c555c61e2a | ||
---|---|---|
DS3231 | ||
ESP32 | ||
ESP8266 | ||
PicoWeb | ||
bitmap | ||
buildcheck | ||
data_to_py | ||
date | ||
encoders | ||
fastbuild | ||
functor_singleton | ||
import | ||
micropip | ||
mutex | ||
ntptime | ||
phase | ||
power | ||
pyboard_d | ||
quaternion | ||
random | ||
resilient | ||
reverse | ||
sequence | ||
soft_wdt | ||
timed_function | ||
watchdog | ||
LICENSE | ||
PICOWEB.md | ||
QUATERNIONS.md | ||
README.md | ||
RSHELL_MACROS.md | ||
SERIALISATION.md |
README.md
micropython-samples
The first part of this repo comprises assorted code ideas for MicroPython and also serves as an unofficial MicroPython FAQ. Some samples are targeted at Pyboard variants. Some are intended as pointers for programmers rather than being complete solutions. Egregious bugs will be fixed but I may not accept feature requests.
Section 5 is an index to complete applications and modules which are documented and supported.
Please also see the official examples.
0. Index
- Installation guides
1.1 Installing MicroPython libraries
1.2 Fastbuild Build scripts and udev rules
1.3 A web framework Microdot.
1.4 Buildcheck Check firmware build date
1.5 A USB pitfall A problem with platforms which share their filesystem with the PC. - Hardware information and drivers
2.1 ESP32 Pinout and notes on the reference board
2.2 SSD1306 Write large fonts to the SSD1306.
2.3 Pyboard D Some information remains absent or hard to find in the docs.
2.4 DS3231 precision RTC Use cheap hardware to calibrate Pyboard RTC. - Essays General thoughts.
3.1 Resilient A guide to writing resilient WiFi code
3.2 Serialisation Review of MicroPython's five serialisation libraries
3.3 Measurement of relative timing and phase of fast analog signals For Pyboard.
3.4 Import subtleties Ways to save RAM with import statements. - Code samples Samples prefixed Pyboard are Pyboard specific
4.1 Pyboard Mutex Share data between threads and ISR's.
4.2 Pyboard watchdog Access a Pyboard hardware WDT.
4.3 Software Watchdog Cross-platform soft WDT.
4.4 Reverse Reversal algorithms for bytearrays.
4.5 Timed function Time execution with a decorator.
4.6 ESP8266 MQTT benchmark Test performance of MQTT with official library.
4.7 Rotary incremental encoder Fast, simple, proven algorithm.
4.8 Pseudo random number generators
4.9 Verifying incrementing sequences Test communications drivers.
4.10 Bitmaps Non-allocating ways to access bitmaps.
4.11 Functors and singletons Useful decorators.
4.12 Quaternions Scale, move and rotate 3D objects with minimal mathematics.
4.13 A Pyboard power meter One of my own projects.
4.14 NTP time More portable than official driver with other benefits.
4.15 Date Small and simple classes for handling dates. - Module Index Supported code. Device drivers, GUI's, utilities.
5.1 uasyncio Tutorial and drivers for asynchronous coding.
5.2 Memory Device Drivers Drivers for nonvolatile memory devices.
5.3 Inertial Measurement Units Gravity, gyro and magnetic sensors.
5.4 Other hardware drivers
5.5.Communications
5.6 Displays Fonts, graphics, GUIs and display drivers
5.7 Pyboard micropower
5.8 Pyboard DSP Fourier transforms and filters.
5.9 rshell Fork of rshell with text macros.
5.10 Hard to categorise Other modules.
1. Installation guides
1.1 Installing MicroPython libraries
Please read the official docs. This doc describes a simple way to override modules which have been incorporated in firmware as frozen bytecode.
1.2 Fastbuild
Scripts for building MicroPython for various target hardware types and for updating your local source. Now detects and builds for Pyboard D. See docs
1.3 A Web Framework
Paul Sokolovsk's PicoWeb requires his fork of MicroPython and requires some modification to run under official MicroPython.
The solution preferred by MicroPython maintainers is Microdot.
1.4 Buildcheck
Raise an exception if a firmware build is earlier than a given date.
1.5 A USB pitfall
By default the Pyboard's /flash/boot.py
enables MSC (mass storage) mode. This
makes the Pyboard look like a USB stick, making its filesystem visible to the
PC. This helpful feature ignores a fundamental flaw which leads to filesystem
corruption. This is because the USB standard requires mass storage devices to
behave like disks with static content. By contrast a Pyboard can independently
modify the "disk" contents causing chaos.
To fix this, edit /flash/boot.py
so that the usb_mode
line reads:
pyb.usb_mode('VCP')
On reboot the Pyboard will no longer appear as a mass storage device on the PC. Various tools are available to manage the device's storage via USB. I use rshell.
2. Hardware information and drivers
2.1 ESP32
Pinout diagram for the reference board with notes and warnings about reserved pins etc. See this doc. See also this excellent resource.
2.2 SSD1306
A means of rendering multiple larger fonts to the SSD1306 OLED display. The
Writer
class which performs this has been substantially improved and may now
be found as part of this repository.
2.3 Pyboard D
Assorted information not yet in the official docs or hard to find.
2.4 DS3231 precision RTC
This is a low cost precision battery backed real time clock (RTC) accurate to +-2 minutes/year. Two drivers are provided, one portable across platforms and one which is Pyboard specific.
The Pyboard-specific driver provides a facility to calibrate the Pyboard's RTC from the DS3231. Calibration to high precision may be achieved in five minutes.
The drivers are documented here.
Index
3. Essays
3.1 Resilient
A guide to writing reliable ESP8266 networking code. Probably applies to other WiFi connected MicroPython devices.
3.2 Serialisation
A discussion of the need for serialisation and of the relative characteristics of five libraries available to MicroPython. Includes a tutorial on a Protocol Buffer library.
3.3 Measurement of relative timing and phase of fast analog signals
This describes ways of using the Pyboard to perform precision measurements of analog signals of up to around 50KHz. It is documented here.
3.4 Import subtleties
This doc describes a way to save RAM with Damien's lazy
loader, a reload
function, and ways to use wildcard imports.
Index
4. Code samples
4.1 Pyboard mutex
A class providing mutual exclusion enabling hard interrupt handlers and the main program to access shared data in a manner which ensures data integrity.
4.2 Pyboard watchdog
Access the simpler of the Pyboard's watchdog timers.
4.3 Software watchdog
A software watchdog timer with a fixed or variable timeout. Supports temporary suspension and permanent cancellation. The latter can be useful when debugging code to prevent a machine reboot when the application fails, terminates or is interrupted with ctrl-c. See code and comments in the test script.
4.4 Reverse
Fast reverse a bytearray in Arm Thumb assembler: this
reverses the byte order of the array so [1,2,3,4]
becomes [4,3,2,1]
.
Also includes cross-platform Python code to bit-reverse (fast-ish) 8, 16 and 32 bit words.
4.5 Timed function
Time a function's execution using a decorator and implement timeouts using a closure.
Index
4.6 ESP8266 MQTT benchmark
This benchmark tests the performance of MQTT by
periodically publishing while subscribed to the same topic. Measures the
round-trip delay. Uses the official umqtt.simple
library. Adapt to suit your
server address and desired QOS (quality of service, 0 and 1 are supported).
After 100 messages reports maximum and minimum delays.
This connect utility connects in station mode using saved connection details where possible.
4.7 Rotary Incremental Encoder
These devices produce digital signals from a shaft's rotary motion in such a way that the absolute angle may be deduced. Specifically they measure incremental change: it is up to the code to keep track of absolute position, a task which has some pitfalls. This doc discusses this and points to some solutions in MicroPython code.
4.8 Pseudo random number generators
On the Pyboard V1.1, true random numbers may be generated rapidly with
pyb.rng()
which uses a hardware random number generator on the
microcontroller.
There are a few use-cases for pseudo random number generators. Some platforms
lack a hardware generator (e.g. the Pyboard Lite) and some ports don't support
uos.urandom
. There is also a case for running a RNG in an interrupt service
routine.
Pseudo random number generators provide repeatable sequences of numbers which can be an advantage, for example in testing. The RNG is seeded with an initial value. On each call to the function it will return a random number, but (given the same seed) the sequence of numbers following initialisation will always be the same.
See random.py for usage and timing documentation. The yasmarang generator is also included, along with my own cheap random. The latter constrains calculations to 30 bits, allowing its use in an ISR. It comes with no guarantees of random quality and the only statistical test is that the mean converges on the right value. None of these generators are suitable for cryptography.
