Measuring the temperature from a pt100 RTD using a breadboard circuit.

Thread Starter

Reyleight

Joined Dec 13, 2016
25
So I have 8 pt100 RTD's that I need to measure the temperature of, and I know the resistance of the RTD changes as its temperature goes up therefore I would use the resistance to obtain the temperature. I currently have my wheatstone bridge set up and I have a 2.5 V source going into my wheatstone bridge. My setup currently is as shown in my uploaded circuit configuration with a balanced circuit having all the resistors at 100 ohms and the RTD being the only thing that has a change in resistance. Currently I get the correct voltage difference out when measuring across the resistors when my RTD changes resistance ( since it is hooked up to a heat pump).

The thing is that I have to get readings off of Labjack, a program that measures the voltage of the wires connected to it. In this case I am using a labjack U3, so if a wheatstone bridge where my RTD resistance is 115 Ohms, and the others are 120 ohms and a voltage source of 2.5 V my V_bridge would be .0872 V, and I get approximately that. I know I need a DIP in-amp for my breadboard, but I do not know what a good one to buy would be that could meet my specifications. Any good recommendations?

How do I get the labjack which displays the voltage of the circuit, in this case my voltage across my bridge to be a temperature? I have the file with the equation relating Resistance and temperature attached. One thing to note is that the labjack can not read voltages higher than 2.4 V so my V_output of the in-amp should not exceed that.

As far as dealing with negative temperatures, I may be dealing with some slightly negative temperatures but if it was truly negative then it would freeze the water flowing through the heat pump so I don't believe that is correct. That being said I think I'll just be dealing with temps greater than 0 degrees Celsius.

http://www.micropik.com/PDF/pt100.pdf This website also shows the relationship between resistance and temperature for this particular RTD.

Would appreciate any help! Thanks!

TLDR: Recommended DIP In-amp for a temperature sensor (pt100) with a voltage source of 2.5 V ? How do I obtain my temperature with my v_output using the equation in the z251 pdf file?
 

Attachments

Thread Starter

Reyleight

Joined Dec 13, 2016
25
Is this homework (or a lab project of some kind)?
It's a project, I'll end up setting up a step by step procedure to be able to set the circuit to read the temperature of the RTD's via labjack. This system has been sitting unused for years. Afterwards I'll end up ordering a more permanent circuit (custom built).
 

Thread Starter

Reyleight

Joined Dec 13, 2016
25
Just looking at the fundamental variables, if you were able to measure the current going through the RTD, while also measuring the voltage across it, the resistance would be the ratio of the two quantities. An instrument such as the one here should be able to measure the RTD current:
https://www.amazon.com/SEOH-Galvano...kcompar-20&ascsubtag=58509d19e86f9e0b94ee57d4
Right, I had measured my current with a multimeter. Then I did v_diff / current to get the resistance however it didn't give me a value anywhere close to what the resistance of the RTD actually was. (I measured it). The thing is I have to use that voltage across the resistors to calculate the temperature. Or when I use an in-amp the V_out. Which with a gain of 1 should be the same as the voltage difference across the wheat stone bridge.
 

drc_567

Joined Dec 29, 2008
1,156
If you look at the labjack webpage, there is a device .. tick something ... That is supposed to work with the RTD. So take a look at that item .... If it requires an op-amp, that can be fixed up.
 

Reloadron

Joined Jan 15, 2015
7,501
You have a few options. Using a RTD within a Wheatstone Bridge you are going to need a instrumentation amplifier to get the signal up to a usable level. When using a bridge or any method resulting in a current passing through the RTD there can be problems with the RTD having a "rtd self heating effect" and the link just being one example, a Google of RTD Self Heating Effect should bring up a few dozen examples and hits.

Within industry one pretty common solution to the use of RTDs or Thermocouples is the use of RTD or Thermocouple Transmitters, many of which are programmable. Here is an example of what I am talking about. They come literally in a dozen flavors with a dozen prices. Most will output a 4 to 20 mA current loop. You place a resistor in the current loop of for example 250 or 500 Ohms resulting in a 1 to 5 Volt or 2 to 10 Volt voltage drop proportional to the measured temperature. Then you scale your analog to digtal converter be it Labjack or similar for the engineering units, Degrees F or Degrees C.

Aside from a 100 Ohm RTD there are a good number of off the shelf solutions like the LM 35 Precision Centigrade Temperature Sensors. If you choose the Wheatstone Bridge option keep in mind your resistors used in the bridge should be 1% precision resistors with a good temperature coefficient. The INA you choose should be a low noise flavor and preferably with programmable gain requiring a low parts count since you will need to fabricate a circuit board for everything.

