An RP2040 Harp Core that implements the Harp Protocol to serve as the basis of a custom Harp device.
- Synchronization to an external Harp Clock Synchronizer signal.
- Parsing incoming harp messages
- Dispatching messages to the appropriate register
- Sending harp-compliant timestamped replies
See the examples folder to get a feel for incorporating the harp core into your own project.
Additionally, here are a few examples that use the RP2040 Harp Core as a submodule in the wild:
- harp.device.environment-sensor
- harp.device.lickety-split
- harp.device.sniff-detector
- harp.device.treadmill
- harp.device.cuttlefish
- harp.pico.cam-trigger
- harp.pico.ephys-sync
The easiest way to use this library is to include it as submodule in your project. To see how to structure your project to incorporate the RP2040 Harp Core as a library, see the examples above--or read on.
Download (or clone) the Pico SDK to a known folder on your PC. (This folder does not need to be a sub-folder of your project.) From the Pico SDK project root folder, install the Pico SDK's dependencies with:
git submodule update --init --recursive
Next, in a sub-folder of your project, add harp.core.rp2040 as a submodule with:
git submodule add git@github.com:AllenNeuralDynamics/harp.core.rp2040.git
At the top of your project's CMakeLists.txt, you will need to include and initialize the Pico SDK. You can do so with:
include(${PICO_SDK_PATH}/pico_sdk_init.cmake)
pico_sdk_init()Your must also point to the folder of the Harp.Core.RP2040's CMakeLists.txt with
add_subdirectory(/path/to/cmakelist_dir build) # Path to harp.core.rp2040's CMakeLists.txt(Note that you must change path/to/cmakelist_dir above to the actual path of this project's CMakeLists.txt.)
At the linking step, you can link against the Harp core libraries with:
target_link_libraries(${PROJECT_NAME} harp_core harp_sync)Recommended, but optional: define the PICO_SDK_PATH environment variable to point to the location where the pico-sdk was downloaded. i.e:
PICO_SDK_PATH=/home/username/projects/pico-sdk
On Linux, it may be preferrable to put this in your .bashrc file.
From this directory, create a directory called build, enter it, and invoke cmake with:
mkdir build
cd build
cmake ..
If you did not define the PICO_SDK_PATH as an environment variable, you must pass it in here like so:
mkdir build
cd build
cmake -DPICO_SDK_PATH=/path/to/pico-sdk ..
After this point, you can invoke the auto-generated Makefile with make
Press-and-hold the Pico's BOOTSEL button and power it up (i.e: plug it into usb). At this point you do one of the following:
- drag-and-drop the created *.uf2 file into the mass storage device that appears on your pc.
- flash with picotool
Native packages exist in Bonsai for communicating with devices that speak Harp protocol. For more information on reading data or writing commands to your custom new harp device, see the Harp Tech Bonsai notes.
The Harp Core consumes the USB serial port, so printf messages must be rerouted to an available UART port.
The Pico SDK makes this step fairly straightforward. Before calling printf you must first setup a UART port with:
#define UART_TX_PIN (0)
#define BAUD_RATE (921600)
stdio_uart_init_full(uart0, BAUD_RATE, UART_TX_PIN, -1) // or uart1, depending on pinTo read these messages, you must connect a 3.3V FTDI Cable to the corresponding pin and connect to it with the matching baud rate.
Additionally, in your CMakeLists.txt, you must add
pico_enable_stdio_uart(${PROJECT_NAME} 1) # UART stdio for printf.for each library and executable using printf and you must link it with pico_stdlib.
printf messages are sprinkled throughout the Harp Core code, and they can be conditionally compiled by adding flags to your CMakeLists.txt.
To print out details of every outgoing (Device to PC) messages, add:
add_definitions(-DDEBUG_HARP_MSG_OUT)To print out details of every received (PC to Device) messages, add:
add_definitions(-DDEBUG_HARP_MSG_IN)- Harp Protocol Repo
- pyharp python library for connecting to harp-compliant devices and sending read/writes.