|
High linearity analog optocouplers provide the versatility required to meet a wide range of analog isolation needs. For designers of high voltage applications, high linearity analog optocouplers can reliably send analog signals across very high voltage area and low voltage area without distortion. This article examines the internal operation and servo control mechanism of high linearity analog optocouplers in detail. Application examples are also presented, ranging from motor control current sensing to traditional current loop communication in process control.
Selecting the Ideal Optocoupler for High Voltage Applications
Standard digital optocouplers have long been used to address the optoisolation needs in high voltage applications. Combining a digital optocoupler with signal processing circuitry meets the need of high voltage isolation, but this complicates the design and is not suitable for applications that require analog in and analog out. Some linear optocouplers available in the market do provide analog isolation, however they fail to deliver necessary performance such as linearity, gain accuracy, along with high enough working insulation voltage. Isolation amplifier optocouplers can also be considered, but designers must consider the trade off between cost and performance.
An analog optocoupler with high linearity is ideal to isolate analog signals in a wide variety of applications that require excellent stability, linearity and bandwidth. An optimally designed circuit is capable of handling different type of signals including unipolar/bipolar, AC/DC and inverting/noninverting. Certain applications require very high isolation voltage. For example, in motor drive high side current sensing and phase current sensing applications, the working voltage could be as high as 1 kV. An optocoupler needs to be specially constructed to work under such harsh conditions. The following examples use Avago Technologies' HCNR200/201 optocouplers to illustrate the wide range of isolation applications that can benefit from high linearity and up to 1.4 kV working insulation voltage.
Current Sensing and Voltage Monitoring Applications
High linearity is critical for current sensing and voltage monitoring in various application areas, such as motor control drives, switching power supply feedback loop, and inverter systems. As part of the motor control drives, variable-speed motor drives are finding increasing applications not only in industrial applications but also home appliances.
Among the key components such as IGBT/ MOSFET, gate drivers, and of course the microcontroller unit (MCU), analog current and voltage sensors are critical to feed back to the MCU for stable and protected system control. Because of the presence of high voltages, it is necessary, and often mandated by safety and regulatory agencies, that people operating the motors and low voltage digital electronics are protected through galvanic isolation. An optocoupler with very high insulation voltage (5 kVrms/1 min rating) is required to handle DC bus voltage monitoring, DC bus current sensing, and AC phase current sensing, as well temperature and positioning sensing.
Figure 1 shows these applications (framed in the box named Analog Isolation Block) in a typical motor drive block diagram. From this figure, one can figure out resistors R2 and R5 are used to measure the HV DC bus voltage and DC bus current respectively, while resistors R3 and R4 are used to measure motor phase current. Parameters such as temperature and position can be sensed by appropriate sensors attached to the motor, whose output is fed to another Analog Isolation Block. All the parameters are then transferred across the isolation barrier and collected by MCU. Figure 2 A and B show a simplified schematic of the Analog Isolation Block for unipolar input and bipolar input circuit respectively, which are discussed in next section.
Figure 1. A typical motor drive block diagram.
Figure 1. A typical motor drive block diagram.
Figure 2. Simplified schematic of the Analog Isolation Block for (A) unipolar input, and (B) bipolar input.
Figure 2. Simplified schematic of the Analog Isolation Block for (A) unipolar input, and (B) bipolar input.
Theory of Operation
The operation [3, p. 15] of the circuit may not be immediately obvious just from inspecting Figure 2A, particularly the input part of the circuit. The op-amp always tries to maintain the same inputs voltages at its two inputs in a linear feedback close loop connection. Thus, the input side op-amp A1 always tries to place zero volts across the photodiode PD1. Now, if some positive voltage VIN+ is applied at the input, the op-amp output would tend to swing to the negative rail causing the LED current to flow. This VIN+ will cause a current flowing through R1, and the LED light output will be detected by PD1 and generates and a current IPD1 flowing from the "+" terminal to GND1. Assuming that A1 is a perfect op-amp, no current flows into the inputs of A1; therefore, all of the current flowing through R1 will flow through PD1. Since the "+" input of A1 is at 0 V, the current through R1, and therefore IPD1 as well, is equal to VIN+/R1, or
IPD1 = VIN+/R1.
Notice that IPD1 depends ONLY on the input voltage and the value of R1 and is independent of the light output characteristics of the LED. As the light output of the LED changes with temperature, amplifier A1 adjusts IF to compensate and maintain a constant current in PD1. Also notice that IPD1 is exactly proportional to VIN+, giving a very linear relationship between the input voltage and the photodiode current. The relationship between the input optical power and the output current of a photodiode is very linear. Therefore, by stabilizing and linearizing IPD1, the light output of the LED is also stabilized and linearized. And since light from the LED falls on both of the photodiodes, IPD2 will be stabilized as well.
Since PD1 and PD2 are identical to each other, IPD2 shall be equal to IPD1 ideally, while being varied a coefficient K3 in reality. So we have
IPD2 = K3 x IPD1,
where K3 is the transfer gain defined in the data sheet (K3 = IPD2/IPD1 = 1). Amplifier A2 and resistor R2 form a trans-resistance amplifier that converts IPD2 back into a voltage, VOUT, where
VOUT = IPD2 x R2.
Combining the above three equations yields an overall expression relating the output voltage to the input voltage,
VOUT/VIN+ = K3 x (R2/R1).
Therefore the relationship between VIN+ and VOUT is constant, linear, and independent of the light output characteristics of the LED. The gain of the Analog Isolation Block circuit can be adjusted simply by adjusting the ratio of R2 to R1.
Figure 2A is in a unipolar configuration that accommodates only positive voltage input. Figure 2B is configured to accommodate bipolar input (a signal that swings both positive and negative). Two current sources, IOS1 and IOS2, are added to offset the signal so that it appears to be unipolar to the optocoupler. Current source IOS1 provides enough offset to ensure that IPD1 is always positive.
The second current source, IOS2, provides and an offset to obtain a net circuit offset voltage of a desired value (e.g., a 0 V may be desired if both positive and negative power supplies are used, whereas a midway voltage could be more appropriate for the case of single positive power supply circuit). Current sources IOS1 and IOS2 can be implemented as simply as resistors connected to suitable voltage sources. A note is that the offset performance is dependent on the matching of IOS1 and IOS2 and is also dependent on the gain of the optocoupler.
|
