After getting your kit and finding the necessary components, but before assembling your kit, you should take the time to do a mockup build and setup your Raspberry.
In order to make it easy to connect to the PlanktoScop, you may want to install [avahi](https://en.wikipedia.org/wiki/Avahi_%28software%29) or the [Bonjour](https://en.wikipedia.org/wiki/Bonjour_%28software%29) service on any computer you will use to access the PlanktoScop interface. This will allow you to connect to the PlantoScop using an address similar such as http://planktoscope.local instead of an IP address.
The first boot to the desktop may take up to 120 seconds. This is normal and is caused by the image expanding the filesystem to the whole SD card. DO NOT REBOOT before you reach the desktop.
While you're here, a wise thing to do would be to change the default password for the `pi` user. This is very warmly recommended if your PlanktoScop is connected to a shared network you do not control. Just select the first option `1 Change User Password`.
You may also want to change the default hostname of your Raspberry. To do so, choose option `2 Network Options` then `N1 Hostname`. Choose a new hostname. We recommend using `planktoscope`.
We need to activate a few things for the PlanktoScop to work properly.
First, we need to activate the camera interface. Choose `5 Interfacing Options`, then `P1 Camera` and `Yes`.
These steps can also be done from the Raspberry Pi Configuration GUI tool that you can find in `Main Menu > Preferences`. Go to the `Interfaces` tab. Pay attention, here the Serial Port must be enabled, but the Serial Port Console must be disabled.
You can then run the following to make sure your Raspberry has the necessary components to install and build everything it needs and to create the necessary folders:
Start by following [Adafruit's guide](https://learn.adafruit.com/circuitpython-on-raspberrypi-linux/installing-circuitpython-on-raspberry-pi). You can start at the chapter `Install Python Libraries`.
It is recommended to test this setup by creating this small script under the name `test/blinkatest.py` and running it (you can use the editor nano if you are using the terminal).
The device appearing at addresses 60 and 70 is our motor controller. Address `0d` is the fan controller and `3c` is the oled screen (we'll set up both a bit further down). Your version of the RGB Cooling Hat may not have the screen, it's fine as the screen is not necessary for proper operation of the Planktoscope.
In case the motor controller does not appear, shutdown your Planktoscope and check the wiring. If your board is using a connector instead of a soldered pin connection (as happens with the Adafruit Bonnet Motor Controller), sometimes the pins on the male side need to be bent a little to make good contact. In any case, do not hesitate to ask for help in Slack.
We are going to use gpsd to parse the GPS data. We need to set it up by editing `/etc/default/gpsd`. This file is source just before starting gpsd and allows to configure its working.
```sh
sudo nano /etc/default/gpsd
```
Change the `USB_AUTO` line to read `false`
```sh
USBAUTO="false"
```
Also change the `DEVICES` line to add the device we are going to use `/dev/serial0`:
```sh
DEVICES="/dev/serial0"
```
Finally, we want to add the parameter `-n` to `GPSD_OPTIONS`:
```sh
GPSD_OPTIONS="-n"
```
Save your work, and restart gpsd by running the following:
```sh
sudo systemctl restart gpsd.service
```
If you wait a bit, you can run `gpsmon` to check that your configuration is correct. You should get an output similar to this:
#### Bonus Configuration: Automatic time update from GPSD
The Adafruit GPS HAT allows your PlanktoScop to automatically sets its time to the GPS received one. Moreover, since the PPS (Pulse Per Second) output is activated, you can even set your PlanktoScope to act as a stratum 1 timeserver.
We are first going to make sure that your PlanktoScope receives proper PPS signal. We need to add the following line at the end of `/boot/config.txt`:
```
sudo nano /boot/config.txt
# Add the following line at the end of the line
dtoverlay=pps-gpio,gpiopin=4
```
We also need to activate the pps module of the kernel, by editing `/etc/modules`:
```
sudo nano /etc/modules
# Add the following line at the end of the line
pps-gpio
```
Now install `pps-tools` so we can check that this is properly running.
```sh
sudo apt install pps-tools
```
Finally, in the `/etc/default/gpsd` file, we need to add our pps device to the line `DEVICES`. Append `/dev/pps0`:
```sh
DEVICES="/dev/serial0 /dev/pps0"
```
Reboot your PlanktoScope now and check that the PPS signal is properly parsed by the PlanktoScope. You can do this by running `sudo ppstest /dev/pps0`:
`gpsmon` should also show a PPS signal in the `GSA + PPS` box.
We now need to install the software that will act as timeserver, both locally and globally. Its name is Chrony. It's a more modern replacement for `ntp`, using the same underlying protocol. Let's go ahead and install it:
```sh
sudo apt install chrony
```
We need to edit the configuration of chrony, to activate both the GPS time synchronization and to allow clients to request time updates directly from our microscope.
Edit the file `/etc/chrony/chrony.conf` and replace its content with the following:
Before restarting `chrony`, we need to make sure the timesync service from systemd is deactivated:
```sh
sudo systemctl stop systemd-timesyncd.service
sudo systemctl disable systemd-timesyncd.service
```
Final step, let's start `chrony` with its new configuration and restart `gpsd`:
```sh
sudo systemctl restart chrony
sudo systemctl restart gpsd
```
To check that everything is working as intended, wait a few minutes, and then type `chronyc sources -v`. This command will show the time sources `chrony` is using, and right at the top there should be our NMEA source. Make sure its line starts with `#*`, which means this source is selected:
The other servers are here just as fallback measures, in case the GPS is not working for an unknown reason.
This part is now complete! Everytime you start your Planktoscope, it will set its own time after a few minutes (once a GPS signal is acquired).
The ultimate step will have to be done on the other equipment on the network where you want to use this time source. You will need to add the line `server planktoscope.local` to your ntp configuration file either at `/etc/ntp.conf` or at `/etc/chrony/chrony.conf` and then restart your ntp service.
You can find more information in this hardware module in Adafruit documentation at [Installing Adafruit GPS HAT](https://learn.adafruit.com/adafruit-ultimate-gps-hat-for-raspberry-pi/overview) or on this page to [use Python Thread with GPS HAT](http://www.danmandle.com/blog/getting-gpsd-to-work-with-python/)
### Install RGB Cooling HAT
To setup the RGB Cooling HAT, you just need to clone and build the WiringPi library:
To install Node.js, npm and Node-RED onto a Raspberry Pi, you just need to run the following command. You can review the content of this script [here](https://raw.githubusercontent.com/node-red/linux-installers/master/deb/update-nodejs-and-nodered).
Make sure Node-RED is correctly installed by reaching the following page from the browser of your pi http://localhost:1880 or http://planktoscope.local:1880 from another computer on the same network.
From Node-RED gui in your browser, choose the Hamburger menu top right, and then Import. You can paste the code directly from the lastest version of the GUI available [here](https://raw.githubusercontent.com/PlanktonPlanet/PlanktonScope/blob/master/flows/main.json).
We need to install the latest OpenCV version. Unfortunately, it is not available in the repositories. We are going to install it directly by using pip.
First, we need to install the needed dependencies, then we will directly install opencv:
You can now check that opencv is properly installed by running a python interpreter and importing the cv2 module.
```sh
pi@planktoscope:~ $ python3
Python 3.7.3 (default, Dec 20 2019, 18:57:59)
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import cv2
>>> cv2.__version__
'4.1.0'
>>> quit()
```
If all goes well, the displayed version number should be `4.1.0`.
More detailed information can be found on this [website](https://www.pyimagesearch.com/2019/09/16/install-opencv-4-on-raspberry-pi-4-and-raspbian-buster/).
To update the interface and make sure you run the latest version, you need to copy the json config file from the cloned repository to the Node-RED library:
