More Than a Hobby..

How to Backup your Raspberry Pi SD Card

In this guide, we will be showing you various ways on how to backup your Raspberry Pi and also how to restore it. Backing up is a crucial task that you should be doing often, especially if you make many changes or are storing data on it.

One thing to note is that some backup methods are going to be way more thorough the others, for instance backing up your Raspberry Pi SD Card image is going to be more reliable than just backing up all the files to a USB device since the image is a replication of all partitions on the SD Card.

We will be exploring two different methods of backing up your Raspberry Pi in this guide, the two different methods that we will be showing you how to do is the following:

 Backing up your Raspberry Pi to a computer using a SD Card Reader
 Backup the Raspberry Pi to a USB device

Personally, we recommend backup to a computer as it is the most robust method. Mainly since the Raspberry Pi doesn’t have the best USB speeds or network speeds so creating backups can take considerable time and put a fair bit of stress on the Raspberry Pi’s limited resources. As an added bonus the backup we make using a tool like win32diskimager is a 1 to 1 copy of the SD Card meaning every byte should be the same.

Backing up the Raspberry Pi is an important step in making sure you don’t lose any data when you go to make big changes to your system. This is important if you have stuff that you would prefer not to lose if you’re looking at doing a complicated Pi project, upgrade or something else to the operating system.

 Backing up your Raspberry Pi SD Card

To start off we are first going to show you how to backup your Raspberry Pi SD Card as an image. To do this tutorial you will need to have a SD Card reader handy.

To begin this tutorial, please first turn off your Raspberry Pi by running the following command on your Pi’s terminal:

sudo shutdown now

Once it has shut down, disconnect the power and remove your SD Card. Place the SD Card into your SD Card reader and proceed onwards with this section of the guide.

This section of the guide will be split into 3 parts, one for Windows, one for Mac OS X and one for Linux based systems.

 Backing up your Raspberry Pi SD Card on Windows

1. To backup our Raspberry Pi’s SD Card on Windows we will be utilizing the imaging tool that is called win32diskimager. Win32diskimager is an incredibly useful tool that can read and write images to USB Sticks or SD/CF Cards. It is a tool many will be familiar with since it is used often to write the initial Raspbian image to your Raspberry Pi.

We will start off this section of the guide by downloading and installing this tool by going to win32diskimager’s homepage.

2. With win32diskimager now installed, open it up. You should be greeted with a screen like we have shown below.

For now we need to make sure the right drive has been selected by clicking the dropdown box under “Device:” (1.) and ensuring the correct drive letter of our SD Card is selected.

Afterwards, click the folder icon (2.) that is directly to the left of the device dropdown box.

Selecting drive to backuph


3. You will now be shown a file selection screen, however we won’t be selecting any file. Instead navigate to the folder you want to keep your backup file. Once within this folder, type in the name you want to give your backup file in the textfield (1.) at the bottom as shown in the screenshot below.

Make sure you have .img at the end of the name you choose to use. For example, we named our backup file RaspberryPiMyLifeUp.img

Once you have typed in your file name you want to utilize, click the Open button.

Backing up Raspberry Pi Selecting a disk image


4. Finally with our device set and our new filename set we can begin the backup process. To do this we simply just need to click the “Read” button. Please note that this can take some time as it is a complete backup of your SD Card, meaning every single byte is replicated.

Raspberry Pi Backup SD Card

 Restoring your Raspberry Pi Backup on Windows

1. Now when it comes time that you need to make use of your full SD Card backup we will need to once again make use of win32diskimager. Before opening the software, make sure you have inserted your SD Card into a SD Card Reader that is plugged into your computer.

With the win32diskimager software now opened, lets click the dropdown box that is located under the “Device” (1.) header and ensuring that the correct drive letter of our SD Card is selected.


After you have set the device we can now, click the folder icon (2.) that is located to the left of the dropdown box.

Selecting drive to backup

2. Now that we have the file browser loaded up, let’s find the backup file that we created earlier. Once you have found it either double click the file, or single click it and click the “Open” button.

3. With our backup image now selected, we can now proceed with writing the backup file to the SD Card. Please note this will wipe everything currently on the SD Card so make sure that you have the correct device selected. When you are happy you have everything correct, click the “Write” button.

Raspberry Pi write backup image


4. Once the writing process has completed you should now have a SD Card that is in the exact state of when you originally made the backup. Thanks to the full image backup it means all partitions on the card are restored.

 Backup up your Raspberry Pi SD Card on OS X

1. With your SD Card inserted into a card reader on your Mac we can begin the process of making a full image backup of your Raspberry Pi. Unlike Windows, we can easily just use the terminal to do this.

To proceed with this tutorial, start off by opening the Terminal application.

2. With the Terminal application now open on your Mac device, we need to utilize the following command. This command will display all available disks on your device.

diskutil list

Within this list, look for your SD Card by looking for a disk that is about the size of your SD Card. For instance, with a 16gbSD Card you should be looking at a "_partition_scheme" of about 16gb. You will also notice that there is likely a partition called "boot".

Once you have found your SD Card in this list, take notice of the mount location. For instance, my own SD Card was under /dev/disk1.

3. Now still within the terminal on your Mac, we need to utilize the following command. This command will basically create a copy of your SD Cards image, and save it to your home directory as PiSDBackup.dmg (The file format being a disc image)

Make sure you swap out /dev/disk1 with whatever you found when using the diskutil list command.

sudo dd if=/dev/disk1 of=~/PiSDBackup.dmg

This command can take some time to complete as it requires reading the entire SD Card to the disk. The command also provides no feedback on how far along it is, so please be patient and wait for the prompt to enter another command to reappear before removing your SD Card from your Mac.

 Restoring your Raspberry Pi Backup on OS X

1. Now that you need to restore your Raspberry Pi to its backup we need to again use the diskutil list command to find our SD Card. Remember to as before, take note of partition sizes to find your SD Card.

diskutil list

2. Now before we can write to the SD Card we will need to unmount it. The reason for this is that OSX will attempt to write to it at the same time, unmounting the SD Card prevents this from happening.

Run the following command on your Mac device to unmount the card, again making sure that you replace /dev/disk1 with the one you found using the diskutil list command.

diskutil unmountDisk /dev/disk1

3. Finally we can now write the image back to the SD Card, please be prepared for this to take some time as it involves rewriting the entire SD Card. Remember to change out /dev/disk1 with the mount location you grabbed using the diskutil list command.

sudo dd if=~/PiSDBackup.dmg of=/dev/disk1

Like reading the SD Card to a disk image file, the process of writing the image also takes a long time. As an added note, the dd tool doesn't show any write progress so please be patient and wait till the enter command prompt reappears.

4. Once the write process has completed we can now eject the SD Card from the Mac so we can continue using our Raspberry Pi Backup. To eject the SD Card, we will need to utilize the command below.

Your Raspberry Pi's SD Card should now be in the exact same state as when you made the original backup.

sudo diskutil eject /dev/rdisk3

 Backing up your Raspberry Pi SD Card on Linux

1. Before starting to backup up your Raspberry Pi's SD Card on Linux we will first run a command with the reader not plugged in. The reason for this is makes it much easier to see which device is which.

df -h

This command will return something like what is shown below.

Filesystem      Size  Used Avail Use% Mounted on
/dev/root        15G  4.0G  9.6G  29% /
devtmpfs        458M     0  458M   0% /dev
/dev/mmcblk0p1   41M   21M   21M  52% /boot

2. Now insert your SD Card reader back into your Linux computer and again run the following command again, but this time take note of the additional entries.

df -h

This command will return something like what you got above, but with an additional entry, in our case /dev/sda1 is that additional entry, with your entry, remove the partition number. For instance, /dev/sda1 will become /dev/sda. We need to do this as we want to write to the entire drive and not just a singular partition.

Filesystem      Size  Used Avail Use% Mounted on
/dev/root        15G  4.0G  9.6G  29% /
devtmpfs        458M     0  458M   0% /dev
/dev/mmcblk0p1   41M   21M   21M  52% /boot
/dev/sda1       3.7G   75M  3.6G   3% /media/boot

3. We can now utilize dd, the same tool we make use of on Mac OS X since it is also built on Linux. We can utilize the following command to begin dumping the SD Cards image to our home directory.

Make sure you replace /dev/sda with the filesystem that you grabbed using the df -h command.

sudo dd if=/dev/sda of=~/PiSDBackup.img

The process of backing up your Raspberry Pi can take some serious time, so be patient and wait. The dd tool provides no feedback, so you will have to wait until the input command returns to your terminal.

Once it reappears you will have successfully backed up your Raspberry Pi.

 Restoring your Raspberry Pi Backup on Linux

1. Now that you have made a backup of your Raspberry Pi you will want to at some stage make use of this. To do this we will need to again go through the process of finding out the location of our filesystems.

Use the df -h command like we did in the first segment of this guide, though this time you might have more than one partition pop up for your SD Card such as /dev/sda1 and /dev/sda2. Take note of all new entries as you will need to unmount all of them.

2. Now that we have all our partition locations ready, we can unmount each of them by running the following command for each one. Switching out /dev/sda1 for the locations that you got in the previous step.

sudo unmount /dev/sda1

3. With all the partitions now unmounted let's write our backup image to the SD Card. We can do that by running the following command. Remember swap out /dev/sda with your devices mount location.

sudo dd bs=4M if=~/PiSDBackup.img of=/dev/sda

 Backing up your Raspberry Pi to a USB Drive

1. Before plugging in your USB device that you want to keep the backups on let's first run the following command to find out the current filesystems that are available to us.

df -h

This command will return something like what is shown below.

Filesystem      Size  Used Avail Use% Mounted on
/dev/root        15G  4.0G  9.6G  29% /
devtmpfs        458M     0  458M   0% /dev
/dev/mmcblk0p1   41M   21M   21M  52% /boot

2. Now insert your USB Device into your Raspberry Pi and run the following command, take note of any new entries that pop up.

df -h

This command will return something like what is shown below.

Filesystem      Size  Used Avail Use% Mounted on
/dev/root        15G  4.0G  9.6G  29% /
devtmpfs        458M     0  458M   0% /dev
/dev/mmcblk0p1   41M   21M   21M  52% /boot
/dev/sda1       3.7G   75M  3.6G   3% /media/pi/MYBACKUPDRIVE

3. With our USB device now showing on the list, we need to look at the "Mounted on" location. In our case this is /media/pi/MYBACKUPDRIVE. This will be the location that we will write our backup images to. So, make note of it for later in the tutorial.