4.9 Verifying incrementing sequences
When testing communications applications it is often necessary to check for missing, duplicated, or out-of-order messages. To do this, the transmitter test script ensures that messages include an incrementing message number. The receiver script verifies the sequence. The CheckMid class does this, also detecting transmitter reboots
Index
4.10 Bitmaps
A bitmap stored in a pre-allocated, fixed size bytearray may be viewed in two ways:
- As a set of positive integers whose values are constrained within limits.
- As a fixed size one dimensional array of booleans.
These views provide a Pythonic interface while retaining the non-allocating performance advantage relative to native sets and lists.
The file bitmap.py offers classes supporting these views.
The constraint 0 <= value <= max_value
applies where max_value
is a
constructor arg. The max_value
arg defines the size of the underlying
bytearray. For example if max_value
is 255, the bytearray will use 32 bytes.
The constraint applies to member values of a set, and to index values of a
boolean array. So the set will be capable of storing integers from 0 to 255,
and the array will accept indices in the same range.
These classes are lightweight. For example the IntSet
class does not include
all the dunder (magic) methods required to match the native set
class. These
may readily be added as required.
4.11 Functors and singletons
Two simple class decorators for objects useful in hardware interfacing. Documented here.
Singletons denote classes for which only a single instance can ever occur. They can be useful for hardware interface classes. Their use avoids the need for a global instance: the sole instance may be retrieved efficiently from anywhere in the code.
A functor is a class which is accessed via function call syntax. There is only
one instance, like a singleton. Initial access calls the constructor, with
subsequent accesses being via __call__
. As an object it can retain state. As
an example, a functor might have a continuously running task: successive calls
modify the behaviour of the task.
4.12 Quaternions
Quaternions have a reputation for being mathematically difficult. Surely true if you intend using them to write Maxwell's equations (as per Maxwell). If your ambitions are limited to manipulating three dimensional objects or processing IMU data they can be remarkably simple.
The Quaternion
class lets you create, move, transform and rotate 3D objects.
They assume no maths beyond familiarity with an xyz
coordinate system.
Includes a demo where a wireframe cube rotates in response to movements of a
BNo055 IMU, plus a demo of moving wireframe graphics. Neither demo uses trig
functions. See the docs.
4.13 A pyboard power meter
This uses a Pyboard to measure the power consumption of mains powered devices. Unlike simple commercial devices it performs a true vector (phasor) measurement enabling it to provide information on power factor and to work with devices which generate as well as consume power. It uses the official LCD160CR display as a touch GUI interface. It is documented here.
4.14 NTP Time
The official code has a number of drawbacks, mainly a lack of portability.
- It does not check the host device's epoch. This version returns the number of seconds since the host device's epoch. [Official version is now fixed].
- It uses socket timeouts while the docs recommend select.poll as being more portable. This version remedies that.
- This version has very basic support for local time in the form of an offset in hours relative to UTC.
- In the event of a timeout this version returns 0: the caller should check for this and re-try after a period.
Code is here.
It seems impractical to write a portable version of settime
as the
machine.RTC
class is not yet fully portable.
It's worth noting that NTP queries do time out. The default timeout is 1s: on
many WANs this is quite demanding, yet for obvious reasons a longer value does
not benefit precision. Another issue is that issuing NTP queries too frequently
causes the time server to send a KoD (Kiss of Death) packet. The official
version issues a large negative number which causes .settime
to throw an
exception. The version here returns 0: the caller should test for this. I
suspect that this also occurs occasionally in non-KoD circumstances.
Finally, an option is to run the NTP daemon on a local server which can be as simple as a Raspberry Pi. This gives high accuracy and should eliminate timeouts. Install with:
$ sudo apt install -y ntp ntpdate
and point the MicroPython device at the local server with:
ntptime.host="192.168.0.10" # Server address.
ntptime.time()
4.15 Date
The official datetime module
is fully featured but substantial. This Date
class has no concept of time,
but is very compact. Dates are stored as a small int. Contrary to normal MP
practice, properties are used. This allows basic arithmetic syntax while
ensuring automatic rollover. The speed penalty of properties is unlikely to be
a factor in date operations.
The Date
class provides basic arithmetic and comparison methods. The
DateCal
subclass adds pretty printing and methods to assist in creating
calendars.
The classes are documented here
Index
5. Module index
This index references applications and device drivers that I have developed, in some cases as collaborations. This is acknowledged in their respective docs.
Unlike the code samples these are fully documented and supported.
5.1 uasyncio
Tutorial
Main focus: using uasyncio to drive and communicate with hardware devices.
