thermister formula

Electronics Computer Programming Q&A
Post Reply
formula KZ1000
Posts: 9
Joined: Thu Dec 04, 2003 1:01 am
Contact:

thermister formula

Post by formula KZ1000 » Thu Dec 04, 2003 7:57 am

Hi I am new to this board and just learning about PIC chips.<p>I am building a tire temp gauge for my racecar.<p>I am currently using thermister that I got from Radio Shack part number 271-110.<p>It has a table on the back that has the ohms to temp values.<p>I now just have it connected to an ohms meter and after I get the values I have to compare them to the chart to get the temp.<p>I am going to use a PIC chip to display the temp, but I cannot figure out the formula that is used to derive the values.<p>It is not a linier but a curved scale.<p>It has been to long since I took Business calculus, I think it is a log function.<p>Unfortunately I do not have the actual chart with me at work here.<p>I Googled looking for such a formula and found the following:
T=(1/((2.656*10^-3)+(2.317*10^-7)ln(R)^3)) -273.15<p>R is resistance in Kohms
ln() is the natural logarithm function (Base e)
T is temperature in °C<p>Is this formula good for all thermisters or just the one that the site was referring to?<p>I will post the chart that came with the thermister when I get home.<p>Thanks for all the help.<p>Eddie

hlreed
Posts: 349
Joined: Wed Jan 09, 2002 1:01 am
Location: Richmond, TX
Contact:

Re: thermister formula

Post by hlreed » Thu Dec 04, 2003 9:52 am

Eddie,
You do not need to know the formula since you have a table, converting ohms to temperature.
I would convert the ohms to volts, use a PIC that has ADC and output the temperature from an internal volts to temperature table.
Harold L. Reed
Microbes got brains

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: thermister formula

Post by MrAl » Thu Dec 04, 2003 2:46 pm

Hello Eddie,<p>No need to post the chart, unless you still
want too...<p>The first thing to note is that the chart is
incorrect for the entry of 90 degrees C.
The correct value should be 1.266, not 1.366
(in K ohms).<p>I had developed a formula several years ago for
this very thermistor, and since i wanted close
to 1/10 percent accuracy (one tenth of one percent)
i ended up using an exponential with a fourth
degree algebraic exponent.<p>Here's the formula for R:<p>R = e^(A4 x^4 + A3 x^3 + A2 x^2 + A1 x + A0),
{with R in k ohms, x is temperature in C}<p>where
A4= +4.913E-10
A3= -2.941E-7
A2= +1.2093E-4
A1= -4.2966E-2
A0= +3.3053<p>With these coefficients, the accuracy is
just a hair over 1/10 of a percent, and so
is much better then the claimed 1.0% accuracy
of the thermistor itself.<p>Since this is a formula for R (not for T)
you have to solve for the zeros in order to
determine T from R. If you would rather
have an inverse function let me know, as that
would be easier to use in practice.<p>It's also possible to come up with a simpler
formula with less constants since your
application wont need to be that accurate.<p>Another idea is to store the table and use
linear interpolation between stored values.<p>Take care for now,
Al
LEDs vs Bulbs, LEDs are winning.

L. Daniel Rosa
Posts: 400
Joined: Tue Dec 25, 2001 1:01 am
Location: Bellingham, WA
Contact:

Re: thermister formula

Post by L. Daniel Rosa » Thu Dec 04, 2003 5:34 pm

FWIW, the LM334 and LM335 are both linear temperature sensors with a Kelvin scale. The '334 is a current regulator, which is good for long wire runs or connections with an uncertain resistance. The '335 behaves like a temperature dependant Zener. Both are three terminal devices that with the support components (very few to none) can function remotely with two wires.

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: thermister formula

Post by MrAl » Fri Dec 05, 2003 4:44 am

Hello,<p>L. Dan:
That's an interesting idea too. I'd want to
check the max operating temperature for that
device first though. <p>Eddie:
Here's a much simpler formula for getting
the temperature T (in deg C) knowing the
resistance R (in k ohms)<p>T = 1.855 x^2 - 36.1 x + 98.4
where
x=ln(R)<p>The above is accurate to 1.5 deg C but the
inaccuracy reduces to 1 deg C max if you
elminiate both endpoints. This means from
-45 deg C to +105 deg C the accuracy is within
about 1 deg C which is probably good enough for
your application.<p>Take care,
Al
LEDs vs Bulbs, LEDs are winning.

formula KZ1000
Posts: 9
Joined: Thu Dec 04, 2003 1:01 am
Contact:

Re: thermister formula

Post by formula KZ1000 » Fri Dec 05, 2003 5:41 am

Thank you all for your help.<p>I will most likely use MrAl formula, as it best fits my needs.<p>I did not want to use the table, as the values are to spread out.<p>Also, a thanks goes to Nuts and Volts for their great magazine this form.<p>I am just starting to explore the wonderful world of electronics.<p>After finishing this small project, I hope to build my own data acquisition system for my car.<p>Image

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: thermister formula

Post by MrAl » Sat Dec 06, 2003 3:05 am

Hello again Eddie,<p>Nice car :-)<p>I didnt realize these cars could benefit from
tire temperature measurements.<p>Take care,
Al
LEDs vs Bulbs, LEDs are winning.

formula KZ1000
Posts: 9
Joined: Thu Dec 04, 2003 1:01 am
Contact:

Re: thermister formula

Post by formula KZ1000 » Tue Dec 09, 2003 8:59 am

Al,<p>Thanks, yes this is how I set camber and tire pressure.<p>After a run we measure 3 point on each tire each side and the center.<p>If the center is hotter than both outside edges that pressure is to high.<p>If outside edge is hottest than to much negative camber, etc...<p>Though I have moved on to a different project as I picked up a used tire pyrometer this weekend for cheaper than it would cost to build.<p>Thanks for all your help I am sure I will use it for another project sometime.<p>Eddie

myp71
Posts: 585
Joined: Wed Apr 23, 2003 1:01 am
Location: Colorado
Contact:

Re: thermister formula

Post by myp71 » Tue Dec 09, 2003 6:33 pm

Hi,<p>just for my info is this temp real time or a touch and tell?<p>Do you touch the circuit to the tires to get the temp or is the system inside the tries somehow with a readout inside the car?<p>Myp71

formula KZ1000
Posts: 9
Joined: Thu Dec 04, 2003 1:01 am
Contact:

Re: thermister formula

Post by formula KZ1000 » Wed Dec 10, 2003 4:58 am

Myp71,<p>I use one probe and manually insert the probe about a 1/4 inch into the tire.<p>Eddie

Post Reply

Who is online

Users browsing this forum: No registered users and 5 guests