Adding a fix to make sure the MS8607 is detected correctly.

Author: PaulZC

Firmware/OpenLog_Artemis/OpenLog_Artemis.ino

@@ -11,6 +11,8 @@
   This firmware runs the OpenLog Artemis. A large variety of system settings can be
   adjusted by connecting at 115200bps.
 
+  The Board should be set to SparkFun Apollo3 \ SparkFun RedBoard Artemis ATP.
+
   v1.0 Power Consumption:
    Sleep between reads, RTC fully charged, no Qwiic, SD, no USB, no Power LED: 260uA
    10Hz logging IMU, no Qwiic, SD, no USB, no Power LED: 9-27mA
@@ -69,10 +71,11 @@
   (done) Add support for the MPRLS0025PA micro pressure sensor
   (done) Add support for the SN-GCJA5 particle sensor
   (done) Add IMU accelerometer and gyro full scale and digital low pass filter settings to menuIMU
+  (done) Add a fix to make sure the MS8607 is detected correctly: https://github.com/sparkfun/OpenLog_Artemis/issues/54
 */
 
 const int FIRMWARE_VERSION_MAJOR = 1;
-const int FIRMWARE_VERSION_MINOR = 7;
+const int FIRMWARE_VERSION_MINOR = 8;
 
 //Define the OLA board identifier:
 //  This is an int which is unique to this variant of the OLA and which allows us
@@ -82,7 +85,7 @@ const int FIRMWARE_VERSION_MINOR = 7;
 //    the variant * 0x100 (OLA = 1; GNSS_LOGGER = 2; GEOPHONE_LOGGER = 3)
 //    the major firmware version * 0x10
 //    the minor firmware version
-#define OLA_IDENTIFIER 0x117 // Stored as 279 decimal in OLA_settings.txt
+#define OLA_IDENTIFIER 0x118 // Stored as 280 decimal in OLA_settings.txt
 
 #include "settings.h"
 

Firmware/OpenLog_Artemis/autoDetect.ino

@@ -1273,6 +1273,7 @@ deviceType_e testDevice(uint8_t i2cAddress, uint8_t muxAddress, uint8_t portNumb
 
 //Given an address, returns the device type if it responds as we would expect
 //This version is dedicated to testing muxes and uses a custom .begin to avoid the slippery mux problem
+//However, we also need to check if an MS8607 is attached (address 0x76) as it can cause the I2C bus to lock up if not detected correctly
 deviceType_e testMuxDevice(uint8_t i2cAddress, uint8_t muxAddress, uint8_t portNumber)
 {
   switch (i2cAddress)
@@ -1283,7 +1284,53 @@ deviceType_e testMuxDevice(uint8_t i2cAddress, uint8_t muxAddress, uint8_t portN
     case 0x73:
     case 0x74:
     case 0x75:
+      {
+        //Ignore devices we've already recorded. This was causing the mux to get tested, a begin() would happen, and the mux would be reset.
+        if (deviceExists(DEVICE_MULTIPLEXER, i2cAddress, muxAddress, portNumber) == true) return (DEVICE_MULTIPLEXER);
+        
+        //Confidence: Medium - Write/Read/Clear to 0x00
+        if (multiplexerBegin(i2cAddress, qwiic) == true) //Address, Wire port
+          return (DEVICE_MULTIPLEXER);
+      }
+      break;
     case 0x76:
+      {
+        //Ignore devices we've already recorded. This was causing the mux to get tested, a begin() would happen, and the mux would be reset.
+        if (deviceExists(DEVICE_MULTIPLEXER, i2cAddress, muxAddress, portNumber) == true) return (DEVICE_MULTIPLEXER);
+
+        // If an MS8607 is connected, multiplexerBegin causes the MS8607 to 'crash' and lock up the I2C bus... So we need to check if an MS8607 is connected first. 
+        // We will use the MS5637 as this will test for itself and the pressure sensor of the MS8607
+        // Just to make life even more complicated, a mux with address 0x76 will also appear as an MS5637 due to the way the MS5637 eeprom crc check is calculated.
+        // So, we can't use .begin as the test for a MS5637 / MS8607. We need to be more creative!
+        // If we write 0xA0 to i2cAddress and then read two bytes:
+        //  A mux will return 0xA0A0
+        //  An MS5637 / MS8607 will return the value stored in its eeprom which _hopefully_ is not 0xA0A0!
+
+        // Let's hope this doesn't cause problems for the BME280...! We should be OK as the default address for the BME280 is 0x77.
+        
+        qwiic.beginTransmission((uint8_t)i2cAddress);
+        qwiic.write((uint8_t)0xA0);
+        uint8_t i2c_status = qwiic.endTransmission();
+
+        if (i2c_status == 0) // If the I2C write was successful
+        {
+          qwiic.requestFrom((uint8_t)i2cAddress, 2U); // Read two bytes
+          uint8_t buffer[2];
+          for (uint8_t i = 0; i < 2; i++)
+          {
+            buffer[i] = qwiic.read();
+          }
+          if ((buffer[0] != 0xA0) || (buffer[1] != 0xA0)) // If we read back something other than 0xA0A0 then we are probably talking to an MS5637 / MS8607, not a mux
+          {
+            return (DEVICE_PRESSURE_MS5637);
+          }
+        }
+
+        //Confidence: Medium - Write/Read/Clear to 0x00
+        if (multiplexerBegin(i2cAddress, qwiic) == true) //Address, Wire port
+          return (DEVICE_MULTIPLEXER);
+      }
+      break;
     case 0x77:
       {
         //Ignore devices we've already recorded. This was causing the mux to get tested, a begin() would happen, and the mux would be reset.

Firmware/OpenLog_Artemis/menuAttachedDevices.ino

@@ -51,6 +51,7 @@ bool detectQwiicDevices()
 
   //First scan for Muxes. Valid addresses are 0x70 to 0x77 (112 to 119).
   //If any are found, they will be begin()'d causing their ports to turn off
+  //testMuxDevice will check if an MS8607 is attached (address 0x76) as it can cause the I2C bus to lock up if we try to detect it as a mux
   uint8_t muxCount = 0;
   for (uint8_t address = 0x70 ; address < 0x78 ; address++)
   {
@@ -73,6 +74,11 @@ bool detectQwiicDevices()
           Serial.printf("detectQwiicDevices: multiplexer found at address 0x%02X\r\n", address);
         muxCount++;
       }
+      else if (foundType == DEVICE_PRESSURE_MS5637)
+      {
+        if (settings.printDebugMessages == true)
+          Serial.printf("detectQwiicDevices: MS8607/MS5637 found at address 0x%02X. Ignoring it for now...\r\n", address);        
+      }
     }
   }