NEO6MV2 GPS receiver<\/h2>\n\n\n\n
The manufacturer of the GPS module is u-blox<\/strong>. The NEO-6 receivers are a whole family of chips. The exact one I\u2019m dealing with today is the NEO-6-M-0-001, and the module it\u2019s mounted on is called NEO6MV2<\/a>. <\/p>\n\n\n <\/p>\n\n\n\n The documentation lists a few interesting details about the receiver itself. The most important are:<\/p>\n\n\n\n The complete module has a status LED, EEPROM to store configuration, and a backup battery that helps with warm\/hot starts.<\/p>\n\n\n\n Only the UART interface is brought out to the NEO6MV2<\/a> module header. It is the primary form of communication. The module also has a \u00b5FL connector for attaching an antenna. You can buy the module with an antenna included. It\u2019s usually a small antenna that works pretty well outdoors, but is definitely worse indoors. I had to wait a long time with the antenna stuck to the window.<\/p>\n\n\n\n After powering up, the device immediately transmits NMEA messages on the RX pin. What is this enigmatic acronym?<\/p>\n\n\n\n NMEA (National Marine Electronics Association) is a communication protocol between marine electronic devices. Interesting, right? However, it is the standard when it comes to GPS devices. <\/p>\n\n\n\n Data comes as ASCII-encoded messages. They are therefore somewhat human-readable. A single message can contain a maximum of 82 characters. Each sentence begins with a dollar sign ($), then there is the sentence identifier, followed by subsequent data separated by commas. The sentence ends with the character sequence \u201c\\n\\r\u201d.<\/p>\n\n\n\n For the GPS system all messages start with \u201c$GP\u201d. Dozens of message types are defined, but for GPS I found 19 useful ones, which are described here: http:\/\/aprs.gids.nl\/nmea\/<\/a><\/p>\n\n\n\n In the default configuration, the NEO-6 outputs only a few of them:<\/p>\n\n\n\n Some data is repeated across several messages. Of course, you can edit what the receiver outputs on the UART interface. You can use the UBX protocol for this, but it\u2019s easier to connect the module to a computer and use the u-center<\/strong> application. It\u2019s a powerful tool that will, of course, read data from the NEO-6 module<\/a>, but also configure it freely, and even update the module\u2019s firmware. To connect a module that only has UART, you can use a UART<->USB converter<\/a>, which you can buy in my store<\/a>.<\/p>\n\n\n\n Today I\u2019ll use the Nucleo F401RE<\/a> board. For firmware configuration and programming I used STM32CubeIDE version 1.0.2 and the HAL libraries F4 1.24.1.<\/p>\n\n\n\n Allow me, perhaps for the first time on the blog, to skip the schematic. Connecting the module is really simple and boils down to powering it and wiring the TX and RX pins to the Nucleo. I powered the GPS board from 5V, and connected communication to UART1:<\/p>\n\n\n\n I also brought out a test pin for timing measurements \ud83d\ude42<\/p>\n\n\n\n Now Cube. In addition to the standard Nucleo settings such as HCLK = 84 MHz, set UART1 to 9600 baud 8n1<\/strong>. Also enable the global interrupt for UART1<\/strong> in the NVIC Settings tab.<\/p>\n\n\n\n You can now generate the project \ud83d\ude42 After generation, it\u2019s very important to add -u_printf_floa<\/em>t to the linker options because we\u2019re going to use it. <\/p>\n\n\n <\/p>\n\n\n\n I haven\u2019t written anything based on UART reception for a long time. But I managed hehe \ud83d\ude42 This time I opted for receiving single characters in an interrupt so as not to blockingly wait for incoming data. Step by step.<\/p>\n\n\n\n First, we should initialize the library. I created a struct variable to store all the data related to our module. Thanks to this structure, you can connect several independent NEO-6 modules<\/a> because it contains a pointer to the UART and its own receive and working buffer for it.<\/p>\n\n\n\n Initialization therefore fills the structure with initial data, assigns the UART for the receiver, and starts listening on the serial interface.<\/p>\n\n\n\n Let me discuss how the STM32 acquires all the data from the GPS receiver. As you remember, the NEO-6<\/a> sends data by itself, once per second by default. The UART interrupt is set for every single received byte. Why? We\u2019re never sure how many characters will arrive. Besides, this way it\u2019s easier to control the moment when the end of the message arrives by recognizing the \u2018\\n\u2019 character.<\/p>\n\n\n\n Receiving a character is handled in the NEO6_ReceiveUartChar<\/em> function, which should be placed in the UART receive-complete interrupt.<\/p>\n\n\n\n Let\u2019s look inside this function. It operates on the receive buffer from the GPS data structure. This is a circular\/ring buffer with no beginning or end. New data is appended at the next position in the buffer, and consumed data is removed from the \u201cstart\u201d. Meanwhile, that start and end keep moving depending on what is being done on the buffer.<\/p>\n\n\n\n As overflow handling, I chose to drop incoming data. Fortunately, by setting a sufficiently large buffer, and as long as the microcontroller isn\u2019t blocked on some task, such choking shouldn\u2019t occur. It\u2019s good to set the buffer to a value of 2x the longest message. As we know from the NMEA standard, it will be 81 characters at most, so the minimum buffer is 162 bytes. I set 256 because it\u2019s such a nice, round number.<\/p>\n\n\n\n Okay. What happens when there is space? Three conditions are checked:<\/p>\n\n\n\n At the very end I re-enable UART RX listening for one character.<\/p>\n\n\n\n There is a function called <\/p>\n\n\n\n which you must place in the program\u2019s main loop. It checks whether a complete message has already arrived. If so, it fetches the whole thing from the ring buffer into the working buffer and analyzes (parses) the message. How can such a message be parsed?<\/p>\n\n\n\n Two (three) functions are very useful for this:<\/p>\n\n\n\n Notice one characteristic thing in NMEA. Data is always separated by a comma and each piece of information has its defined place. An example of what the module prints after getting a fix:<\/p>\n\n\n <\/p>\n\n\n\n There\u2019s also a small piece after the asterisk which is a checksum. I\u2019ll skip it during parsing. I trust I\u2019m receiving correct data.<\/p>\n\n\n\n It would be good to leverage those commas. For this, the strtok<\/em> function is perfect. <\/p>\n\n\n\n How do you use it? It takes as arguments a pointer to the string to be split and the token<\/em> by which it will split. Our token is a comma.<\/p>\n\n\n\n What does it do with that comma? It inserts 0x00 in its place, i.e. the end-of-C-string character, and returns a pointer to this modified string. Now you can easily operate on it.<\/p>\n\n\n\n Alright, but did the rest of the buffer just disappear or what? What about the characters after the comma?!<\/p>\n\n\n\n Well, the next call to strtok<\/em> will return a pointer to the next string, i.e. the one after the last found token, meaning the characters between the first and second comma. It\u2019s important that you don\u2019t pass a pointer to the next string as the first argument now. To continue splitting the same buffer, you should now pass NULL as the first argument.