Manufacturers of optocouplers characterize their devices for CTR (current-transfer ratio), and an optocoupler’s data sheet will list its CTR ranges. Unfortunately, the CTR range may be too wide for some applications, requiring you to screen incoming devices when you need a tighter tolerance.
Optocouplers are commonly used in the feedback path of a switched-mode power supply. The CTR determines the DC gain of a power supply’s open-loop frequency response, which affects the loop’s phase margin. An optocoupler influences multiple operating parameters, so CTR variations between devices can cause production tests to reject a power supply. When this happens, you must measure CTR as part of your troubleshooting.
With the circuit in the Figure 1, you can measure CTR at VOUT with a multimeter. Besides serving as the VCC source for the DUT (device under test) circuits, IC5’s 10-V output also provides virtual ground to IC3, so the latter can operate with a single, unipolar power supply.
 |
|
| Figure 1. | This circuit lets you directly read a DUT's CTR across VOUT using a multimeter. |
At TDK, we use this circuit to test the TLP521-1(GR) optocoupler. We measure the device’s CTR under fixed IF (forward current) and VCE (collector-to-emitter voltage). The network surrounding IC1 forms a current source that provides a fixed current (IF = 5 mA) that drive’s the DUT’s photodiode. The circuit around IC2 forms a shunt-regulation network that keeps the DUT’s VCE at a fixed voltage at 5 V. CTR is defined as IC/IF.
We calculate IC (the collector current) and IF by measuring the voltages across resistors R2 and R3:
and
The rest of the circuits are mainly for signal conditioning.
Analog multiplexer IC3 functions as a divider that produces its output, W, with a transfer function of:
IC3’s output feeds amplifier IC4, removes the offset voltage of 10 V, and amplifies the signal with a gain of:
The resistor divider network of R12 and R11 scales down the voltage from IC4 by a factor of 1/3. Thus,
Voltage VOUT represents the optocoupler’s CTR.
Resistors R5, R7, and R13 will let you calibrate the initial error of IC1, IC2, and IC3. We adjust the value of R7 to trim the initial error of IC1’s reference voltage. We measure the voltage across R3 and adjust R7 until the voltage across R3 is 2.15 V. Potentiometer R5 lets us trim the error in IC2. Measuring the voltage across VCE of the DUT lets us trim R5 until VCE = 5 V.
Using the circuit, we’ve found that IC3’s gain contributes most of the error in VOUT. By measuring VR3 (hence IF), VR2 (hence IC), and VOUT simultaneously, we can trim R13 so that VOUT is reflecting an actual ratio of IC/IF.
IC5 is a 10-V TS7810 whose output is labeled as VGND (virtual ground). This voltage acts as the VCC source that supplies test current to IC1, IC2, and the DUT circuits. It also provides virtual ground potential to IC3 (AD734), which lets the device operate from single supply voltage.
The test circuit lets us evaluate the CTR of the TLP521-1(GR), but we also use it to test many types of optocouplers with a CTR that is within 1 to about 3.8 without making changes to it. Factors that may limit the ability of the circuit to measure the CTR range include the input and output “rail-to-rail” capability of IC3 and the power dissipation of IC2. If you need to adapt the test circuit because an optocoupler’s operating parameters force the circuit out of its linear region, then you can change the values of resistors R1, R2, R3, R7, and R8 as needed.
