Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Quickly Metering Air Temperature

Status
Not open for further replies.

Dacr0n

Member
This is going to be somewhat of a continuation of the thread called "Proportional Control Circuit w/ Thermocouple".


My focus at the moment is finding a good system to quickly detect air temperature changes in air flowing from a circular tube about 3" in diameter.

So far I have messed with Thermistors and Thermocouples. Both seem to be quick but not quick enough for my liking.

Airflow meters in some automobiles consist of simply one wire spanning across the intake tube, perpendicular to the air flow. This is some kind of very thin resistance wire that has current passed through it to warm the wire up, but also provides the computer with a resistance that changes depending on the amount of air flowing by and cooling the wire.


What I am wondering now is, could I use this kind of system on a smaller scale? Say....if I had a ultra thin wire spanning across where I want to measure the temp or maybe even a "T" for more accuracy and a circuit to read the resistance and provide an output for the controller?


What do you guys think of this idea?
 
Last edited:
The problem you have is called WIND CHILL.
The movement of the air also cools the wire. For instance, if the air is 30 degrees C and moving at 1m/s, the wire will get cooler when the air flows at 2m/s.
If the wire is not heated, the situation is different. You can use very thin nicrome wire from a rheostat (or wire wound resistor). Just remove the ceramic with a hammer and you have the sensor you want. Choose a high value.
 
The problem is going to be the thermal mass of the sensor. ie: A larger object, of the same material, takes longer to change temperature than a smaller one given a change in the ambient temperature. So keep the sensors mass low with the largest surface area possible.
 
That's why I suggested using a wire wound resistor with the highest resistance. It will have the thinest wire and create the greatest change in resistance per degree.
 
If by "wire wound resistor" you mean something like the attached photo, thats going to be way too slow.

I've already tried tiny TCs and thermistors like the one in the second photo.

If I get an ultra thin nichrome wire will that work to read the temp via the resistance when the hot air heats it?
 

Attachments

  • WW4.jpg
    WW4.jpg
    2.1 KB · Views: 137
  • thermistor.jpg
    thermistor.jpg
    9.4 KB · Views: 144
You have to get a 3watt wire wound resistor and smash off the ceramic outside.

Yes Ni-Chrome wire will change resistance when it is heated.

The easiest thing is to smash a good 40watt 120v or 240v globe
 
Last edited:
Could you elaborate a little bit more about this globe thing, where I could find one. and how I could use it.

Ohh and by the way... Im running on a 12v DC system.
 
If you want to measure the temperature difference from ambient then you will need two probes. One in the airflow and one outside to compare with.

From a table of resistance change versus temperature (**broken link removed**) copper wire is the best common metal for measuring temperature based upon the percentage change versus temperature. Calculated from Table 1, copper varies 0.22% per K(degree C), whereas tungsten varies 0.08% per K, and nichrome varies only 0.0004% per K.

The question is: how small a temperature difference do you need to measure?
 
I don't need to know the difference... I just need a quick component that gives me resistance (x) at my target temp. The circuit needs to control a heating element so that it heats until the sensor resistance is at target and proportionally controls it there.

How accurate am I trying to be? +/- 5 degrees C
 
Last edited:
What the previous poster has failed to take into account is resistance. The resistance of Tungsten is much higher than copper.
That's why I suggested Tungsten - as the filament is made from Tungsten. Ni-Chrome is used for the wire-wound resistors.
 
Last edited:
Please do not make absolute statements about what I did, or did not, take into account based upon your own level of understanding. The "previous poster" is well aware that the resistance of tungsten is higher than copper (about 3.3 times actually, is that "much"?).

My point was that the percentage change in resistance is what determines the relative change in signal level for a small temperature change, not the absolute resistance. Otherwise you would use nichrome (as was originally suggested) which has 18 times the resistance of tungsten. But with only 0.0004% change per K it would be hard to detect a 5 degree change. It's the percent change that determines how easy it is to measure a small change in temperature.

The low resistance of copper just means you will need to use a very small copper wire (such as a single strand of a 28AWG stranded copper wire) and wind it into a coil to get a reasonable resistance to measure with.

That is not to say that tungsten won't work. Since he only needs +/- 5 degrees accuracy, that would cause a 0.4% change in the tungsten resistance. You should be able to reliably detect that with reasonably careful circuit design.
 
I don't need to know the difference... I just need a quick component that gives me resistance (x) at my target temp. The circuit needs to control a heating element so that it heats until the sensor resistance is at target and proportionally controls it there.
The heating element will also have thermal mass and will continue to produce heat after power input is reduced. A properly designed PID control loop should take care of any lag in the system.
What exactly are you trying to build?
 
the percentage change in resistance is what determines the relative change in signal level

nichrome (as was originally suggested) which has 18 times the resistance of tungsten. But with only 0.0004% change per K it would be hard to detect a 5 degree change. It's the percent change that determines how easy it is to measure a small change in temperature.

The low resistance of copper just means you will need to use a very small copper wire (such as a single strand of a 28AWG stranded copper wire) and wind it into a coil to get a reasonable resistance to measure with.

That is not to say that tungsten won't work. Since he only needs +/- 5 degrees accuracy, that would cause a 0.4% change in the tungsten resistance. You should be able to reliably detect that with reasonably careful circuit design.


So if I am going more for accuracy, would going with the tungsten be a better choice since it probably has less mass than a 28AWG wire and respond quicker?

What is the % change per K of tungsten?
 
Last edited:
The tungsten filament is 60 ohms and you should try it first to see if you can measure any increse in resistance for say 5 degrees rise in temp.

You have to start somewhere.
 
I tried the filament of a lamp and the change is miniscule.

The only other thing to do is take the coating off a "thermal bead" to make its response time faster.
 
What exactly are you trying to build?

I am trying to get a better understanding of heating systems, PID Control, and control theory by building a small scale heating system with highly accurate variable control.

The link in the first post has more info.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top