Now with our backup location now handy we can download the backup script that we are going to use, this script was written by a user on the Raspberry Pi forums called Jinx.

Let's clone the script to our Raspberry Pi by running the following two commands.

cd ~
git clone https://github.com/lzkelley/bkup_rpimage.git

4. With the script now saved to the Raspberry Pi we can start to make use of it. We can do an initial backup by running the command below on our Raspberry Pi. Make sure you replace /media/pi/MYBACKUPDRIVE with your own mount location that you grabbed in the previous step.

sudo sh bkup_rpimage.sh start -c /media/pi/MYBACKUPDRIVE/rpi_backup.img

Basically this script will create a dummy image then launch an rsync process to copy all the files from the system to the dumby image. Please note that the initial backup can take up to an hour to complete.

5. Now that we have created our initial backup file and know that the script is working as intended we can move onto automating the backup. To do this we will be making use of cronjobs.

One thing to decide on how is whether you want an incremental backup or multiple backups. An incremental backup just updates the original backup and doesn't generate a new file.

Run the following command on your Raspberry Pi to begin editing the crontab.

sudo crontab -e

6. In the crontab editor, add one of the following lines to the bottom of the file. This will make a backup every day. If you want to edit the cron timings you can use the Crontab guru website to easily work them out.

Incremental Backup

0 0 * * * sudo sh ./home/pi/bkup_rpimage/bkup_rpimage.sh start -c /media/pi/MYBACKUPDRIVE/rpi_backup.img

Multiple Backup

0 0 * * * sudo sh ./home/pi/bkup_rpimage/bkup_rpimage.sh start -c /media/pi/MYBACKUPDRIVE/rpi_$(date +%Y-%m-%d).img

Now save the file by pressing Ctrl + X then pressing Y and then hitting Enter.

7. You should now have an automated backup system up and running that will continually backup your Raspberry Pi to your USB device. To restore these images follow our Restoring guides located in the SD Card section of this guide.

I hope that you now know how to backup your Raspberry Pi SD card on both your main system or backing up directly to a USB device. You can now continue on with working on some Raspberry Pi projects or just using your Pi as you would normally. If you run into any issues or have any feedback feel free to drop a comment below.

openHABian hassle-free openHAB Setup





sudo raspi-config

sudo openhabian-config

 Clean cache:

sudo rm -rf /var/lib/openhab2/tmp/*

sudo rm -rf /var/lib/openhab2/cache/*

 Stop / Start openhab2:

sudo systemctl status openhab2

sudo systemctl stop openhab2

sudo systemctl start openhab2

 install htop (a better version of top)

sudo apt-get install htop

sudo htop


################# openHAB1 ####################################################
## Ip =
## Release = Raspbian GNU/Linux 9 (stretch)
## Kernel = Linux 4.9.59-v7+
## Platform = Raspberry Pi 2 Model B Rev 1.1
## Uptime = 0 day(s). 0:0:43
## CPU Usage = 55.1 % avg over 4 cpu(s) (4 core(s) x 1 socket(s))
## CPU Load = 1m: 3.39, 5m: 0.93, 15m: 0.32
## Memory = Free: 0.71GB (75%), Used: 0.23GB (25%), Total: 0.95GB
## Swap = Free: 0.09GB (100%), Used: 0.00GB (0%), Total: 0.09GB
## Root = Free: 11.71GB (87%), Used: 1.70GB (13%), Total: 14.02GB
## Updates = 0 apt updates available.
## Sessions = 1 sessions
## Processes = 137 running processes of 32768 maximum processes

Welcome to __ _____ ____ _
____ ____ ___ ____ / / / / | / __ )(_)___ _____
/ __ \/ __ \/ _ \/ __ \/ /_/ / /| | / __ / / __ `/ __ \
/ /_/ / /_/ / __/ / / / __ / ___ |/ /_/ / / /_/ / / / /
\____/ .___/\___/_/ /_/_/ /_/_/ |_/_____/_/\__,_/_/ /_/
openHAB 2.2.0-1 (Release Build)

Looking for a place to get started? Check out 'sudo openhabian-config' and the
documentation at http://docs.openhab.org/installation/openhabian.html
The openHAB dashboard can be reached at http://openHAB1:8080
To interact with openHAB on the command line, execute: 'openhab-cli --help'

[12:34:13] openhabian@openHAB1:~$

Stopping SD Card Corruption on Raspberry Pi’s Raspbian

The following are instructions for minimizing SD card writes for Raspberry Pi’s “Raspbian” Distribution.

If you’re like me, you’ve run into a corrupted SD card too many times to not become hell-bent on making it never happen again. I have the following setup, and it seems to be working well for me.

The biggest offender for Filesystem writes on any linux system is logging. If you are like me, you don’t really look at /var/log after a recycle anyways. This area, and /var/run, a location where lock files, pid files and other “stuff” shows up, are the most common areas for mess-ups. Take a look at your blinking FS light on the board. Our goal is to make that light stay off as long as possible.

Set up tmpfs mounts for worst offenders. Do other tweaks.

Linux has with it the concept of an in-memory filesystem. If you write files to an in-memory filesystem, they will only exist in memory, and never be written to disk. There are two common mount types you can use here: ramfs, which will continue to eat memory until your system locks up (bad), and tmpfs, which sets a hard upper limit on mount size, but will swap things out if memory gets low (bad for raspberry pi, you will probably be hard stopping your device if it is low on memory).

We will first solve the usual corruption culprit and then move on to making sure we are covered when our programs decide to blow up.

The following two lines should be added to /etc/fstab:

none        /var/run        tmpfs   size=1M,noatime         00
none        /var/log        tmpfs   size=1M,noatime         00

UPDATE (unverified): I have been told that /var/run is now a symlink to a tmpfs filesystem, anyways, so you may not need to add /var/run anymore, and adding it may cause issues.

There’s more, however. By default, linux also records when a file was last accessed. That means that every time you read a file, the SD card is written to. That is no good! Luckily, you can specify the “noatime” option to disable this filesystem feature. I use this flag generously.

Also, for good measure, i set /boot to read-only. There’s really no need to regularly update this, and you can come back here and change it to “defaults” and reboot when you need to do something.

After this, /etc/fstab should look as follows:

proc            /proc               proc    defaults                    0   0
/dev/mmcblk0p1  /boot               vfat    ro,noatime                  0   2
/dev/mmcblk0p2  /                   ext4    defaults,noatime            0   1
none            /var/run            tmpfs   size=1M,noatime             0   0
none            /var/log            tmpfs   size=1M,noatime             0   0

UPDATE (unverified): I have been told that /var/run is now a symlink to a tmpfs filesystem, anyways, so you may not need to add /var/run anymore, and adding it may cause issues.

Go ahead and reboot now to see things come up. Check the Filesystem light on your raspberry pi after it’s fully booted. You should see no blinking at all.

Disable swapping

As a note, since i have done the changes above, i have not corrupted an SD card. I’m not saying I’ve tried very hard, but it is much better, even with power plug pulls, which i tried a few of after doing these changes.

One protection against SD card corruption is an optional, but potentially “I’m glad i did that” change to disable swapping.

The raspberry pi uses dphys-swapfile to control swapping. It dynamically creates a swap partition based on the available RAM. This tool needs to be used to turn off swap, and then needs to be removed from startup.

Run the following commands to disable swapping forever on your system:

sudo dphys-swapfile swapoff
sudo dphys-swapfile uninstall
sudo update-rc.d dphys-swapfile remove

After doing this, call free -m in order to see your memory usage:

pi@raspberrypi ~ $ free -m
             total       used       free     shared    buffers     cached
Mem:           438         59        378          0          9         27
-/+ buffers/cache:         22        416
Swap:            0          0          0

If you reboot, and run a free -m again, you should still see swap at 0. Now we don’t have to worry about tmpfs filesystems swapping out to hard disk!


As you go on and install other tools and frameworks on your raspberry pi (like ROS), you need to be aware of where caches and logfiles are written to. If, for example, you use ROS, the nodes will, by default, log to ~/.ros/log/. For these sorts of things, and for ROS in particular, point these logs to a folder on /dev/shm. This mount is created by default on linux boxes, and is a tmpfs filesystem that is globally writable by default. Mine has 88 MB.
For example, pointing ROS logs to this file could be done by adding the following to your .bashrc:

export ROS_LOG_DIR=/dev/shm/rosLogs

If you feel like creating more mounts, feel free, but I have run into ownership issues that required having another script on startup that had to be run as root to chown directories to their proper owners.

Getting to 100%

The only way to fully protect against SD card corruption is to mount your root filesystem as readonly. For me, this was too much of a usability issue. I am editing files and installing new packages too often for this to be feasible. The steps listed above should, however, cover you 98% of the time. Try not to pull power while you’re editing files or installing new packages on your device, and you should be fine. Still make backup images of your SD card every once in a while! This is a best practice no matter what!

Happy Hacking!

Use the Adafruit PCA9685 with a Raspberry PI

Use the Adafruit PCA9685 with a Raspberry PI, in Java Raspberry PI
to drive up to 16 servos

The Servo Driver we are talking about here is the Adafruit 16-Channel 12-bit PWM/Servo Driver - I2C interface - PCA9685.

The original tutorial from Adafruit is available here. The language it uses is Python. 

This board has an I2C (pronounce I "square" C) interface, that needs to be enabled on the Raspberry PI. 
I2C stands for Inter-Integrated-Circuit. 
Arduino & Adafruit provide python libraries that go along with their components. The code we present here is in big part an adaptation of this code.

Enable I2C on the Raspberry PI

We will show how to enable I2C on a Model B Raspberry PI. There is a lot of documentation on the web on how to do it on whatever version on the RasPI you have. We are not going to duplicate the web in this document... 
We will need to edit several configuration files. Use the editor you prefer (nanovigedit, ...), as root (sudo).

