[En-Nut-Discussion] Question about ADC stability

[Rich Wellner]

I've hooked an lm35 to adc0 and am surprised at the results I'm getting.

    double temp;
    u_short value = 0;
    //int i=0;

    ADCInit();
    ADCSetChannel(ADC0);
    ADCBufferFlush();
    ADCSetMode(FREE_RUNNING);
    ADCStartConversion();

    for(;;){
    while (ADCRead(&value))
            NutThreadYield();
    printf("raw value: %i\n", value);
    printf("Temp C: %f\n", (double)value / 1023.0 * 5.0 *100);

    }

It goes without saying, I hope, that I haven't been changing the temp 
dramatically.  I've also confirmed that the sensor it putting out a 
steady .232V, which is what I expect in my slightly warm workroom at the 
moment.  Are there additional things that need to be done to get steady 
readings from this adc?  Also, if I switch to SINGLE_CONVERSION then I 
get a more correct reading on average (e.g. about 23C), but it's still 
highly variable as with the readings below.

raw value: 96
Temp C: 46.920824
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 79
Temp C: 38.611922
raw value: 71
Temp C: 34.701857