Temperature Measurment Over a Large Surface

[madddog]

Can anyone give me some help.

I want to measure the temperature of an airconditioning condenser coil. Since this is a large surface area I would like to take place eight thermistors around the top third and eight more around the bottom third. Since my data acquisiton system has limited I/O I want to input the top group of sensors into one analog channel and the bottom group into a second.

My question: Should I wire the thermistors in series or parallel if I want to take an average reading at the two locations? Would the response time be effected either way? Also which way would give me the most sensitive reading?

The resistance of the thermistors is 10K and I will be driving them with a voltage source via remote sense resistor.

 

[nbucska]

Why don't you measure them sequentially, so you can have separate calibration factors, calculate max, min, average
etc. ? <nbucska@pcperipherals.com>

 

[uWaveFDL]

Thermal imaging an option?

 

Hello,

You can't wire Thermistors in Parallel nor in series. They will have to be wired individually. This means you will need some type of device which can read all sixteen channels. There are commercial devices available to do this which use an embedded processor and sends the results to a computer which is saved in a file. This file can be looked at using a spreadsheet such as Microsoft Excel. You can plot over a period of time and get meaningful graphics. If you are interested in more information please contact me.

Good Luck

TemperatureToys@aol.com

 

[madddog]

TemperatureToys

Thanks but, can you explain why you can't wire thermistors in paraller or series. They are nothing but temperature sensitive resistors and you can wire resistors in series or parallel to obtain a specific resistance.

 

[nbucska]

Thermistors have wide variation in value and characteristics
it would be almost impossible to find matching ones, which
could be averaged by parallel or serial connection.

Your option: measure them individually, using individual calibration factors convert reading to temp. Average temp.'s.

<nbucska@pcperipherals.com>

 

[pebe]

There is an alternative to thermisters - ie. diodes.
I have used a forward biassed diode as a temperature sensor in a heating cabinet.

The temp coeff of any diode is -2mv per degree C. I think you should be able to connect a number in series and feed them from a (low) constant current source. You could then feed the resultant voltage drop across the string of diodes, to an op-amp. The change in voltage would give you the AVERAGE temp change across the surface area.

In the application I used I could easily sense to 0.25 degrees F.

 

[BrianR]

You will not get any sort of accuracy connecting thermistors in series or in parallel.

You may connect thermocouples in parallel to get an average.

Why do you want to do this measurement?

If you are trying to measure the heat transfer of the system by measuring differential air temperature close to the input and output of the heat exchanger then parallal thermocouples will do the job. Similarly knowing the fluid flow rate, a measurement of the fluid inlet and outlet temperatures will work.

 

[madddog]

Basically I want to measure the coil temperature averaged over the top third and bottom third. Then spray conditioned water at various locations on the outside of the coil and measure the dip in temp as the water evaporates, removing heat. Measuring the temp with sensors in series / parallel would average the temp over the entire surface even though the spray heads may not wet all of the coils surface at the same time.

Do you have any suggestions?

Thanks

 

[BrianR]

madddog,

>>Basically I want to measure the coil temperature averaged over the top third and bottom third.

Two groups of parallel connected thermocouples would achieve this.

>>Then spray conditioned water at various locations on the outside of the coil and measure the dip in temp as the water evaporates, removing heat.

Heat exchanger efficiency is generally greatest when the water is sprayed over the cooler end of the coil where, along with air movement it can reduce the coolant outlet temp below ambient. If sprayed over the hot end then ambient is as cool as the outlet can get.

>>Measuring the temp with sensors in series / parallel would average the temp over the entire surface even though the spray heads may not wet all of the coils surface at the same time.

The coolant flow through the coil will do a pretty good job of averaging the temp over a short length of tube, but if you really want individual measurements you'll need to buy a multi channel multiplexer for your data logger, and that will require at least 3 output lines to control the switching of 2 pairs of 8 channels into 2 analog inputs or 4 outputs to multiplex 16 analog inputs into 1 input.


Do you have any suggestions?

Thanks

 

[DRWeig]

Hi,

If I can throw in on this one from an HVAC controls perspective, Precon makes an averaging thermistor sensor that goes up to 25 feet long with 9 thermistors inside. They're matched to within 0.1 degree F, and they are wired in a series/parallel configuration that gives you the equivalent average resistance at the two leads. You could do the same thing if you don't want to buy a factory built component, but you need to get matched sensors as someone stated above.

These are used in bazillions of air handling units. Let me know if you want manufacturer's data....

Happy measuring!

Old Dave