  1. Edit /etc/modules 
    Make sure its content eventually looks like this:
    Save your modifications, if you have done any.
  2. Install some I2C Utilities (in python) 
     Prompt> sudo apt-get install python-smbus
     Prompt> sudo apt-get install i2c-tools
  3. Edit /etc/modprobe.d/raspi-blacklist.conf, comment the two lines it contents: 
     blacklist spi-bcm2708
     blacklist i2c-bcm2708

     # blacklist spi-bcm2708
     # blacklist i2c-bcm2708
    Save your modifications, if you have done any.
  4. By now (the PCA9685 needs to be wired for that), you should be able to run the i2cdetect utility, which will will use to know the address of the servo driver:
     Prompt> sudo i2cdetect -y 1
          0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
     00:          -- -- -- -- -- -- -- -- -- -- -- -- --
     10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     40: 40 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     70: 70 -- -- -- -- -- -- --
    As seen above, the I2C addresses range from 0x03 to 0x77 (binary 0000011 to 1110111). This way, several boards can be used simultaneously from the RasPI, as long as their I2C addresses are different, even if the GPIO interface pins are connected to several different boards. 
    For example, if you have connected
    • A PCA9685 (servo driver)
    • A BMP180 (Pressure, Altitude, Temperature)
    • A LSM303 (Magnetometer & Accelerometer)
    you would have
     Prompt> sudo i2cdetect -y 1
          0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
     00:          -- -- -- -- -- -- -- -- -- -- -- -- --
     10: -- -- -- -- -- -- -- -- -- 19 -- -- -- -- 1e --
     20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     40: 40 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     70: 70 -- -- -- -- -- -- 77
    where, in addition to the above,
    • 0x19 would be the magnetometer address (LSM303)
    • 0x1e would be the accelerometer address (LSM303)
    • 0x77 would be the BMP180 address


The connection setting is quite simple.

PCA9685Raspberry PI
VCC 1. 3.3 VDC Power
GND 6. 0V (Ground)
SDA 3. SDA0 (I2C)
SCL 5. SCL0 (I2C)
Raspberry PI P1 Connector map Click to enlarge Click to enlarge 
The Raspberry PI, connected to a PCA9685, with 2 servos.

This diagram shows the name of the pins you want to use on the GPIO Connector of the Raspberry PI. As we said before, we need to use the pins 1, 3, 5, and 6.


You will need to know what address to talk to to reach the PCA9685. Run the i2cdetect utility.

 Prompt> sudo i2cdetect -y 1
      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
 00:          -- -- -- -- -- -- -- -- -- -- -- -- --
 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
 40: 40 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
 70: 70 -- -- -- -- -- -- --

We will use the 0x40 address to reach the PCA9685. 
The Java code we are interested in is in the class named adafruiti2c.AdafruitPCA9685. Check it out in the repository. We are going to comment only on the main method of this class, which is a basic example. It assumes that 2 servos are connected to the PCA9685, on the 15 and 14 slots.
The PCA9685 has 16 such slots, numbered - on the board - from 0 to 15. See them on the picture at the top of this document. The slots we are talking about are the 16 columns of 3 holes drilled at the bottom of the board. Their number is printed on top of each column of 3 holes.

  public static void main(String[] args)
    AdafruitPCA9685 servoBoard = new AdafruitPCA9685();   // 0x40 is the default address
    servoBoard.setPWMFreq(60); // Set frequency to 60 Hz
    int servoMin = 150;   // Min pulse length out of 4096
    int servoMax = 600;   // Max pulse length out of 4096
    final int CONTINUOUS_SERVO_CHANNEL = 14;
    final int STANDARD_SERVO_CHANNEL   = 15;
    for (int i=0; i<10; i++)
      System.out.println("i=" + i);
      servoBoard.setPWM(STANDARD_SERVO_CHANNEL,   0, servoMin);
      servoBoard.setPWM(CONTINUOUS_SERVO_CHANNEL, 0, servoMin);
      servoBoard.setPWM(STANDARD_SERVO_CHANNEL,   0, servoMax);
      servoBoard.setPWM(CONTINUOUS_SERVO_CHANNEL, 0, servoMax);
    servoBoard.setPWM(CONTINUOUS_SERVO_CHANNEL, 0, 0); // Stop the continuous one
    System.out.println("Done with the demo.");

This code loops 10 times and makes the servos go from a value of 150 to a value of 600, and vice versa. It waits for 1000 milliseconds between each message to the servos. 
For this demo, the standard servo we use is the Standard servo - TowerPro SG-5010 - 5010
As seen in its documentation, the servo rotates on 180°, 90° on each side. Position "0" (1.5ms pulse) is middle, "90" (~2ms pulse) is all the way to the right, "-90" (~1ms pulse) is all the way to the left. 
We see we use a 60 Hz frequency. This means we will have 60 cycles per second. Each cycle contains 4096 ticks. 
The value minServo is 150. 
The value maxServo is 600. 
A full cycle (of 4096 ticks) at 60 Hz will take 1 / 60 ~ 16 ms. 
150 ticks would take 150 / (60 * 4096) seconds, which is ~ 0.61 ms. 
600 ticks would take 600 / (60 * 4096) seconds, which is ~ 2.44 ms. 

More generically

  • TickTime = 1 / (freq * 4096).
  • PulseTime = TickTime * nbTicks.

It is quite easy - and convenient - to put those formulas in a spreadsheet. 
Based on those formulas, we should have:

Pulse Width
in ms
Nb Ticks
1 246
1.5 369
2 492

The values given in the Adafruit tutorial (150 & 600) happens to be a bit different of what the documentation of the servo says..., but they actually work fine, and this is the ones we use in the code:

Pulse Width
in ms
Nb Ticks
1 150
1.5 375
2 600


We run the code from a script named servo:

 sudo java -cp $CP adafruiti2c.AdafruitPCA9685
 Prompt> ./servo

And here is the result

 Prompt> ./servo
 Connected to bus. OK.
 Connected to device. OK.
 Setting PWM frequency to 60 Hz
 Estimated pre-scale: 100.72526
 Final pre-scale: 101.0
 Done with the demo
The console output

And what next?

