Shunt references require a minimum cathode current to function properly. This current is provided in each device's data sheet and can vary widely between devices. A biasing resistor is used to verify that these shunt references operate properly so that a biasing current flows into the shunt reference cathode without affecting the feedback signal through the optocoupler. There are two common topologies for biasing a shunt reference for SSR; one places a biasing resistor in parallel with the optocoupler diode, while the other places this biasing resistance directly from the system output to the shunt reference cathode.

Figure 2-3 Biasing a Shunt Reference Topology 1

When the biasing resistor, R_bias, is placed in parallel with the optocoupler diode, as shown in Figure 2-4, the minimum biasing resistance can be generally solved with Equation 2 where V_F(drop) is the expected forward voltage drop of the diode, and I_KA(min) is the minimum cathode current of the chosen shunt reference.

The forward voltage drop across the optocoupler diode is non-linear and varies with forward current and temperature. When a flyback converter is subjected to the max load condition, the steady state forward current of the optocoupler is at the minimum value (due to the requirement of higher power output from the PWM controller), which decreases the forward voltage drop of the optocoupler diode. This shows that under max load conditions, the biasing current being provided to the shunt reference is minimized, meaning that when selecting R_bias, the minimum expected forward voltage drop, V_F(drop), can used in Equation 2 to prevent accidental under-basing for a flyback converter under max-load conditions. This can maintain that the shunt reference is properly biased under all load conditions.

Figure 2-4 Biasing a Shunt Reference Topology 2

In the second topology shown in Figure 2-4, the biasing resistance between the flyback output and the shunt references cathode pin is established. The shunt reference controls the optocoupler current and acts as an error amplifier. Choosing a biasing resistance for this topology is similar to the previous topology; however, the expected voltage drop across this biasing resistance varies more under different load conditions due to an additional voltage drop across the optocoupler resistance that must be considered.

Equation 3 shows the formula for calculating the max biasing resistance, R_bias. The cathode voltage, V_KA, is determined by the expected feedback current through the optocoupler diode on the secondary side and can be calculated as the output voltage, V_out, minus the voltage drop of the optocoupler diode and resistance on the secondary side. Decreasing the biasing resistance below this calculated value increases the biasing current above I_KA(min), which dissipates extra power with no clear improvements in performance. Under max-load conditions, expect the feedback current flowing through the optocoupler diode, IFB (secondary), to be​ lower than this current under a no-load condition (standby mode).This shows a reduced voltage drop across these components, increasing the cathode voltage, V_KA, at the shunt reference. Therefore, similarly to the first topology, the bias current, I_bias, varies under different load conditions, where when the flyback converter is subjected to the max-load condition, the biasing current is at the minimum value. Choose the bias resistance, R_bias, to provide the minimum cathode current, I_KA(min), to the shunt reference under the desired max-load conditions so that this reference is properly biased across the entire load range.

In both topologies, the current flowing into the cathode of the shunt reference, I_KA, is the sum of the current flowing through the optocoupler diode, I_F(drop), and the biasing current I_bias. The voltage applied to the REF pin determines the amount of cathode current the shunt reference shunts to the ground. One can think that because of this, increasing I_bias decreases the current through the optocoupler diode, I_F(drop), thus impacting the feedback loop; however, this is not the case. This is due to the near-infinite transconductance of the shunt reference when the REF voltage is near the internal reference, which is shown in Figure 2-5.

Figure 2-5 V_ref Gain Comparison

Figure 2-5 shows that if excess bias current is provided to a shunt reference, the REF voltage does not need to increase by any notable amount to allow this extra current to be shunted to ground. It is biased close to the ON current. This shows that overbiasing the shunt reference won't have any notable impact on the feedback current through the optocoupler diode, I_{FB(secondary)}.