Solder Reflow Oven Build

At first glance, it appeared that measuring a negative temperature would be a bit more difficult because of the way in which the MAX31855 handled these values.  Basically, the chip works backwards for negative, so a temperature of -0.25°C would be represented as 1111 1111 1111 11 and a temperature of -0.50°C would be represented as 1111 1111 1111 10 and so on.  Since this is basically the opposite of the positive temperatures, I found that if I just XOR'ed the result with 1111 1111 1111 11 (0x3FFF)and added 1, I could use the same method as I used for the positive results.  Here's a piece of my test code used to obtain positive or negative temperatures in Celsius:

   extern uint8_t volatile temp[4];  \\ temp[0] contains the top 8 bits and temp[1] contains the bottom 6 bits plus two unused bits which will be shifted out

   uint16_t volatile temp_result;

// Move the binary temperature result into temp_result using only the lower 14 bits

  temp_result = temp[0];

  temp_result = temp_result << 8;

  temp_result |= temp[1];

  temp_result = temp_result >> 2;

// The if statement checks the 14th bit.  Assuming there is no fault condition, this bit will be 1 if the result is negative.

   if ( (temp_result & 0x2000) == 0){

       sprintf(buffer, "Result is: \n%d.%d%cC",temp_result >> 2, (temp_result & 0x03) * 25, 0x7F);

       Dogs102x6_stringDraw(1, 0, buffer, DOGS102x6_DRAW_NORMAL);

}

   else{

  temp_result = (temp_result ^ 0x3FFF) + 1;

       sprintf(buffer, "Result is: \n-%d.%d%cC",temp_result >> 2, (temp_result & 0x03) * 25, 0x7F);

       Dogs102x6_stringDraw(1, 0, buffer, DOGS102x6_DRAW_NORMAL);    

}

Additional code would need to be added to check for faults prior to making the temperature measurement. 

At first glance, it appeared that measuring a negative temperature would be a bit more difficult because of the way in which the MAX31855 handled these values.  Basically, the chip works backwards for negative, so a temperature of -0.25°C would be represented as 1111 1111 1111 11 and a temperature of -0.50°C would be represented as 1111 1111 1111 10 and so on.  Since this is basically the opposite of the positive temperatures, I found that if I just XOR'ed the result with 1111 1111 1111 11 (0x3FFF)and added 1, I could use the same method as I used for the positive results.  Here's a piece of my test code used to obtain positive or negative temperatures in Celsius:

   extern uint8_t volatile temp[4];  \\ temp[0] contains the top 8 bits and temp[1] contains the bottom 6 bits plus two unused bits which will be shifted out

   uint16_t volatile temp_result;

// Move the binary temperature result into temp_result using only the lower 14 bits

  temp_result = temp[0];

  temp_result = temp_result << 8;

  temp_result |= temp[1];

  temp_result = temp_result >> 2;

// The if statement checks the 14th bit.  Assuming there is no fault condition, this bit will be 1 if the result is negative.

   if ( (temp_result & 0x2000) == 0){

       sprintf(buffer, "Result is: \n%d.%d%cC",temp_result >> 2, (temp_result & 0x03) * 25, 0x7F);

       Dogs102x6_stringDraw(1, 0, buffer, DOGS102x6_DRAW_NORMAL);

}

   else{

  temp_result = (temp_result ^ 0x3FFF) + 1;

       sprintf(buffer, "Result is: \n-%d.%d%cC",temp_result >> 2, (temp_result & 0x03) * 25, 0x7F);

       Dogs102x6_stringDraw(1, 0, buffer, DOGS102x6_DRAW_NORMAL);    

}

Additional code would need to be added to check for faults prior to making the temperature measurement.