Maker Pro
Maker Pro

RTD Sensor

p.f.sanford

Sep 15, 2010
7
Joined
Sep 15, 2010
Messages
7
Hello Again,

I am trying to use an rtd sensor to measure temperature up to 600 degrees F, then measure it with an It can be off by about as much as 10 degrees at 600. I have found that rtd sensors are the cheapest way to accomplish this.

I see many ways to wires these depending on accuracy. Is it possible to just wire one lead to the A2D pin of the microchip and the other to ground? If so, how much would that effect it? Below is the link of the digikey part Im looking at. I also see they have they same one with same specs except 100 ohms. What exactly does that mean?

Thanks,
Pat

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=615-1045-ND
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I am trying to use an rtd sensor to measure temperature up to 600 degrees F, then measure it with an

Measure it with an.... what?

It can be off by about as much as 10 degrees at 600. I have found that rtd sensors are the cheapest way to accomplish this.

Whoa! I'd use a thermocouple. The first question you have to ask yourself is "How am I going to attach a cable to this device if it may exceed the temperature at which solder melts?"

I wouldn't have thought that an RTD was the cheapest method, but being off by 10 degrees in that sort of range indicated that you're only going to have minimal correction.

Check out this recent thread.

I see many ways to wires these depending on accuracy. Is it possible to just wire one lead to the A2D pin of the microchip and the other to ground?

Yes.

If so, how much would that effect it?

I think it would make it impossible to read a temperature with it. At a minimum you would require a constant current flowing through it so that the A2D input had something to read. The concern then is the resolution of the A2D since the voltages (and the changes in voltages) are quite small.

Below is the link of the digikey part Im looking at. I also see they have they same one with same specs except 100 ohms. What exactly does that mean?

It means it has a resistance of 100 ohms at 0 deg C (or at some other specified temperature).
 
Last edited:

p.f.sanford

Sep 15, 2010
7
Joined
Sep 15, 2010
Messages
7
Pardon me if I some of my questions are stupid... i really dont know what the hell im doing here.

I dont know why I did not look at the thermocouple. Ive looked at them at mouser but saw like $60 as the cheapest. Now Im finding these for a couple dollars.

http://www.virtualvillage.com/1m-3-3ft-k-type-thermocouple-temperature-sensor-probe-003820-033.html

Would that one be appropriate to measure ambient air temperature?

And how would I wire it in the circuit, connecting one to the A2D pin on the microchip?

Thanks,
Pat
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Go here, type "thermocouple" in the search box, then press enter.

You will see 2 devices listed. Read their specs, decide which approach you prefer, and go from there.

There are many devices that perform these functions. These were just quick and easy for me to find :)
 

p.f.sanford

Sep 15, 2010
7
Joined
Sep 15, 2010
Messages
7
Im confused. Is an amplifier or converter neccisary. I was under the understanding that I could simply read the voltage generated from the thermocouple with the microchips Im using (pic16f690). It has an ADC pin just for that.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Yes. Either the voltage differences are too small, or the impedance of the signal is too high, or both.

Imagine a resistor which varies from 100 ohms to 134 ohms between 0 and 100C. With 1mA going through it, the voltage varies from 0.1 V to 0.134V. So for 1C resolution, you need to be able to discriminate changes in voltage of 0.34mV.

A 10bit ADC can measure 1024 different voltages between 0 and 5 volts, so 5mV between values (that's 15degC). AND the low order bits may be noisy, so you might have to multiply that by 2 or 4. To get the require resolution, you would need at least 14 bits, preferably 16.

If, on the other hand, you have a circuit that gives you 0 to 1V for 0 to 100C, then a 10 bit ADC *might* be sufficient (a little more gain and it would be).

That all assumes a current of 1mA. The current you use may be 0.1mA, and the problem gets harder. In addition, you have to correct some of the non-linearity in your code. If you're going to have a sensible display (i.e. one that doesn't skip certain temperatures) you will need better than 1degC resolution.

Even worse, you want an output in degF. One degree F is just over 0.5 degC, so you need even more resolution.

OK, if you can get 16 but resolution on an ADC in a uC, then you might be *just* able to do it -- you can take many readings and average them to gain a bit or two (or to cancel out the noise).

Otherwise, you will need something, and why not use the correct something?
 
Top