MENU

Intelligent car audio distribution

Intelligent car audio distribution

Technology News |
By eeNews Europe



Modern cars with hands-free technology usually have one or two built-in microphones, which are hooked up to the head unit via an analog connection. In the future, additional microphones will be needed for new applications and features, such as hands-free calling from the rear seat, in-car communication (ICC) and active noise cancellation. For various reasons, hooking these microphones up with a classic analog connection is not ideal or sometimes not even possible.

In many head units there is no room for all the necessary plug connections, and there is also a huge amount of wiring involved if each microphone needs to be directly connected to the head unit. Furthermore analog-to-digital converter (ADC) channels need to be set up in the head unit for subsequent processing of the microphone signals, which would only be fully utilised in some vehicles or would require a wide variety of head units.

With all that in mind, Analog Devices has developed an audio bus, which enables several microphones to be connected via a simple two-wire line. The “Automotive Audio Bus” (A²B) is a bidirectional bus with a data word width of up to 32 bits and supports sample rates of 44.1 kHz and 48 kHz. As a result, it can be used not only to connect microphones but for almost all in-car audio connections. This means that various audio components, such as head units and amplifiers, can be connected with A²B in a very cost-effective manner.

One of A²B’s major advantages is its simplicity. The principle is based on an I²S/TDM (audio multi-channel) connection, which is normally only used for connections between components on the same circuit board. A²B makes this multi-channel connection via a cable, measuring one or more meters long, whereby only two A²B wires are required for the four TDM signals (SYNC, BCLK, DTx, DRx) through which clock and data are transferred in both directions. The A²B chip has both an I2S/TDM interface and the transceiver for the A²B connection. The A²B connection looks like a normal TDM line both for the master and the slave. For the microphone connection, the slave chip has the option to accept a PDM (pulse-density modulation) signal as an input.


A variety of digital microphones support PDM as digital output. The PDM output emits a “sigma-delta” transformer signal without filtering; the digital filtering takes place in the A²B chip. This allows up to four digital microphone capsules to be directly connected to the A²B bus without any additional circuitry.

Figure 1 shows how A²B simplies wiring.

There is no need for a software stack, which processes incoming and outgoing data in one form or another, to be put in place on the master or the slave. Implementing the bus simply requires a host controller to configure the master and slaves via I²C. The slaves are configured via the bus and require no further intelligence. In addition to the audio data, the bus provides the transfer option for an I²C channel, whereby the host ECU communicates with its I²C master via the I²C slave interface of the A²B master.


This communication is then relayed via the bus and the A²B slave with its I²C master connection to the corresponding component. Both for the host and an external component, this works as if both were directly connected to each other (see Figure 4). As such, even external components (e.g. ADCs) can be configured in the slave node via the bus.

Figure 2: A²B enables I²S/TDM and I²C via a UTP (unshielded twisted pair) cable connection.

A²B is therefore a bus, which can transmit both a multi-channel audio stream (in both directions) and an I²C channel via a two-wire line. A power supply (max. 50 mA per node) is also provided via the bus for slave nodes and their components (e.g. microphones). If the slave node requires more power than can be provided, a local power supply can also be connected at any time.


The bus consists of one master node and currently up to 8 slave nodes. The master is supplied via the I2S/TDM interface, by the host controller or an audio DSP, with the sampling frequency (SYNC), which is then the master clock for the entire bus. The bus itself runs with a clock which corresponds to 1,024 times the sampling frequency, i.e. approximately 50 MHz where the audio sample rate is 48 kHz. In addition to the I2S/TDM connection, the master requires another I²C connection, whereby the A²B master is the slave for the I2C connection. The host provides the entire bus via this interface, where the first slave is initially supplied with voltage and is then discovered. “Discovered” means that the slave which is being spoken to responds and therefore can be programmed and directly addressed by the host.
The host discovers and configures each slave in turn.

Figure 3: Example of A²B system set-up – slave 1, PDM microphone; slave 2, analog microphone and speaker.

As part of this, every A²B slave is informed of how many audio slots it will receive from the master and how many of them it will emit itself, and it will pass on the remainder to the next slave. The same is then configured for the other direction (upstream). Post-configuration every slave knows how many and which data slots it has to issue and how many it will get as input. As such, there is flexibility in how the overall bandwidth of the bus can be configured, meaning that any number of upstream and downstream slots can be specified.


Once the slave chip has been configured, external components (e.g. an ADC) can also be configured in the slave node via I²C. The slave chip relays the I²C communication from the host directly via its I²C master to the corresponding external component. For preparation in the master and slave chip, the host describes several registers, which then determine at which slave the external component is spoken to and with which address. Then the host can communicate with the corresponding component as if a direct connection existed between it and the master, and as if it weren’t located several meters away, connected only via a two-wire line.

The I2C bus is designed as a polling system, i.e. the master determines when and which nodes to read. However, if the external component has an interrupt output to actively inform the host of a particular situation, then the host this output will be connected to a GPIO of the slave. Upon activation, this GPIO triggers an interrupt at the host, via the master’s IRQ pin. The host can then read the component’s corresponding status registers via I2C. The host controller now knows that it can read the interrupt type register in the A²B master, and can work out that the ADAU1977 in the specific slave node was the trigger, read its register via the bus and obtain an exact report as to what the cause of the diagnostic message was.
The bus is also capable of carrying out diagnostics itself and can therefore recognize and locate all possible errors, including a short circuit in the wires, the battery or the ground, and open or exchanged wires.

Figure 4: Path of I²C communication from the master to the external component in the slave node.


Furthermore, the audio data is monitored during runtime with a parity bit and the control data is CRC-protected. This means the host can precisely assess the stability of the connection at all times.

In addition to the microphone connection with phantom power, connections can be made for any audio sinks and sources. As such, sinks and sources can be distributed on the bus as desired and both can also be present in one node together. A²B therefore not only simplifies software efforts but also helps to reduce costs because cost-efficient and light cables (unshielded twisted pair (UTP)) and plugs can be used.

About the author:
Jörg Hauber is Field Applications Engineer, Automotive Infotainment at Analog Devices.

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News

Share:

Linked Articles
10s