<\/p>\n\n\n\n An example straight from the code. To extract the NMEA message header I use strtok<\/em>. Let\u2019s say the message looks like this:<\/p>\n\n\n\n $GPRMC,081836,A,3751.65,S,14507.36,E,000.0,360.0,130998,011.3,E*62<\/p>\n\n\n\n I passed a pointer to the working buffer which contains the full NMEA message. As the token, of course, \u2018,\u2019. Under ParsePointer<\/strong> we now have the string \u201c$GPRMC\u201d. <\/p>\n\n\n\n Next call<\/p>\n\n\n\n Will result in ParsePointer<\/strong> now being the string \u201c081836\u201d, representing the time. We can now convert it to a number and assign it to appropriate variables.<\/p>\n\n\n\n Let\u2019s call strtok<\/em> once again.<\/p>\n\n\n\n Now ParsePointer<\/strong> is \u201cA\u201d. We keep doing this until we obtain all the data, operating on those split strings with functions that convert ASCII to float, for example. And that\u2019s the whole magic of parsing this way. I hope you get it.<\/p>\n\n\n\n There\u2019s no rose without thorns. A big thorn is this super function strtok<\/em> itself. Why? Notice that some data fields are empty, i.e. there are two commas in a row. Especially if there is no fix. Why is this dangerous?<\/p>\n\n\n\n Well, the strtok<\/em> function does not return empty strings.<\/strong> Instead, it returns the next string with some content. As a result, if some data is missing in the sequence of tokens, our parsing gets broken\u2026 You can write some smarter parsing logic, but\u2026 you can also write a better token function. One that will return an empty string.<\/p>\n\n\n\n I found an implementation of such a function on the site loved by programmers with problems<\/a>.<\/p>\n\n\n\n The function is called strtoke<\/strong><\/em> and works exactly like standard strok<\/em> except that it can return an empty string. You can safely replace all calls with this improved one. Now empty fields won\u2019t be skipped. Success!<\/p>\n\n\n\n The sample code I wrote prints to UART2 the data it received from the GPS on UART1. However, it doesn\u2019t always print\u2014only when the module has a fix.<\/p>\n\n\n Let\u2019s also look at the logic analyzer to see how the interrupts and data parsing are doing.<\/p>\n\n\n\n Receiving all messages from the NEO-6<\/a> looks like this.<\/p>\n\n\n The entire frame the module sends is over 100 ms. This one doesn\u2019t have a fix. It won\u2019t lock on my desk\u2026 Adding data after getting a GPS signal, this frame will be a bit longer.<\/p>\n\n\n\n As you can see in the second row there are lots of interrupt calls for character reception. They are triggered often so they look serious. How much do they knock the microcontroller out of its normal operation?<\/p>\n\n\n <\/p>\n\n\n\n The interrupt lasts about 2.8 \u00b5s. Considering that this time also includes handling the test GPIO pin, that\u2019s very little. Such an interrupt is invoked every exactly 1.04 ms when the next character arrives.<\/p>\n\n\n\n On an even lower level you can see a few spikes. I marked the moments of data parsing here. <\/p>\n\n\n Parsing, i.e. splitting, searching strings, and converting data takes about 100 \u00b5s.<\/p>\n\n\n\n Handling GPS is very easy. Just a few tricks are enough to fully enjoy the messages transmitted by the NEO-6 module<\/a>. Reception and data processing are very fast. Also remember that the times I presented relate to an HCLK frequency of 84 MHz.<\/p>\n\n\n\n Where can you use such a GPS? Lately I see an application for it in a measurement device with LoRa communication. We kind of know where the device is, but when producing several of them, we don\u2019t have to worry about manually entering the location. A cool solution was recently suggested to me by one of the readers. It was a measurement drone. It performs air quality measurements and automatically saves them to an SD card together with the measurement location. Great use case.<\/p>\n\n\n\n An additional advantage of using GPS is the accurate time. You can use it to synchronize the RTC or even consider removing the clock from the device altogether.<\/p>\n\n\n\n GPS receivers are everywhere these days. For example in smartphones or cars. Even drones have GPS, which, for example, assists in piloting the machine.<\/p>\n\n\n\n You can find the full project along with the library on my GitHub as usual: link<\/a><\/p>\n\n\n\n If you noticed any mistake, disagree with something, would like to add something important or simply feel like discussing the topic, leave a comment. Remember that the discussion should be polite and in accordance with the rules of the Polish language.<\/span><\/p>\n\n\n![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/NEP6MV2_baner.jpg\")
<\/a><\/figure>\n<\/div>\n\n\n\n
NMEA<\/h3>\n\n\n\n
\n
Connections and Cube configuration<\/h2>\n\n\n\n
![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2020\/07\/Nucleo-64-baner.jpg\")
<\/a><\/figure><\/h1>\n\n\n\n\n
![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_linker-288x300.jpg\")
<\/a><\/figure>\n<\/div>\n\n\nCode<\/h2>\n\n\n\n
typedef struct\n{\n\t\/\/\n\t\/\/ UART stuff\n\t\/\/\n\tUART_HandleTypeDef *neo6_huart;\n\tuint8_t UartBuffer[GPS_UART_BUFFER_SIZE];\n\tuint8_t UartBufferHead;\n\tuint8_t UartBufferTail;\n\tuint8_t UartBufferLines;\n\n\tuint8_t WorkingBuffer[GPS_WORKING_BUFFER_SIZE];\n\n\t\/\/\n\t\/\/ Time and Date\n\t\/\/\n\tuint8_t Hour;\n\tuint8_t Minute;\n\tuint8_t Second;\n\tuint8_t Day;\n\tuint8_t Month;\n\tuint8_t Year;\n\n\t\/\/\n\t\/\/ Position\n\t\/\/\n\tdouble Latitude;\n\tchar LatitudeDirection;\n\tdouble Longitude;\n\tchar LongitudeDirection;\n\tdouble Altitude;\n\n\t\/\/\n\t\/\/ Speed\n\t\/\/\n\tdouble SpeedKnots;\n\tdouble SpeedKilometers;\n\n\t\/\/\n\t\/\/ Satelites parameters\n\t\/\/\n\tuint8_t SatelitesNumber;\n\tuint8_t Quality; \/\/ 0 - no Fix, 1 - Fix, 2 - Dif. Fix\n\tuint8_t FixMode; \/\/ 1 - no Fiz, 2 - 2D, 3 - 3D\n\tdouble Dop; \/\/ Dilution of precision\n\tdouble Hdop; \/\/ Dilution of precision for flat coords\n\tdouble Vdop; \/\/ Dilution of precision for height\n\n}NEO6_State;<\/pre>\n\n\n\nvoid NEO6_Init(NEO6_State *GpsStateHandler, UART_HandleTypeDef *huart);<\/pre>\n\n\n\n
\/* USER CODE BEGIN 4 *\/\nvoid HAL_UART_RxCpltCallback(UART_HandleTypeDef *huart)\n{\n if(huart == GpsState.neo6_huart)\n {\n NEO6_ReceiveUartChar(&GpsState);\n }\n}\n\/* USER CODE END 4 *\/<\/pre>\n\n\n\n\n
void NEO6_Task(NEO6_State *GpsStateHandler);<\/pre>\n\n\n\n
Parsing NMEA<\/h3>\n\n\n\n
\n
![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_raw_data.jpg\")
<\/a><\/figure>\n<\/div>\n\n\nchar *strtok(char *s, const char *delim);<\/pre>\n\n\n\n
char* ParsePoiner = strtok((char*)GpsStateHandler->WorkingBuffer, \",\");<\/pre>\n\n\n\n
ParsePoiner = strtok(NULL, \",\");<\/pre>\n\n\n\n
ParsePoiner = strtok(NULL, \",\");<\/pre>\n\n\n\n
The strtok trap<\/h2>\n\n\n\n
Result<\/h2>\n\n\n\n
![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_no_fix.jpg\")
<\/a><\/figure>\n<\/div>\n\n![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_fix_data.jpg\")
<\/a><\/figure>\n<\/div>\n\n\n![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_receiving_data-1024x115.jpg\")
<\/a><\/figure>\n<\/div>\n\n\n![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_interrupt_time-1024x113.jpg\")
<\/a><\/figure>\n<\/div>\n\n\n![\"\"](\"http:\/\/msalamon.pl\/wp-content\/uploads\/2019\/09\/neo6mv2_parse_time-1024x119.jpg\")
<\/a><\/figure>\n<\/div>\n\n\nSummary<\/h2>\n\n\n\n