Drive the Raspberry PI, on its own cart... 
Fuel the Raspberry PI with a solar panel, oriented by servos to face the sun for greater efficiency.

Something to be aware of...

I followed the PWM instructions, came up with several spreadsheets for the calculations, and... it just did not work correctly. 
Both standard and continuous servos are supposed to be triggered with pulses having width ranging from 1 to 2 ms. As seen above, at 60 Hz, that makes 246 to 492 ticks. 
In reality, after many manual trials, for a frequence of 60 Hz, I used:

For the Standard Servo Between 122 & 615 ticks
For the Continuous Servo Between 340 & 410 ticks, stop at 375

Those values are suprisingly different from the theorical ones..., but they work. 
See in the code the classes named adafruiti2c.samples.DemoContinuous and adafruiti2c.samples.DemoStandard
Those are the ones I will use as bases for robotic projects.

Serial Port Sample

Serial Port Sample

This sample has been verified to work with the following configuration:

Windows 10 IoT Core: version 10.0.10586.0Visual Studio: Visual Studio 2015 update 1Windows SDK: version 10586 (Included with Visual Studio Update 1. Check Visual Studio install options)

We’ll create a simple app that allows communication between a desktop and an IoT device over a serial interface.

This is a headed sample. To better understand what headed mode is and how to configure your device to be headed, follow the instructions here.

Load the project in Visual Studio

You can find the C# and C++ versions of this sample here. Make a copy of the folder on your disk and open the project from Visual Studio.

This app is a Universal Windows app and will run on both the PC and your IoT device.

Wiring the serial connection

You have two options for wiring up your board:

  1. using the On-board UART controller
  2. using a USB-to-TTL adapter cable such as this one

On-board UART (MinnowBoard Max)

The MinnowBoard Max has two on-board UARTs. See the MBM pin mapping page for more details on the MBM GPIO pins.

  • UART1 uses GPIO pins 6, 8, 10, and 12.
  • UART2 uses GPIO pins 17 and 19.

In this sample we will use UART2.

Make the following connections:

  • Insert the USB end of the USB-to-TTL cable into a USB port on the PC
  • Connect the GND wire of the USB-to-TTL cable to Pin 1 (GND) on the MBM board
  • Connect the RX wire (white) of the USB-to-TTL cable to Pin 17 (TX) on the MBM board
  • Connect the TX wire (green) of the USB-to-TTL cable to Pin 19 (RX) on the MBM board

Note: Leave the power wire of the USB-to-TTL cable unconnected.

On-board UART (Rasperry Pi2)

The Rasperry Pi2 has one on-board UART. See the Raspberry Pi 2 Pin Mappings page for more details on the MBM GPIO pins.

  • UART0 uses GPIO pins 6 (GND), 8 (TX) and 10 (RX).

Make the following connections:

  • Insert the USB end of the USB-to-TTL cable into a USB port on the PC
  • Connect the GND wire of the USB-to-TTL cable to Pin 6 (GND) on the RPi2 board
  • Connect the RX wire (white) of the USB-to-TTL cable to Pin 8 (TX) on the RPi2 board
  • Connect the TX wire (green) of the USB-to-TTL cable to Pin 10 (RX) on the RPi2 board

Note: Leave the power wire of the USB-to-TTL cable unconnected.

On-Board UART (DragonBoard 410c)

The DragonBoard has two on-board UARTs.

  • UART0 uses GPIO pins 3, 5, 7, and 9.
  • UART1 uses GPIO pins 11 and 13.

In this sample, UART1 will be used. Make the following connections:

  • Insert the USB end of the USB-to-TTL cable into a USB port on the PC
  • Connect the GND wire of the USB-to-TTL cable to pin 1 (GND)
  • Connect the RX wire (white) of the USB-to-TTL cable to pin 11 (UART1 TX)
  • Connect the TX wire (green) of the USB-to-TTL cable to pin 13 (UART1 RX)

NOTE: Leave the power wire of the USB-to-TTL cable unconnected.

Using USB-to-TTL Adapter

Note: Only USB-to-TTL cables and modules with Silicon Labs chipsets are natively supported on MinnowBoard Max and Raspberry Pi2.

You will need:

Make the following connections:

  • Insert the USB end of the USB-to-TTL cable into a USB port on the PC

  • Insert the USB end of the USB-to-TTL module into a USB port on the RPI2 or MBM device

  • Connect the GND pin of the USB-to-TTL module to the GND wire of the USB-to-TTL cable

  • Connect the RX pin of the USB-to-TTL module to the TX wire (green) of the USB-to-TTL cable

  • Connect the TX pin of the USB-to-TTL module to the RX wire (white) of the USB-to-TTL cable

Leave the power pin of the USB-to-TTL cable unconnected. It is not needed.

Below is an image of our USB-to-TTL module connected to a USB port in our RPi2. The GND, TX, and RX pins of the module are connected to the GND, RX, TX wires of the USB-to-TTL cable that is connected to our PC.

Deploy and Launch the SerialSample App

Now that our PC and RPi2 or MBM are connected, let’s setup and deploy the app. If you are not familiar with how to set the target device and target architecture in Visual Studio see this section for details.