Ron
 

Thread Starter

Reyleight

Joined Dec 13, 2016
25

Thread Starter

Reyleight

Joined Dec 13, 2016
25
You have a few options. Using a RTD within a Wheatstone Bridge you are going to need a instrumentation amplifier to get the signal up to a usable level. When using a bridge or any method resulting in a current passing through the RTD there can be problems with the RTD having a "rtd self heating effect" and the link just being one example, a Google of RTD Self Heating Effect should bring up a few dozen examples and hits.

Within industry one pretty common solution to the use of RTDs or Thermocouples is the use of RTD or Thermocouple Transmitters, many of which are programmable. Here is an example of what I am talking about. They come literally in a dozen flavors with a dozen prices. Most will output a 4 to 20 mA current loop. You place a resistor in the current loop of for example 250 or 500 Ohms resulting in a 1 to 5 Volt or 2 to 10 Volt voltage drop proportional to the measured temperature. Then you scale your analog to digtal converter be it Labjack or similar for the engineering units, Degrees F or Degrees C.

Aside from a 100 Ohm RTD there are a good number of off the shelf solutions like the LM 35 Precision Centigrade Temperature Sensors. If you choose the Wheatstone Bridge option keep in mind your resistors used in the bridge should be 1% precision resistors with a good temperature coefficient. The INA you choose should be a low noise flavor and preferably with programmable gain requiring a low parts count since you will need to fabricate a circuit board for everything.

Ron
Yes you're completely right about the 1% precision resistors, which they are. I'll look into transmitters you mentioned. Right now for the INA I was thinking of a INA 129P for my uses which I believe is low noise and programmable gain.

I did think that the voltage drop across the bridge should be scaled to the appropriate temperature I just didn't know how to do it. Since when dividing that voltage by the current I got a very small resistance that would have given me an even smaller temperature. You also mention the INA is to get the signal to a usable level, what do you mean by this? By increasing the gain I increase the voltage drop across the bridge as a voltage out of the INA. However why would I want to increase that voltage? And if I increase it by say a magnitude of 2 does that mean whenever I calculate my resistance do I have to x2 that as well?
 

Reloadron

Joined Jan 15, 2015
7,501
As to your INA either of the INA12x Precision, Low Power Instrumentation Amplifiers would do fine. More on that later.

You also mention the INA is to get the signal to a usable level, what do you mean by this?
Here is what I am getting at. Keep in mind I have never used the LabJack modules but have used similar for data acquisition. My read on the LabJack U3 Series is there is a U3 LV (Low Voltage) and a LU3 HV (High Voltage). Looking at the data sheet the Analog In is a 12 bit A/D converter. So whatever the full scale voltage is for a channel the result will be 2 ^ 12 or 4096 bits. So if I have a U3HV with a +/- 10 Volt input you will get 4096 to -4096 bits with -10 volts to 10 volts applied. To scale the channel to read volts for example in code it would look something like: Analog Input / 4096 * 10. If we apply 1 Volt the bit count should be 409.6 so we get 409.6 / 4096 * 10 = 1 Volt. There is a little more to it but that is about how it works with most A to D modules. You scale to the engineering unit be it PSI, Degrees C or Degrees F, whatever it is.

As to the circuit itself. The following is a rough drawing of a bridge with a 100 Ohm RTD.
RTD CKT.png

The 100 Ohm is R4 in the drawing. If we do a plot of the current through the RTD is looks a little like this:
RTD CKT.png RTD Plot.png

What I did here is over 10 seconds time I stepped the RTD resistance value between 100 Ohms (32 Deg F.) and 138.5 Ohms (212 Deg F) in one ohm steps. The gain of the Instrument Amp is set at 10. The current runs from 1.946 mA to about 1.961 mA with the INA input ranging from 0 volts to about 73.5 mV and a Vout of about 735 mV. Sorry the plot is sort of ugly.

Ron
 

Thread Starter

Reyleight

Joined Dec 13, 2016
25
As to your INA either of the INA12x Precision, Low Power Instrumentation Amplifiers would do fine. More on that later.



Here is what I am getting at. Keep in mind I have never used the LabJack modules but have used similar for data acquisition. My read on the LabJack U3 Series is there is a U3 LV (Low Voltage) and a LU3 HV (High Voltage). Looking at the data sheet the Analog In is a 12 bit A/D converter. So whatever the full scale voltage is for a channel the result will be 2 ^ 12 or 4096 bits. So if I have a U3HV with a +/- 10 Volt input you will get 4096 to -4096 bits with -10 volts to 10 volts applied. To scale the channel to read volts for example in code it would look something like: Analog Input / 4096 * 10. If we apply 1 Volt the bit count should be 409.6 so we get 409.6 / 4096 * 10 = 1 Volt. There is a little more to it but that is about how it works with most A to D modules. You scale to the engineering unit be it PSI, Degrees C or Degrees F, whatever it is.

As to the circuit itself. The following is a rough drawing of a bridge with a 100 Ohm RTD.
View attachment 116896

The 100 Ohm is R4 in the drawing. If we do a plot of the current through the RTD is looks a little like this:
View attachment 116896 View attachment 116897

What I did here is over 10 seconds time I stepped the RTD resistance value between 100 Ohms (32 Deg F.) and 138.5 Ohms (212 Deg F) in one ohm steps. The gain of the Instrument Amp is set at 10. The current runs from 1.946 mA to about 1.961 mA with the INA input ranging from 0 volts to about 73.5 mV and a Vout of about 735 mV. Sorry the plot is sort of ugly.

Ron
Ohhh, what you did does make sense. Most of the channels on the U3 already show you the output voltage from the INA, is there a way to obtain the RTD resistance via that voltage?
 

Reloadron

Joined Jan 15, 2015
7,501
What temperature range are you looking for?
A RTD sensor, like a thermocouple is unfortunately not linear, that is why I always liked the Temperature Transmitters I mentioned earlier or a sensor like the LM 35. With a 100 Ohm RTD we can say 100 Ohms is 32 Degrees F but 200 Ohms is not 64 Degrees F. :)

Analog linearization of resistance temperature detectors is a pretty good read on the subject. The RTD has advantages over the thermocouple but neither really has a linear output unfortunately. Again, while working in an industrial environment this is why I loved those little temperature transmitters. The programmable ones were really sweet.

Another detail before I forget is anytime we use a bridge circuit be it with a force sensor, temperature sensor, pressure sensor or whatever sensor the applied excitation voltage needs to be very stable. That is very important.

Ron
 

Thread Starter

Reyleight

Joined Dec 13, 2016
25
What temperature range are you looking for?
A RTD sensor, like a thermocouple is unfortunately not linear, that is why I always liked the Temperature Transmitters I mentioned earlier or a sensor like the LM 35. With a 100 Ohm RTD we can say 100 Ohms is 32 Degrees F but 200 Ohms is not 64 Degrees F. :)

Analog linearization of resistance temperature detectors is a pretty good read on the subject. The RTD has advantages over the thermocouple but neither really has a linear output unfortunately. Again, while working in an industrial environment this is why I loved those little temperature transmitters. The programmable ones were really sweet.

Another detail before I forget is anytime we use a bridge circuit be it with a force sensor, temperature sensor, pressure sensor or whatever sensor the applied excitation voltage needs to be very stable. That is very important.

Ron
0-100 Degrees C would be the highest I would need to read at, and I doubt it would even reach 100 C. Yeah I see what you mean by it not being a linear output. As for the excitation voltage I'll make sure of that.
 

Reloadron

Joined Jan 15, 2015
7,501
OK, then I would consider the following. Make a bridge about like I drew it. When choosing Rg (Gain Resistor) I would look around for a 5 K ten turn or even better 5 K twenty turn trimmer pot. With about 1.6K the Rg will get you close to 212 Deg F = 2.4 Volts. As to actually checking the temperature you can try using some crushed ice and distilled water and make a Ice Bath Slurry as best you can. While not perfect it will be close. Next place the sensor in a pot of boiling water and adjust your Rg for 2.4 Volts. Again, unless you are at sea level under ideal conditions it will not be 100 C but will be close enough. That should give you a zero and span. Do not connect to the analog input channel until Rg is adjusted. I am guessing these devices have some form of input protection?

Actually this 2K ten turn pot would be a good Rg.

Ron
 

wayneh

Joined Sep 9, 2010
17,496
I've used a Labjack U3-HV to collect data. I used Excel to handle all the data processing and calculations. The LabJack is easily controlled with Visual Basic within Excel, so I had it collecting something like 10 voltages every minute over night, adding the data points to a chart in Excel.
 

Reloadron

Joined Jan 15, 2015
7,501
I've used a Labjack U3-HV to collect data. I used Excel to handle all the data processing and calculations. The LabJack is easily controlled with Visual Basic within Excel, so I had it collecting something like 10 voltages every minute over night, adding the data points to a chart in Excel.
Oh yes, when we collect data which we may want to save using Visual Basic and Excel is a great way to go. I also liked using Access as a data base to place data in. If you plan to collect data rather than just view current data then think about Excel or Access. Something I never did try was using any of the Open Office equivalent programs to store data, Calc and Base would be interesting to try.

Since I have never used LabJack modules I have never looked at their code samples but most modules for data acquisition like this normally include some good visual basic examples.

Ron
 
Top