SPI serial bus access with Node.js on Linux boards like the Raspberry Pi or BeagleBone. All methods have asynchronous and synchronous forms.
spi-device supports Node.js versions 10, 12, 14, 15 and 16.
npm install spi-device
Determine the temperature using a TMP36 analog temperature sensor wired to channel 5 on an MCP3008 SPI A/D converter.
const spi = require('spi-device');
// The MCP3008 is on bus 0 and it's device 0
const mcp3008 = spi.open(0, 0, err => {
// An SPI message is an array of one or more read+write transfers
const message = [{
sendBuffer: Buffer.from([0x01, 0xd0, 0x00]), // Sent to read channel 5
receiveBuffer: Buffer.alloc(3), // Raw data read from channel 5
byteLength: 3,
speedHz: 20000 // Use a low bus speed to get a good reading from the TMP36
}];
if (err) throw err;
mcp3008.transfer(message, (err, message) => {
if (err) throw err;
// Convert raw value from sensor to celcius and log to console
const rawValue = ((message[0].receiveBuffer[1] & 0x03) << 8) +
message[0].receiveBuffer[2];
const voltage = rawValue * 3.3 / 1023;
const celcius = (voltage - 0.5) * 100;
console.log(celcius);
});
});
spi-device enables low-level access to SPI devices. Often, high-level access is required. When this is the case, high-level packages can be built using spi-device. An example of such a package is mcp-spi-adc which provides a high-level API for accessing an MCP3008 SPI A/D converter and will generally be more useful than the low-level demonstration code shown above.
All methods have asynchronous and synchronous forms.
The asynchronous form always take a completion callback as its last argument. The arguments passed to the completion callback depend on the method, but the first argument is always reserved for an exception. If the operation was completed successfully, then the first argument will be null or undefined.
When using the synchronous form any exceptions are immediately thrown. You can use try/catch to handle exceptions or allow them to bubble up.
/dev/spidev0.n
, 1 for /dev/spidev1.n
, .../dev/spidevn.0
, 1 for /dev/spidevn.1
, ...Asynchronous open. Returns a new SpiDevice object. The completion callback gets one argument (err). The SpiDevice object returned should not be used before the completion callback is called.
/dev/spidev0.n
, 1 for /dev/spidev1.n
, .../dev/spidevn.0
, 1 for /dev/spidevn.1
, ...Synchronous open. Returns a new SpiDevice object.
Asynchronous message transfer. An SPI message is an array of one or more read+write transfers. The completion callback gets two arguments (err, message). Returns this.
Synchronous message transfer. An SPI message is an array of one or more read+write transfers. Returns this.
Asynchronously read device configuration options. The completion callback gets two arguments (err, options) where options is an object describing the device configuration options. Returns this.
Synchronously read device configuration options. Returns an object describing the device configuration options.
Asynchronously write device configuration options. The completion callback gets one argument (err). Returns this.
Synchronously write device configuration options. Returns this.
Asynchronous close. Frees system resources used by this instance. The completion callback gets one argument (err). Returns null.
Synchronous close. Frees system resources used by this instance. Returns null.
SPI mode number 0.
SPI mode number 1.
SPI mode number 2.
SPI mode number 3.
An SPI message is an array of one or more read+write transfers. A transfer is an object with the properties listed below. Most of the properties are optional. Note that although both sendBuffer and receiveBuffer are optional, at least one one of them must be specified.
Device configuration options can be optionally specified when a device is
opened with the open
or openSync
methods. They can also be specified at a
later point with the setOptions
or setOptionsSync
methods. When calling
these methods, only the options that need to be set need to be specified in the
options object passed to those methods. All options are optional and the
appropriate defaults will be used for options that are not specified.
The options supported varies from system to system and will depend on the device drivers used on those systems.
Configurations options can be read with the getOptions
and getOptionsSync
methods.
IMPORTANT The semantics of chipSelectHigh have changed with Linux kernel 5. To the best of my knowledge, the chipSelectHigh option no longer serves any purpose when used from user space with Linux kernel 5 and should not be used. With Linux kernel 5, the chip select is assumed to be active low. With Linux kernel 5, if an SPI device has has active high chip select, it's chip select must be controlled manually with a GPIO using a module such as onoff. The chipSelectHigh option has been crossed out below but it's still available for usage on older kernels.