Working with the Beaglebone Black ================================= Overview -------- This document covers the KubOS Linux features which are specific to the Beaglebone Black target. Reference Documents ------------------- Beaglebone Documentation ~~~~~~~~~~~~~~~~~~~~~~~~ - `Beaglebone Black Web Page `__ - `Beaglebone Black Wiki `__ - `Beaglebone Black Hardware Diagrams `__ - `Beaglebone Black System Reference Manual Rev C `__ Kubos Documentation ~~~~~~~~~~~~~~~~~~~ - :doc:`first-linux-project` - Basic tutorial for creating your first KubOS Linux SDK project - :doc:`using-kubos-linux` - General guide for interacting with KubOS Linux - :doc:`KubOS Linux on the Beaglebone Black ` - Steps to build and load KubOS Linux for the Beaglebone Black USB Connection -------------- As documented in section 7.5 of the :title:`Beaglebone Black System Reference Manual`, an FTDI cable can be connected to the serial debug connector in order to establish a debug console connection. This connection will be passed through to a Kubos Vagrant image as `/dev/FTDI`. Peripherals ----------- The Beaglebone Black has several different ports available for interacting with peripheral devices. Currently, users should interact with these devices using the standard Linux functions. A Kubos HAL will be added in the future to abstract this process. .. note:: KubOS Linux for the Pumpkin MBM2 can be used instead of KubOS Linux for the Beaglebone Black. In this case, some buses and pins won't be available, since they aren't exposed in the MBM2's CSK headers, or are dedicated to other uses. See the :ref:`peripherals-mbm2` section for more information. UART ~~~~ The Beaglebone Black has 5 UART ports available for use: +--------------+--------+--------+---------+---------+ | Linux Device | TX Pin | RX Pin | RTS Pin | CTS Pin | +==============+========+========+=========+=========+ | /dev/ttyS1 | P9.24 | P9.26 | | | +--------------+--------+--------+---------+---------+ | /dev/ttyS2 | P9.21 | P9.22 | | | +--------------+--------+--------+---------+---------+ | /dev/ttyS3 | P9.42 | | P8.34 | P8.36 | +--------------+--------+--------+---------+---------+ | /dev/ttyS4 | P9.13 | P9.11 | P8.33 | P8.35 | +--------------+--------+--------+---------+---------+ | /dev/ttyS5 | P8.37 | P8.38 | P8.32 | P8.31 | +--------------+--------+--------+---------+---------+ .. note:: /dev/ttyS3 (UART3) is TX-only. /dev/ttyS1 and /dev/ttyS2 do not have RTS/CTS due to a pin conflicts with other buses. Users can interact with these ports in their applications using Linux's `termios `__ interface. `A tutorial on this interface can be found here `__ Additionally, the ports can be used from the command line: The ``stty -F {device} [parameters]`` command can be used to configure the port. For example, this command will set the baud rate of `/dev/ttyS1` to 4800:: $ stty -F /dev/ttyS1 4800 The ``echo`` command can be used to transmit basic data out of the TX pin. For example:: $ echo "Hello!" > /dev/ttyS1 The ``cat`` command can be used to read any data from the RX pin. For example:: $ cat < /dev/ttyS1 I2C ~~~ The Beaglebone Black has two user-accessible I2C buses. +--------------+---------+---------+ | Linux Device | SCL Pin | SDA Pin | +==============+=========+=========+ | /dev/i2c-1 | P9.17 | P9.18 | +--------------+---------+---------+ | /dev/i2c-2 | P9.19 | P9.20 | +--------------+---------+---------+ `I2C Standards Doc `__ KubOS Linux is currently configured to support the I2C standard-mode speed of 100kHz. Users will need to add their peripheral device to the system and then open the bus in order to communicate. Once communication is complete, the bus should be closed and the device definition should be removed. Since the peripheral devices will be different for each client, they will need to be `dynamically added in the userspace (method 4) `__. The bus is then opened using the standard Linux ``open`` function and used for communication with the standard ``write`` and ``read`` functions. These functions are described in the `Linux I2C dev-interface doc `__. The buffer used in the ``write`` and ``read`` functions will most likely follow the common I2C structure of "{register, value}" The user program should look something like this: .. code-block:: c /* Add device to system */ system("echo i2cdevice 0x20 > /sys/bus/i2c/devices/i2c-1/new_device"); /* Open I2C bus */ file = open("/dev/i2c-1"); /* Configure I2C bus to point to desired slave */ ioctl(file, I2C_SLAVE, 0x20); /* Start of communication logic */ buffer = {0x10, 0x34}; write(file, buffer, sizeof(buffer)); read(file, buffer, lengthToRead); /* End of communication logic */ /* Close I2C bus */ close(file); /* Remove device */ system("echo 0x20 > /sys/bus/i2c/devices/i2c-1/delete_device); SPI ~~~ The Beaglebone has one SPI bus available with a pre-allocated chip select pin. **SPI Bus 1** +------+-------+ | Name | Pin | +======+=======+ | MOSI | P9.30 | +------+-------+ | MISO | P9.29 | +------+-------+ | SCLK | P9.31 | +------+-------+ | CS0 | P9.28 | +------+-------+ Users can interact a device on this bus using Linux's `spidev interface `__ The device name will be ``/dev/spidev1.0``. An example user program to read a value might look like this: .. code-block:: c #include #include #include #include #include #define SPI_DEV "/dev/spidev1.0" int fd; uint8_t mode = SPI_MODE_0; uint8_t bits = 8; uint32_t speed = 100000; uint16_t delay = 0; uint8_t register, shift_reg; uint8_t value; fd = open(SPI_DEV, O_RDWR); /* Register to read from */ register = 0xD0; struct spi_ioc_transfer tr = { .tx_buf = (unsigned long)®ister, .rx_buf = (unsigned long)®ister, .len = 1, .speed_hz = speed, .bits_per_word = bits, .cs_change = 0, .delay_usecs = delay, }; /* Send request to read */ ioctl(fd, SPI_IOC_MESSAGE(1), &tr); /* Setup buffer to read to */ tr.tx_buf = &value; tr.rx_buf = &value; /* Read data */ ioctl(fd, SPI_IOC_MESSAGE(1), &tr); close(fd); ADC ~~~ The Beaglebone Black has seven analog input pins available: +------+-------+ | Name | Pin | +======+=======+ | AIN0 | P9.39 | +------+-------+ | AIN1 | P9.40 | +------+-------+ | AIN2 | P9.37 | +------+-------+ | AIN3 | P9.38 | +------+-------+ | AIN4 | P9.33 | +------+-------+ | AIN5 | P9.36 | +------+-------+ | AIN6 | P9.35 | +------+-------+ The pins are available through the Linux device ``/sys/bus/iio/devices/iio\:device0/``. A single raw output value can be read from each of the pins via ``/sys/bus/iio/devices/iio\:device0/in_voltage{n}_raw``, where `{n}` corresponds to the AIN number of the pin. Information about setting up continuous data gathering can be found in `this guide from TI `__. To convert the raw ADC value to a voltage, use this equation: .. math:: V_{in} = \frac{D * (2^n - 1)}{V_{ref}} Where: - :math:`D` = Raw ADC value - :math:`n` = Number of ADC resolution bits - :math:`V_{ref}` = Reference voltage The Beaglebone Black uses 12 resolution bits and a reference voltage of 1.8V, so the resulting equation is .. math:: V_{in} = \frac{D * (4095)}{1.8} GPIO ~~~~ The Beaglebone Black has many GPIO pins available for general use. Pinout diagrams are available on the `Beaglebone website `__. Any pin that is not dedicated to a previously mentioned peripheral is available for use. CLI and Script Interface ^^^^^^^^^^^^^^^^^^^^^^^^ To interact with a pin from the command line or from a script, the user will first need to generate the pin's device name: :: $ echo {pin} > /sys/class/gpio/export For example, to interact with pin P8.11, which corresponds with GPIO_45, the user will use: :: $ echo 45 > /sys/class/gpio/export Once this command has been issued, the pin will be defined to the system as '/sys/class/gpio/gpio{pin}'. The user can then set and check the pins direction and value. :: Set pin as output: $ echo out > /sys/class/gpio/gpio45/direction Set pin's value to 1: $ echo 1 > /sys/class/gpio/gpio45/value Get pins's value: $ cat /sys/class/gpio/gpio45/value Once finished, the pin can be released: :: $ echo {pin} > /sys/class/gpio/unexport Application Interface ^^^^^^^^^^^^^^^^^^^^^ This functionality can also be used from a user's application with Linux's sysfs interface. An example program might look like this: .. code-block:: c #include #include #include #include #include #include int fd; int pin = 45; int value = 1; /* Define the pin to the system */ fd = open("/sys/class/gpio/export", O_WRONLY); write(fd, &pin, sizeof(pin)); close(fd); /* Set the pin's direction */ fd = open("/sys/class/gpio/gpio45/direction", O_WRONLY); write(fd, "out", 3); close(fd); /* Set the pin's value */ fd = open("/sys/class/gpio/gpio45/value", O_WRONLY); write(fd, &value, 1); close(fd); /* Read the value back */ fd = open("/sys/class/gpio/gpio45/value", O_RDONLY); char strValue[3]; write(fd, strValue, 1); value = atoi(strValue); close(fd); /* Release the pin */ fd = open("/sys/class/gpio/unexport", O_WRONLY); write(fd, &pin, sizeof(pin)); close(fd); User Data Partitions -------------------- The Beaglebone Black has two user data partitions available, one on each storage device. eMMC ~~~~ The user partition on the eMMC device is used as the primary user data storage area. All system-related `/home/` paths will reside here. /home/usr/bin ^^^^^^^^^^^^^ All user-created applications will be loaded into this folder during the ``kubos flash`` process. The directory is included in the system's PATH, so applications can then be called directly from anywhere, without needing to know the full file path. /home/usr/local/bin ^^^^^^^^^^^^^^^^^^^ All user-created non-application files will be loaded into this folder during the ``kubos flash`` process. There is currently not a way to set a destination folder for the ``kubos flash`` command, so if a different endpoint directory is desired, the files will need to be manually moved. /home/etc/init.d ^^^^^^^^^^^^^^^^ All user-application initialization scripts live under this directory. The naming format is 'S{run-level}{application}'. microSD ~~~~~~~ /home/microsd ^^^^^^^^^^^^^ This directory points to a partition on the microSD device included with the base Beaglebone Black board .. todo:: EEPROM - /home/eeprom (header characters here) This directory points to the available space of the EEPROM storage included with the Beaglebone Black board. There are 4KB of space available for use. .. note:: While EEPROM storage is more stable and safe than MMC/SD, it also has a much more limited number of writes available. This storage should be used carefully.