SLYT883 October   2026 LM74900-Q1 , LM74910H-Q1

 

  1.   1
  2. 1Introduction
  3. 2The differences between ORing and priority switching
  4. 3Priority switching details
  5. 4Conclusion
  6. 5About the author
  7. 6Trademarks

Priority switching details

In general, the circuits used for ORing and priority switching are not particularly different from one another, but the behavior of the priority switching implementation requires some attention to detail.

While the ORing solution enables both power paths if they are somewhat equal, the priority switching approach keeps the auxiliary power path disabled until the voltage of the main power path drops below a certain threshold. Reaching this threshold disables the main power path and enables the auxiliary power path, causing a sudden load shift from one supply to the other and generating a sudden voltage step on both supplies. Because of this sudden load shift, the voltage on the main power supply will often spike for a short period before completely collapsing, although this depends on the failure causing the initial drop.

Because the auxiliary supply switches based on the voltage on the main supply, a voltage spike can cause the system to switch back to the primary supply, which immediately lowers the primary voltage and causes the system to revert to the auxiliary supply. This results in uncontrolled oscillation between the two power supplies and creates a large ripple on the output, which may disrupt system functionality. The reason behind the voltage shift on each supply is the sum of all parasitic resistances and inductances within the power path, such as printed circuit board traces, wiring, and the MOSFET drain-to-source on-resistance (RDS(on)).

Figure 3 shows an example of such behavior. During a disconnect of the primary power supply, a short burst occurs right when the input voltage drops below the switchover threshold. Channel 1 in Figure 3 shows the voltage on the auxiliary supply, while Channel 3 shows the output voltage. The auxiliary supply voltage in this scenario has a lower value than the primary supply voltage, with the switchover threshold set right above the voltage level of the auxiliary supply.


 Oscillation during input switchover in a priority switching system

Figure 3 Oscillation during input switchover in a priority switching system

Depending on what problem causes the main supply to drop, this oscillation may even be continuous; for example, if the cable for the main supply is partially damaged and causing a huge voltage drop instead of a complete disconnect. In many cases the hysteresis built into the ideal diode controllers, together with a sufficiently high input capacitance, can suppress this behavior. But in some cases the hysteresis may not be enough to avoid unnecessarily large input capacitors.

Although it is not possible to change the internal hysteresis, on some devices there is a simple solution to increase it externally. Many devices such as the LM74910H-Q1 offer a fault output that you can use to add external hysteresis with minimal component count. On the LM74910H-Q1, the fault pin is an open-drain output that pulls low for overvoltage conditions and other faults. This output can also control an external transistor, which alters the feedback divider for the overvoltage input. Figure 4 illustrates a typical implementation.


 External hysteresis

Figure 4 External hysteresis

Since the fault output is an open-drain output, you will need a bias resistor to turn on the bipolar junction transistor. This resistor connects to the software pin of the integrated circuit, which offers a switchable low-current path to the input that becomes disabled when the system shut downs to minimize quiescent current.

As long as the primary power supply is above the switchover threshold, the auxiliary controller is in an overvoltage fault state, disabling the power switches and pulling the fault output low. Pulling down the fault output disables transistor Q1; therefore, the feedback divider for the overvoltage input only consists of R1 and R2. This results in an overvoltage feedback voltage defined by the ratio of both resistors and the switchover threshold, as expressed by Equation 1:

Equation 1. V o v _ f a l l i n g = V s e n s e _ f a l l i n g     R 1 R 2

As soon as the primary supply voltage drops below this overvoltage threshold, the fault output no longer pulls low and the transistor switches on, connecting R3 in parallel to R2 and reducing the voltage appearing at the overvoltage input. Equation 2 now gives a new value for the rising threshold:

Equation 2. V o v _ r i s i n g = V s e n s e _ r i s i n g     R 1 R 2     R 3 R 2   +   R 3

Changing the value of R3 easily adjusts the resulting hysteresis, offering great design flexibility.