GPS example sketch stuck after a couple of iterations

[mehow]

Board: CubeCell GPS-6502 Version: 1.4.0 Example: OnBoardGPS > PeriodicLowPowerMode

The sketch gets stuck on "GPS Searching..." after a few iterations. More precisely it gets stuck in GPS.begin() while checking for the current baudrate:

23:49:00.892 -> GPS Current baudrate detected:9600
23:49:01.004 -> GPS baudrate updating to 9600
23:49:01.040 -> GPS baudrate updated to 9600
23:50:01.975 -> GPS search timeout.
23:50:04.002 -> 2 seconds later into lowpower mode....
23:51:06.118 -> GPS Searching...
23:51:06.118 -> GPS Current baudrate detecting...
23:51:06.336 -> GPS Current baudrate detected:9600
23:51:06.444 -> GPS baudrate updating to 9600
23:51:06.520 -> GPS baudrate updated to 9600
23:51:17.425 -> GPS search timeout.
23:51:19.456 -> 2 seconds later into lowpower mode....
23:52:21.566 -> GPS Searching...
23:52:21.566 -> GPS Current baudrate detecting...
23:52:21.774 -> GPS Current baudrate detected:9600
23:52:21.879 -> GPS baudrate updating to 9600
23:52:21.953 -> GPS baudrate updated to 9600
23:52:32.880 -> GPS search timeout.
23:52:34.901 -> 2 seconds later into lowpower mode....
23:53:37.016 -> GPS Searching...
23:53:37.016 -> GPS Current baudrate detecting...
23:53:37.232 -> GPS Current baudrate detected:9600
23:53:37.356 -> GPS baudrate updating to 9600
23:53:37.389 -> GPS baudrate updated to 9600
23:53:48.333 -> GPS search timeout.
23:53:50.365 -> 2 seconds later into lowpower mode....
23:54:52.474 -> GPS Searching...
23:54:52.474 -> GPS Current baudrate detecting...
23:54:52.702 -> GPS Current baudrate detected:9600
23:54:52.786 -> GPS baudrate updating to 9600
23:54:52.870 -> GPS baudrate updated to 9600
23:55:03.803 -> GPS search timeout.
23:55:05.813 -> 2 seconds later into lowpower mode....
23:56:07.910 -> GPS Searching...
23:56:07.910 -> GPS Current baudrate detecting...

(there is no more output after that line)

After adding some print statements I learned that the execution gets stuck inside the getNMEA() method if GPSSerial.available() continues to return false. This code:

String Air530ZClass::getNMEA()
{
	uint32_t starttime = millis();
	String nmea = "";
	char buff[128]; 
	int index = 0;
	while(millis() - starttime <1000)
	{
		Serial.print(".");
		if(GPSSerial.available())
		{
			char c = GPSSerial.read();
			Serial.println(c);
			if(c=='$')
			{
				nmea += c;
				index = GPSSerial.readBytesUntil('\r',buff,127);
				buff[index]=0;
				if(buff[index-3]!='*')
					return "0";

				nmea += (String)buff;
				return nmea;
			}
		}
		Serial.print("_");
	}
	Serial.println("XXX");
	return "0";
}

ends up printing

00:56:20.923 -> GPS Searching...
00:56:20.959 -> GPS Current baudrate detecting...
00:56:20.959 -> ._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._.

before getting stuck. Notice that it never exits the loop and instead appears crash or get stuck inside GPSSerial.available().

I checked and the problem is not present in 1.3.0 which makes me suspect this commit.

[rodri16]

Same problem here...

[olman011]

In case anyone is still having this problem, in PlatformIO the solution is to change the platform in the platformio.ini file. If you choose the Cubecell GPS board when starting a project it will default to "platform = asrmicro650x". This platform is deprecated and must be manually changed to "platform = heltec-cubecell". Once this is done the GPS functionality will work correctly.