  1. Download the Visual Studio 2015 SerialSample source project.

  2. Make two separate copies of the app. We’ll refer to them as the ‘Device copy’ and ‘PC copy’.

  3. Open two instances of Visual Studio 2015 on your PC. We’ll refer to these as ‘Instance A’ and ‘Instance B’.

  4. Open the Device copy of the SerialSample app in VS Instance A.

  5. Open the PC copy of the SerialSample app in VS Instance B.

  6. In VS Instance A, configure the app for deployment to your RPi2 or MBM device)

    *For RPi2, set the target device to ‘Remote Machine’ and target architecture to ‘ARM’

    *For MBM, set the target device to ‘Remote Machine’ and target architecture to ‘x86’

  7. In VS Instance B, set the target architecture to ‘x86’. This will be the instance of the sample we run on the PC.

  8. In VS Instance A, press F5 to deploy and launch the app on your RPi2 or MBM.

  9. In VS Instance B, press F5 to deploy and launch the app on your PC.

Using the SerialSample App

When the SerialSample app is launched on the PC, a window will open with the user interface similar to the screenshot shown below. When launched on the RPi2 and MBM, the SerialSample will display the user interface shown below on the entire screen.

Selecting a Serial Device

When the SerialSample app launches, it looks for all the serial devices that are connected to the device. The device ids of all the serial devices found connected to the device will be listed in the top ListBox of the SerialSample app window.

Select and connect to a serial device on the PC and RPi2 or MBM by doing the following:

  1. Select the desired serial device by clicking on the device ID string in the top ListBox next to “Select Device:”.

    • On the PC, the device ID for the USB-to-TTL cable connected in this example begins with ‘\?\USB#VID_067B’.

    • On the MBM, if using the GPIO for serial communication, select the device ID with UART2 in it. UART1 may require using CTS/RTS signals.

    • On the DragonBoard, select the device with QCOM24D4 and UART1 in it. This will likely be the last device in the listbox (you may need to scroll down).

    • On the MBM and RPi2, if using the USB-to-TTL adapter module, select the device ID that begins with \?\USB#. For the USB-to-TTL module used in this example, the device ID should begin with ‘\?\USB#VID_10C4’.

  2. Click ‘Connect’.

The app will attempt to connect and configure the selected serial device. When the app has successfully connected to the attached serial device it will display the configuration of the serial device. By default, the app configures the serial device for 9600 Baud, eight data bits, no parity bits and one stop bit (no handshaking).

Sending and Receiving Data

After connecting the desired serial device in the SerialSample apps running on both the PC and the RPi2 or MBM we can begin sending and receiving data over the serial connection between the two devices.

To send data from one device to the other connected device do the following:

  1. Choose a device to transmit from. On the transmit device, type the message to be sent in the “Write Data” text box. For our example, we typed “Hello World!” in the “Write Data” text box of the SerialSample app running on our PC.

  2. Click the ‘Write’ button.

The app on the transmitting device will display the sent message and “bytes written successfully!” in the status text box in the bottom of the app display.

The device that is receiving the message will automatically display the text in the ‘Read Data:’ window.


  • When connecting to the USB-to-TTL device for the first time from the IoT Device, you may see the error “Object was not instantiated” when you click on Connect. If you see this, un-plug the device, plug it back in and refresh the connection or redeploy the app.
  • If you have more than one Silicon Labs USB-to-TTL devices connected to your IoT device, only the device that was first connected will be recognized. In order to run this sample, connect only one device
  • When connecting USB-to-TTL device to MinnowBoard Max, use a powered USB hub or the bottom USB port

Let’s look at the code

The code for this sample uses the Windows.Devices.SerialCommunication namespace.

The SerialDevice class will be used to enumerate, connect, read, and write to the serial devices connected to the device.

NOTE: The SerialDevice class can be used only for supported USB-to-TTL devices (on PC, Raspberry Pi 2, and MinnowBoard Max) and the on-board UART (on MinnowBoard Max).

For accessing the serial port, you must add the DeviceCapability to the Package.appxmanifest file in your project.

You can add this by opening the Package.appxmanifest file in an XML editor (Right Click on the file -> Open with -> XML (Text) Editor) and adding the capabilities as shown below:

Visual Studio 2015 has a known bug in the Manifest Designer (the visual editor for appxmanifest files) that affects the serialcommunication capability.  If 
your appxmanifest adds the serialcommunication capability, modifying your appxmanifest with the designer will corrupt your appxmanifest (the Device xml child 
will be lost).  You can workaround this problem by hand editting the appxmanifest by right-clicking your appxmanifest and selecting View Code from the 
context menu.
    <DeviceCapability Name="serialcommunication">
      <Device Id="any">
        <Function Type="name:serialPort" />

Connect to selected serial device

This sample app enumerates all serial devices connected to the device and displays the list in the ListBox ConnectDevices. The following code connects and configure the selected device ID and creates a SerialDevice object.

private async void comPortInput_Click(object sender, RoutedEventArgs e)
    var selection = ConnectDevices.SelectedItems; // Get selected items from ListBox

    // ...

    DeviceInformation entry = (DeviceInformation)selection[0];         

        serialPort = await SerialDevice.FromIdAsync(entry.Id);

        // ...

        // Configure serial settings
        serialPort.WriteTimeout = TimeSpan.FromMilliseconds(1000);
        serialPort.ReadTimeout = TimeSpan.FromMilliseconds(1000);                
        serialPort.BaudRate = 9600;
        serialPort.Parity = SerialParity.None;
        serialPort.StopBits = SerialStopBitCount.One;
        serialPort.DataBits = 8;

        // ...
    catch (Exception ex)
        // ...

