Digital Isolation in Hybrid and Electric Vehicles

Hybrid and electric vehicles ( HEV / EV ) are bringing 400V and higher voltage designs to the automotive and transportation sectors. To handle such high voltages and currents in the harsh automotive environment, a highly reliable and long-term stable solution is required to effectively isolate this high voltage from other electronic functional circuits (of course, the most important thing is driving!).

Isolation requirements in transportation applications

Hybrid and electric drivetrains in cars, trucks, and motorcycles have created new, previously unknown challenges in the transportation industry. The original 12V voltage network now needs to be supplemented by 400V or higher batteries and power systems, which has put forward a new set of requirements for automotive OEMs and system module suppliers. All functions in hybrid/electric vehicles, such as high-voltage batteries, DC/DC converters, inverters for driving motors , and on-board charger modules connected to the 230V/380V grid, have isolation requirements (Figure 1).

Figure 1: Typical system architecture of an electric vehicle.

Automotive and transportation applications have different requirements for isolation than industrial applications. Robustness is a must, of course, but so must immunity to magnetic “noise”. The high power levels in the car (e.g. a 100kW motor operating at 400V, meaning 250A of operating current) create strong magnetic fields in the car that must be properly managed. The parts used must last long enough to meet the expected life of the vehicle; for example, they must meet the requirements of large transportation applications for decades. Products used in automotive environments will drive requirements for automotive application quality (Q1) and an operating temperature range of -40 to +125°C.

At the same time, cost pressure in these areas will drive the demand for higher system integration. Therefore, single-chip products with isolation functions, such as CAN transceivers, ADCs or gate drivers, have shown their advantages.

Different digital isolation technologies

In principle, there are four different digital isolation methods: optical, inductive, capacitive and radio frequency. The first three methods are described below.

Optical isolation technology uses a transparent insulating isolation layer to transmit light to achieve optical isolation. By driving the LED (light emitting diode), the digital signal is converted from electricity to light. This light signal is then transmitted through the isolation layer, and then converted back to an electrical signal using an optical detection component (photodiode, phototransistor).

The main advantages of optical isolation are the immunity of light to electric or magnetic fields and the possibility of transmitting static signals. On the receiving side (flipside) of the isolation layer, the operating frequency (transmission speed) of the optical isolator is limited by the relatively slow characteristics of the LED. For hybrid/electric vehicle applications, the limited life of optical isolation is a major disadvantage. Over time, the efficiency of the LED will decrease, requiring a larger signal drive current (usually starting at 10mA), so over time, this optical isolation will eventually fail to function.

Inductive isolation uses the change in magnetic field between two coils to achieve communication across the isolation barrier. One advantage of the inductive isolation method is the difference between common mode and differential transmission, which means it has good noise immunity. The disadvantage of this method is the distortion that can come from the magnetic field, which is common in the motor control environment of hybrid/electric vehicle applications.

Capacitive isolation uses the change in electric field across the isolation barrier. The benefits of capacitive isolation are greater immunity to magnetic fields and long system life. The transmission speed of capacitive isolation is similar to that of inductive isolation.

However, the disadvantage of capacitive isolation is that there is no differential signal (that is, the signal and noise share the same channel). In addition, like inductive isolation, they cannot directly transmit static signals (they must first be encoded with a frequency signal).

Isolation Products

[page]

An example of a capacitive isolation approach is the ISOxxxx series from Texas Instruments. Figure 2 shows a simplified architecture of the ISO72xx system. The ISOxxxx device integrates two die on separate lead frames, as well as the transmit and receive chips, in a single package. The two die are connected only by a bond wire. The actual isolation function implemented at the receiver is a capacitor based on silicon dioxide (SiO2, i.e. glass) with copper and doped silicon as the substrate electrodes (Figure 3). The use of SiO2 brings the advantages of high reliability and long life.

Figure 2: TI’s ISO 72xx family architecture diagram.

Figure 3: Die photo of TI’s ISO72xx family, with silicon dioxide isolation used on the receiver die on the right.

Two channels allow both DC and AC communication and are also fail-safe.

The basic AC channel uses an input signal and transmits it through a differential pair consisting of isolation capacitors after filtering. It is then detected by the input of the Schmitt trigger on the receiving chip, and finally the received signal is output through the output buffer. It can achieve very high-speed transmission, slight pulse width distortion and short transmission delay, but cannot send DC signals.

A DC channel can be used to transmit a DC (or very slow) signal across the isolation barrier. An on-chip oscillator encodes the signal into a PWM signal, which is transmitted across the isolation barrier using a differential signal similar to an AC channel. At the receiving chip, the signal is decoded and sent to an output buffer.

However, the DC channel is also used for fault prevention. For example, if the power supply voltage at the transmitter is not high enough, the oscillator will stop working, which means that the receiver will not detect the data signal, thus issuing a fault indication and outputting a high level. In normal operation (i.e., with sufficient data transmission density), the output multiplexer will ignore the DC channel; but when there is no data transmission on the AC channel within about 4us, the DC channel will be given priority. Once there is a jump in the AC signal, the multiplexer will immediately switch back to this channel.

[page]

There are a variety of isolation components available to meet different configuration requirements (single to quad). They all provide 560V continuous insulation (4kV peak transient) at data rates up to 150Mbps. In addition, these components meet automotive grade application requirements.

Reliability considerations and immunity to external electromagnetic fields

The harsh automotive and transportation environments, coupled with the long service life of vehicles, require components with special characteristics. Mean time to failure (MTTF) is a standard method for determining the reliability of semiconductor circuits. For isolation components, it applies to both the integrated circuit and the isolation principle. Evaluation should be carried out using a 90% confidence level and an ambient temperature of 125°C. Typical capacitor and inductor components have an MTBF of more than 2,000 years and a FIT (failures in 10 ⁹ hours) of less than 60. In contrast, the MTBF of typical optical components is only 30 years and the FIT is nearly 4,000.

In terms of immunity to magnetic fields, Figure 4 compares inductive and capacitive components. Both inductive and capacitive components (ISO721) have high magnetic field immunity far exceeding the IEC61000 standard. However, the capacitive component ISO721 is superior, which is particularly important for harsh automotive environment applications.

Figure 4: Immunity to external magnetic fields.