MLX90614 Sensor Displaying in Fahrenheit

jtm1202

No trouble communicating with the sensor and getting data in Kelvin and Celsius but I would like to display the temperature in Fahrenheit as well. I am close to the goal but there is something small I have overlooked and cannot put my finger on it. When I run the program I have modified I get a reading for Fahrenheit but it is not correct. Example: When celsius reads 20, fahrenheit reads 37.

What else do I need to modify?

Find file attached

Thanks!

ElectricAye

I'm not very familiar with the BS2, but it jumped out at me that 20 C is very close to 68 F. And 68-37= 31, which is very close to 32. Any chance you're subtracting a 32 somewhere in your conversion code where you shouldn't be?

jtm1202

Nice catch! I checked on that and I'm not subtracting 32 anywhere, just adding. Here is what i have

Init:
LOW Reset
INPUT Reset
DEBUG CLS, ' setup report screen
"MLX90614 Infra Red Thermometer", CR,
"=============================", CR,CR,
"Kelvin......... ", CR,
"Celsius........ ", CR,
"Fahrenheit..... ", CR,
"PEC test....... ", CR

'

---

[ Program Code ]

---

getTemperature:
SEROUT Sensor,baud,[0,"!TEMR",xslave,$07]
SERIN Sensor,baud,1000,PECfail,[tempL,tempH,pec]

checkPEC:
z=0
y = xslave<<1+0
GOSUB calculateCRC
y = $07
GOSUB calculateCRC
y = xslave<<1+1
GOSUB calculateCRC
y = tempL
GOSUB calculateCRC
y = tempH
GOSUB calculateCRC
IF z<>pec THEN PECfail
GOTO convertTemperatures

calculateCRC:
w=z^y
READ w,z
RETURN

convertTemperatures:
Kelvin = (temperature/100)*2
KelvinDec = temperature*2

IF (temperature < 13650) THEN overWordsize

Celsius = (temperature/50)-273
CelsiusDec = (temperature//50*2)

Fahrenheit = ((((temperature/50)-273) * 18) + 32) / 10
FahrenheitDec = (((temperature//50*2) * 18) + 32) / 10

GOTO displayTemperatures



overWordsize:

Celsius = (273-(temperature/50) )
CelsiusDec = (temperature//50*2)
DEBUG CRSRXY, 23, 4,"-",DEC Celsius,".",DEC2 CelsiusDec, CLREOL

displayTemperatures:
DEBUG CRSRXY, 23, 3,DEC Kelvin,".",DEC1 KelvinDec,CLREOL
DEBUG CRSRXY, 23, 4,DEC Celsius,".",DEC2 CelsiusDec, CLREOL
DEBUG CRSRXY, 23, 5,DEC Fahrenheit,".",DEC3 FahrenheitDec, CLREOL
DEBUG CRSRXY, 23, 6,"Pass",CLREOL
PAUSE 1000
GOTO getTemperature

PECfail:
DEBUG CRSRXY, 23, 3,CLREOL,CR
DEBUG CRSRXY, 23, 4,CLREOL,CR
DEBUG CRSRXY, 23, 5, CLREOL,CR
DEBUG CRSRXY, 23, 6,"Fail",CLREOL,CR
PAUSE 500
GOTO getTemperature

Tracy Allen

Your Fahrenheit formula needs to add 320, not just 32.

[SIZE=1]Fahrenheit = ((((temperature/50)-273) * 18) + 32[/SIZE][SIZE=1][COLOR=red]0[/COLOR][/SIZE][SIZE=1]) / 10[/SIZE]

I think the other formulas and debugs are a little off, too. For example, it appears that the raw temperature value is returned in units of 0.02 Kelvin. Here is how I would write the conversions and the debugs...

[SIZE=1]Kelvin = temperature * 2   ' in units of 0.01 K, resolution 0.02 K
DEBUG CR, "Kevin=", DEC Kelvin / 100, ".", DEC2 Kelvin

degC = Kelvin - 27315   ' units of 0.01 degC, resolution 0.02, can be negative[/SIZE] [SIZE=1]
DEBUG CR, "degC=", REP "-" \ degC.bit15, DEC ABS degC / 100, ".", DEC2 ABS degC

degF = Kelvin * 9/5 - 45967   ' units of 0.01 degF, can be negative[/SIZE] [SIZE=1]
DEBUG CR, "degF=", REP "-" \ degF.bit15, DEC ABS degF / 100, ".", DEC2 ABS degF[/SIZE]

note that the REP operator is used to handle the negative sign if needed, depending on the sign bit, bit15 of the number.