Drivers
Asynchronous device drivers for switches, pushbuttons and ADC's. Also has
information on interfacing interrupts to uasyncio.
Asynchronous monitor Use a
raspberry Pico and a logic analyser or scope to monitor asynchronous code.
Schedule
Schedule events at specified times and dates.
HTU21D
Asynchronous driver for this temperature and humidity sensor.
I2C Slave
Uses the Pyboard's I2C slave mode to implement a full duplex asynchronous
link. Principal use case is for ESP8266 which has only one UART.
GPS
See section 5.4.
5.2 Memory device drivers
EEPROM Support for EEPROM, FRAM and Flash chips and modules. In all cases devices or sets of devices can be configured as a single memory array supporting access either as an array of bytes or as a filesystem.
5.3 Inertial Measurement Units
BNo055 In my view the
best IMU as it performs internal sensor fusion.
MPU9x50 Driver for
the popular InvenSense MPU9250, MPU9150, MPU6050 devices.
BMP180 Driver for the
Bosch BMP180 pressure/temperature sensor.
Fusion Sensor fusion:
combine IMU readings to produce heading, pitch and roll or Quaternion data.
Quaternions The proper way to deal with 3D
rotations. Amazingly requires less maths than Euler angles.
5.4 Other hardware drivers
Asynchronous GPS driver
Thermal IR Support for the
Adafruit 3538 thermal camera. Includes optional bicubic interpolation.
Audio High quality audio
I/O with the Adafruit VS1053 board. Synchronous and asynchronous options.
IR remotes Support for various
infra red protocols in receiver or transmitter ('blaster') applications.
433MHz remote Supports
433MHz remote controlled wall sockets. Captures the signal enabling replay so
that mains devices may safely be controlled at low cost.
BME280 A bugfix fork of an
abandoned project. Supports the BME280 combined temperature, pressure and
humidity sensor on ESP8266.
5.5 Communications
Asynchronous MQTT
This improves on official MQTT drivers by recovering from WiFi outages and
offering asynchronous operation.
IOT An IOT solution. Provides
a socket-like interface between WiFi clients and a server with wired ethernet.
The interface is resilient in the presence of WiFi outages. The server (e.g.
Raspberry Pi) does any internet work, improving security.
MessagePack Serialisation
with usage almost identical to ujson
but with message compression and other
advantages.
Radio Simplify use of the
officially supported NRF24l01 radio.
5.6 Displays
Fonts, graphics, GUIs and display drivers.
All GUIs except e-paper are based on uasyncio. Where input is supported a callback-based interface is povided.
font-to-py Converts
industry standard font files to Python source which may be frozen as bytecode.
Files use minimal RAM when frozen.
writer
A simple way to render text to displays where the driver is subclassed from
framebuf
.
nano-gui Simple output
only GUI for displays where the driver is subclassed from framebuf
. Repo
includes display drivers for various displays including TFT, OLED, ePaper and
Sharp. Supports a range of hosts with low RAM usage.
micro-gui Derived from
nano-gui and supporting the same displays and hosts, this provides for user
input via pushbuttons or a navigation joystick.
LCD160CR Touch GUI
for the official display module.
TFT-GUI A fast touch GUI
for large displays based on SSD1963 controller with XPT2046 touch controller.
RA8875-GUI Touch GUI for
large displays based on the RA8875 controller (e.g. from Adafruit).
e-paper GUI for the
Embedded Artists 2.7 inch e-paper display. Pyboard only. Sadly these displays
are now obsolete.
5.7 Pyboard micropower
micropower Support for low power applications on Pyboard 1.x and Pyboard D.
5.8 Pyboard DSP
fourier DFT using Arm
Thumb assembler. Primarily for processing data received on an ADC.
Filters FIR filters
using ARM Thumb assembler. Using an online utility you can go from a graph
of required frequency response to a filter implementation.
5.9 rshell
rshell is a fork of Dave Hylands' excellent utility. My fork adds text macros to improve its power and usability, notably when maintaining complex Python packages. The fork includes documentation, but these notes provide a usage example where each project has its own set of automatically loaded macros.__
5.10 Hard to categorise
data to py Convert arbitrary objects to Python files which may be frozen as bytecode. Can be used to freeze images for display.
Index
License
Any code placed here is released under the MIT License (MIT).
The MIT License (MIT)
Copyright (c) 2016 Peter Hinch
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.