Excel Formula to PICAXE

[Lliam20789]

Hello everyone,
I have been experimenting with a NTC thermistor and have finally come up with a conversion formula, but...

temp=0.00032442727*w0^2-0.342224714*w0+115.3133411

will obviously not work in a PICAXE... Does anyone know how to do this?

Thankyou,
Lliam

 

[Brietech]

how accurate does it need to be? If this is a dedicated application, and you have room on your chip (or on i2c memory), you could just make a lookup table for all of the values of w0 you're likely to get.

 

[Lliam20789]

Doesn't need to be overly accurate and a look up table is an option.
So it is as hard as it looks to get the PICAXE to do that?

Thanks for your help so far,
Lliam

 

[Dippy]

Any sort of stuff like this needs breaking down into separate small integer functions and keeping an eye on overflows and negatives.
eg. your 0.00032442727*w0 is the same as w0/3082 approx.

I think your numbers of digits may be trimmed a bit.
I would think that with tolerances of thermistors that w0+115.3133411 maybe a little OTT. How about just w0+115?

So, split it into smaller functions, check for possiblity of overflows etc, then add them back together.
Go on, give it a go. Be brave and get the pen and paper out.

 

[Jeremy Leach]

If you actually draw this out as a graph (or in Excel) you might spot a much simpler formula you could use which will give a good enough result.

 

[puddlehaven]

PH Anderson has an about appnote reading a thermistor and then using an abbreviated lookup table and interpolation to get a resolution of "nominally 0.1 Degrees C over the range of 0.0 to 37.0 degrees C and 0.2 degrees over the range -20.0 to 65 degrees C."

Look here:

http://www.phanderson.com/picaxe/picaxe_thermistor.html

 

[Lliam20789]

Thanks

I tried getting a simpler formula but I'm all out of ideas...
I can try Dippy's idea but I'm not the greatest at maths.

I'll let you know how I go.

Thankyou all for you help,
Lliam