User Tools

Site Tools


doc:appunti:hardware:raspberrypi_thermostat

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
doc:appunti:hardware:raspberrypi_thermostat [2022/01/12 16:44] – [Starting the protherm service at bootstrap] niccolodoc:appunti:hardware:raspberrypi_thermostat [2022/01/12 17:11] (current) – [Installing the old Adafruit LCD library] niccolo
Line 659: Line 659:
  
 ==== Installing the new Adafruit LCD library ==== ==== Installing the new Adafruit LCD library ====
 +
 +**WARNING**: Do not use this library on the old Raspberry Pi model B, it is too CPU intensive! Use the old one.
  
 **WARNING**: The new library is more CPU intensive than the old one. I opened an **[[https://github.com/adafruit/Adafruit_CircuitPython_PCD8544/issues/16|issue on GitHub]]**. So I decided to give-up with the new library and use the old one (see below for installation tips). Here you can see a CPU utilization graph, the first half is using the old library, the second half is using the new one: **WARNING**: The new library is more CPU intensive than the old one. I opened an **[[https://github.com/adafruit/Adafruit_CircuitPython_PCD8544/issues/16|issue on GitHub]]**. So I decided to give-up with the new library and use the old one (see below for installation tips). Here you can see a CPU utilization graph, the first half is using the old library, the second half is using the new one:
Line 708: Line 710:
 ==== Installing the old Adafruit LCD library ==== ==== Installing the old Adafruit LCD library ====
  
-As reported above, the new **Adafruit CircuitPython PCD8544** library is more CPU intensive than the old discontinued one **[[https://github.com/adafruit/Adafruit_Nokia_LCD|Adafruit Nokia LCD]]**. Fortunately enough, the old library does run with Python 3, and I was able to use it on my old Raspberry Pi model B over the operating system derived from Debian 11 Bullseye. So the average CPU usage is **lower than 10%**, instead than **above 30%**!+As reported above, the old and discontinued library **[[https://github.com/adafruit/Adafruit_Nokia_LCD|Adafruit Nokia LCD]]** is less CPU intensive than the new **Adafruit CircuitPython PCD8544**. Fortunately enough, the old library does run with Python 3, and I was able to use it on my old Raspberry Pi model B over the operating system derived from Debian 11 Bullseye. So the average CPU usage is **lower than 10%**, instead than **above 30%**! 
 + 
 +The library is not available from the **[[https://pypi.org/|PyPI]]** repository, so it must be downloaded from the GitHub repository. The library depends upon **Adafruit_GPIO**, which in turn requires **Adafruit_PureIO** and **spidev**. The two Adafruit libraries are available from PyPI, for **spidev** we installed the Debian packaged version. 
 + 
 +<code> 
 +apt install python3-spidev 
 +mkdir -p /usr/local/download/Adafruit-GPIO 
 +cd /usr/local/download/Adafruit-GPIO 
 +pip3 download Adafruit-GPIO 
 +pip3 install Adafruit_PureIO-1.1.9-py3-none-any.whl 
 +pip3 install Adafruit_GPIO-1.0.3-py3-none-any.whl 
 +</code> 
 + 
 +We download the library from its **GitHub repository** and install it using **pip3**: 
 + 
 +<code> 
 +cd /usr/local/download 
 +git clone https://github.com/adafruit/Adafruit_Nokia_LCD.git 
 +cd Adafruit_Nokia_LCD/ 
 +pip3 install . 
 +</code> 
 + 
 +The dependencies tree is the following: 
 + 
 +  * Adafruit_Nokia_LCD 
 +    * Adafruit_GPIO 
 +      * Adafruit_PureIO 
 +      * spidev
  
-FIXME Describe the installation. 
 ==== Installing the Telegram Bot Python library ==== ==== Installing the Telegram Bot Python library ====
  
doc/appunti/hardware/raspberrypi_thermostat.1642002299.txt.gz · Last modified: 2022/01/12 16:44 by niccolo