.
The accident investigation determined the crash was primarily triggered when materials making up part of the IFE, which were supposed to be fire-resistant, caught alight and flames spread to critical control lines. While not definitively established, the source of the fire was believed to be arcing across wires serving the IFE. Though the wires had circuit breakers, the circuit breakers were not designed to trip in response to an arc. Today, such circuits are protected by arc-fault detection devices that do trip when sensing arcs (apart from those generated by normal operation such as switches flipping). Insufficient circuit protection resulted in 229 fatalities.
USB-PD Brings Increased Hazards
While the Swissair MD-11 was brought down by an electrical rather than an electronic fault, electronics engineers are increasingly designing circuits where voltages and currents are sufficient to generate arcing (and potentially life-threatening fires). One example is USB Power Delivery (USB-PD); this is an upgraded version of USB, which allows for higher voltages and currents, up to a maximum of 20V and 5A (for 100W maximum power). USB-PD’s upgrade is a notable step up from, for example, USB Type-C’s 5V and 3A (15W), dramatically increasing the likelihood of hazards.
Apart from the risk associated with higher voltages and currents, USB-PD is used with USB Type-C connectors and cables—both of which increase the chance of accidents. This is because with just 0.5mm between pins, the USB Type-C connector features a pitch just one-fifth that of Type-A and Type-B connectors, which increases the risk that a minor twist of the connector during insertion or removal could cause a short. A build-up of detritus inside the connector could have a similar effect. Moreover, the popularity of USB Type-C has seen the development of a big supply of aftermarket cables—many of which are not up to the job of carrying 100W and yet are not marked as such. Not that such markings guarantee safety; an under-specified cable can be just as easily plugged into a USB-PD socket as a conforming one if a consumer is determined to use it.
Arcing is not the only risk when using USB-PD at higher voltages and currents. Because of the proximity of the main bus power supply pins to other pins in the connector, a short can easily expose downstream electronics to a damage-inducing power surge—and not one just limited to 20V. The inductance of a one-meter USB cable, for example, is sufficient to generate ‘ringing’ that can result in a peak voltage far higher—sometimes even double—than that of the (20V) short-circuit voltage. Depending on the application, the failure of downstream devices subject to these overvoltages could put safety at risk because some of the most vulnerable to damage are those that normally control the maximum current and voltage at which the cable should operate.
Comprehensive Circuit Protection
The risk of arcing or component damage when using USB-PD at its highest current and voltage ratings is such that paying lip service to circuit protection doesn’t cut it. Comprehensive circuit protection in applications that regularly use USB-PD’s highest power modes—for example, when charging the battery of a portable computer—is mandatory.
Transient voltage suppression (TVS) diodes fitted between the USB Type-C receptacle pins and ground are a relatively simple and inexpensive start to the circuit protection task. In the event of a transient short, the TVS diodes ‘clamp’ the peak voltage to a level that connected components can withstand. Unfortunately, while TVS diodes offer a good option for transients, they are not so good for sustained overvoltage events. To deal with these, additional circuitry in the shape of an overvoltage protection device paired with an n-channel MOSFET is needed. Then, during a sustained overvoltage episode, the protection device triggers the nMOSFET to disconnect the load from the input sparing downstream connected devices from overload. But TVS diodes, protection devices, and nMOSFETs still don’t protect against all overvoltage scenarios; occasionally shorting events can occur that bypass the USB cable. In this case, receptacle inductance is very low, allowing the voltage to rise more rapidly than the protection device and nMOSFET can react. The solution lies in yet more clamping devices to hold the rise in voltage for long enough for the protection device to cut in.
Comprehensive protection inevitably adds cost and complexity to a USB-PD application, but there are ways to rein in both by judicious selection of components. Manufacturers now offer integrated devices that incorporate TVS diodes, protection, and clamping devices into single packages (nMOSFETs typically remain as discrete chips). These monolithic chips simplify USB-PD protection design while saving money and space.
Conclusion
Circuit protection will never be at the glamorous end of the electronics development business. But the knowledge that suitable protection could prevent material damage, injury, or even death is certainly cause for immense satisfaction for the skilled engineer that developed the solution.