Perform a read on the serial port

Reading input from serial port is done by Listen() invoked right after initialization of the serial port. We do this in the sample code by creating an async read task using the DataReader object that waits on the InputStream of the SerialDevice object.

Due to differences in handling concurrent tasks, the implementations of Listen() in C# and C++ differ:

  • C# allows awaiting ReadAsync(). All we do is keep reading the serial port in an infinite loop interrupted only when an exception is thrown (triggered by the cancellation token).
private async void Listen()
        if (serialPort != null)
            dataReaderObject = new DataReader(serialPort.InputStream);

            // keep reading the serial input
            while (true)
                await ReadAsync(ReadCancellationTokenSource.Token);
    catch (Exception ex)

private async Task ReadAsync(CancellationToken cancellationToken)
    Task<UInt32> loadAsyncTask;

    uint ReadBufferLength = 1024;

    // If task cancellation was requested, comply

    // Set InputStreamOptions to complete the asynchronous read operation when one or more bytes is available
    dataReaderObject.InputStreamOptions = InputStreamOptions.Partial;

    // Create a task object to wait for data on the serialPort.InputStream
    loadAsyncTask = dataReaderObject.LoadAsync(ReadBufferLength).AsTask(cancellationToken);

    // Launch the task and wait
    UInt32 bytesRead = await loadAsyncTask;
    if (bytesRead > 0)
        rcvdText.Text = dataReaderObject.ReadString(bytesRead);
        status.Text = "bytes read successfully!";
  • C++ does not allow awaiting ReadAsync() in Windows Runtime STA (Single Threaded Apartment) threads due to blocking the UI. In order to chain continuation reads from the serial port, we dynamically generate repeating tasks via “recursive” task creation - “recursively” call Listen() at the end of the continuation chain. The “recursive” call is not a true recursion. It will not accumulate stack since every recursive is made in a new task.
void MainPage::Listen()
        if (_serialPort != nullptr)
            // calling task.wait() is not allowed in Windows Runtime STA (Single Threaded Apartment) threads due to blocking the UI.
    catch (Platform::Exception ^ex)

Concurrency::task<void> MainPage::ReadAsync(Concurrency::cancellation_token cancellationToken)
    unsigned int _readBufferLength = 1024;
    return concurrency::create_task(_dataReaderObject->LoadAsync(_readBufferLength), cancellationToken).then([this](unsigned int bytesRead)
        if (bytesRead > 0)
            rcvdText->Text = _dataReaderObject->ReadString(bytesRead);
            status->Text = "bytes read successfully!";

            Dynamically generate repeating tasks via "recursive" task creation - "recursively" call Listen() at the end of the continuation chain.
            The "recursive" call is not true recursion. It will not accumulate stack since every recursive is made in a new task.

            // start listening again after done with this chunk of incoming data

Perform a write to the serial port

When the bytes are ready to be sent, we write asynchronously to the OutputStream of the SerialDevice object using the DataWriter object.

private async void sendTextButton_Click(object sender, RoutedEventArgs e)
    // ...
    // Create the DataWriter object and attach to OutputStream   
    dataWriteObject = new DataWriter(serialPort.OutputStream);

    //Launch the WriteAsync task to perform the write
    await WriteAsync();   
    // ..

    dataWriteObject = null;	

private async Task WriteAsync()
    Task<UInt32> storeAsyncTask;

    // ...
    // Load the text from the sendText input text box to the dataWriter object

    // Launch an async task to complete the write operation
    storeAsyncTask = dataWriteObject.StoreAsync().AsTask();

    // ...    

Cancelling Read

You can cancel the read operation by using CancellationToken on the Task. Initialize the CancellationToken object and pass that as an argument to the read task.

private async void comPortInput_Click(object sender, RoutedEventArgs e)
    // ...

    // Create cancellation token object to close I/O operations when closing the device
    ReadCancellationTokenSource = new CancellationTokenSource();
    // ...	

private async void rcvdText_TextChanged(object sender, TextChangedEventArgs e)
    // ...

    await ReadAsync(ReadCancellationTokenSource.Token);

    // ...	

private async Task ReadAsync(CancellationToken cancellationToken)
    Task<UInt32> loadAsyncTask;

    uint ReadBufferLength = 1024;

    // ...
private void CancelReadTask()
    if (ReadCancellationTokenSource != null)
        if (!ReadCancellationTokenSource.IsCancellationRequested)

Closing the device

When closing the connection with the device, we cancel all pending I/O operations and safely dispose of all the objects.

In this sample, we proceed to also refresh the list of devices connected.

private void closeDevice_Click(object sender, RoutedEventArgs e)
        ListAvailablePorts(); //Refresh the list of available devices
    catch (Exception ex)
       // ...

private void CloseDevice()
    if (serialPort != null)

    // ...

To summarize:

  • First, we enumerate all the serial devices connected and allow the user to connect to the desired one using device ID

  • We create an asynchronous task for reading the InputStream of the SerialDevice object.

  • When the user provides input, we write the bytes to the OutputStream of the SerialDevice object.

  • We add the ability to cancel the read task using the CancellationToken.

  • Finally, we close the device connection and clean up when done.