On Thu, 25 Dec 2008, cdb wrote:

>  I have been asked to alter one of my temperature controller projects 
> to work as an egg incubator. I need to implement a better heater 
> control mechanism.
> 
> I have decided not to use a PID algorithm as I have no way of knowing 
> the placement of sensors, heater or even size of enclosure, so I 
> decided to use a rate of change self calibration method, I have found 
> some example code for an oven, which is a commercial product and 
> should work. I understand what the code is trying to do, I just don't 
> understand where the constants they've used have come from. The code 
> note just states //25 * 4 = 100 deg C.  Which as you will see doesn't 
> appear correct. The code:
> 
> deltaT = ((3 x deltaT) + ((read_temp() - tmp) x 25))/4
> 
> Ok tmp contains the previous temperature measurement.  The divide by 4 
> is due to the original circuit using a Maxim thermocouple IC so a 
> reading of 400 = 100 deg C - this will be superfluous as I'm not using 
> a thermocouple. 
> 
> I cannot work out what the x25 is for or where they get the 3 x  I'm 
> assuming this is a form of digital filtering and scaling, I also think 
> this code would surely fail if the readings were ever zero. All 
> variables are signed ints.
> 
> The above code is called at a constant time beat so the rate of change 
> time component is taken care of.
> 
> Or perhaps is there another method I should think of?
> 
> Thanks
> 
> Colin

The x25 looks like it's just scaling the measurement.

look at it like this:

new_temp = read_temp()

new_deltaT = (new_temp - old_temp) * scale

deltaT = (deltaT * 0.75) + (new_deltaT * 0.25)

old_temp = new_temp

NOTE: that 0.75 + 0.25 == 1.0, as you increase 0.75 towards 1.0 and 
decrease the 0.25 towards 0 the new_deltaT is having a smaller effect on 
the calculation of deltaT, so yes deltaT is being filtered

Regards
Sergio Masci
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist