The uninterrupted availability of server, communications and telecoms equipment is frequently critical to its client applications, so this equipment typically uses two or more power sources in a redundant power architecture. These sources need mutual protection, otherwise a short-circuit fault in one could quickly overload and damage the others.
The diagram above shows a simple solution using diodes. Each diode allows current to flow in a forward direction only, while preventing either supply from drawing short-circuit current. Therefore the system can continue to function if one supply fails. Although simple and fast, this arrangement has a drawback due to the diodes’ high forward voltage in their normal state of forward conduction. This creates high power and heat dissipation and an unwelcome need for thermal management and extra board space. As systems have increased in power and component density, these problems have become unacceptable.
Active ORing offers a better alternative; it comprises a power MOSFET and controller IC. The MOSFET has an on-state resistance RDS(ON) which, multiplied by the square of the current through the device, creates an internal power loss. However this loss can be substantially lower than that of a Schottky diode for the same current; a ten times efficiency improvement is typically achieved.
For example, for a 20 A application, a Zener diode with a .45 V forward voltage drop would dissipate 9 Watts of power, while a MOSFET with a 2 mOHM RDS(ON) on would dissipate only 0.8W, a more than 10x reduction in power loss.
Yet an active ORing solution requires care in its set-up. A MOSFET passes current in either direction when turned on, allowing the possibility of a large reverse current if the protected power source has a short-circuit failure. This will bring down the system if it persists for long enough. Therefore, the active ORing solution must be very accurate, and able to detect reverse current fault conditions extremely fast. When a fault is detected, the controller must turn the MOSFET off as fast as possible, isolating the input fault from the rest of the system.
The controller IC senses information across the MOSFET to determine the magnitude and polarity of current flowing through it. Transgression of its reverse current threshold indicates a power source failure; therefore sensing of this must be tightly accurate to provide the consistent, fast fault detection required. Speed of response is also critical, as this determines the magnitude and duration of reverse currents. Higher peak reverse currents increase reliability concerns, leading to larger MOSFETs, higher costs and more demand for real estate. These issues can be exacerbated by some of today’s low-impedance power architectures. Conversely, tight, fast control can mean the difference between a damaged system and a safe disconnect of a failed power supply.
To summarise, we can say that the best Active ORing solutions comprise an ORing controller that is fast and accurate in detection and fast in response, together with a MOSFET offering the lowest possible RDS(ON) . These attributes allow minimised losses and dependence on thermal management, with smaller size and greater ease of use.
Vicor offers a complete family of Active ORing solutions, including discrete active ORing controllers as well as full-function system-in-package solutions with high-speed integrated low on-state resistance MOSFETs, in high-density thermally enhanced LGA packages that reduce the power footprint.
For more information on Vicor’s Cool-ORing product family go to the Vicor web site.