I need a thermocouple sensor which have a speed of 50hz or 20 samples per second for a project.

Thread Starter

sidring

Joined Mar 7, 2017
7
Hello All,

I need a thermocouple sensor which have a speed of 50hz or 20 samples per second for a project.

Please suggest some solutions.

Thanks
 

Alec_t

Joined Sep 17, 2013
14,280
Welcome to AAC.
Amplifier >> ADC >> MCU.
If you want detailed help you must give us more details of your requirement, e.g. thermocouple specification, accuracy expected, output required .....
Is this a school/college project?
 

Reloadron

Joined Jan 15, 2015
7,501
You need to define the thermocouple type and application, then the low level signal needs to be amplified and linearized and then passed along to an MCU where a voltage or current proportional to temperature is read and displayed depending on your needs.

Thermocouple Types
Thermocouple Amplifiers and Transmitters

There are several companies which offer "Starter Kits" for reading thermocouple inputs. You really need to well define what Alec has mentioned before anyone can be of much help.

Ron
 

DickCappels

Joined Aug 21, 2008
10,153
...and while at it would you please expand on what you mean by "20 samples per second". Very few thermal sensing technologies can settle to a step change to high accuracy in that short of a time period.
 

Thread Starter

sidring

Joined Mar 7, 2017
7
Hi All,

Thanks for the reply. I am using K type thermocouple and wants to measure liquid temperature with thermocouple.
I mean i want to see the change in temperature within 20ms. if temperature changes within 20ms then thermocouple amplifier should be able to amplify it with in that time period. i wants to see these results on y labview interface.

Thanks
Siddharth
 

Thread Starter

sidring

Joined Mar 7, 2017
7
...and while at it would you please expand on what you mean by "20 samples per second". Very few thermal sensing technologies can settle to a step change to high accuracy in that short of a time period.
Hi,

How can i sense the temperature change with in that short time period. I mean, i need a thermocouple amplifier which can able to amplify thermocouple signal to with in 20ms.
 

Reloadron

Joined Jan 15, 2015
7,501
I really suggest you give this a read Thermocouple Response Times and become familiar with thermocouples and their characteristics as well as uncertainties.
The “Time Constant” or “Response TIme” is defined as the time required to reach 63.2% of an instantaneous temperature change.
Finally larger thermocouples made with heavier gauge wire will have even slower response times. You have not defined temperatures or temperature change, other than K you have not defined your Type K thermocouple. As to a thermocouple amplifier there are dozens of turn key solutions out there and available. A simple Google of Thermocouple Amplifier will bring up a dozen results. You also need to consider the response times of anything connected to the Thermocouple. The amplifier isn't the problem, getting a thermocouple to respond to changing temperature as fast as you want is the problem and again, I see no mention of temperature range or anticipated slew rate?

Ron
 

Thread Starter

sidring

Joined Mar 7, 2017
7
I really suggest you give this a read Thermocouple Response Times and become familiar with thermocouples and their characteristics as well as uncertainties.


Finally larger thermocouples made with heavier gauge wire will have even slower response times. You have not defined temperatures or temperature change, other than K you have not defined your Type K thermocouple. As to a thermocouple amplifier there are dozens of turn key solutions out there and available. A simple Google of Thermocouple Amplifier will bring up a dozen results. You also need to consider the response times of anything connected to the Thermocouple. The amplifier isn't the problem, getting a thermocouple to respond to changing temperature as fast as you want is the problem and again, I see no mention of temperature range or anticipated slew rate?
Ron
Hi Ron,
I am using the Type K thermocouple for 0 - 800 C range.Right now i am using Ad8495 as thermocouple amplifier. I found on some blogs that its response time fast enough(20ms). Means i want to see the temperature change within 20ms in my graph on labview
Is it okk to use it.

one more question.
While using this when i plot the temperature curve on labview there is unexpected dip in temperature. can you tell me why.

Please look at the attached image.

Thanks
Siddhartherror.png
 

Kermit2

Joined Feb 5, 2010
4,162
It is a function of "mass". The physical size of the sensor means it will respond only when the ENTIRE mass of metal has increased in temperature.
The only way to increase response speed is to decrease the physical mass of the sensor.

Secondly, response time is also dependent on difference of temperature between sensor and DUT. A change of .1 degree will take much much longer the trigger a change in thermocouple output than a change of 10 degree would.
 

OBW0549

Joined Mar 2, 2015
3,566
Right now i am using Ad8495 as thermocouple amplifier. I found on some blogs that its response time fast enough(20ms).
Whatever blogs you were reading, don't read them any more. READ THE AD8495 DATA SHEET. If you do that, you'll see that the frequency response of the AD8495 is down 3 db at 25 kHz, which corresponds to a response time of about 15 μs. That's microseconds, not milliseconds. So the AD8495 is over a thousand times faster than you need it to be, and therefore should be well suited for the job.

Means i want to see the temperature change within 20ms in my graph on labview Is it okk to use it.
Of course it's OK to use it. Whether or not you can get 20 ms response time from your thermocouple is another matter entirely, and will largely depend on the size and thermal mass of the thermocouple, the thermal conductivity of the liquid it's immersed in, and whether the liquid is moving and if so, how fast.

And THAT is something only you can determine.
 

Reloadron

Joined Jan 15, 2015
7,501
OK, I understand Type K Thermocouple. Here is what I am getting at and what Kermit2 pointed out. Below are pictured two Type K thermocouples. They are open bare wire thermocouples. The large image uses AWG 12 diameter (2.05mm) wire gauge, while the small inset thermocouple uses AWG 22 (.65mm) wire gauge. The small diameter wire will respond to a temperature change much quicker but will not stand up to higher temperatures.

Both are Type K bare wire thermocouples. However, the larger AWG 12 wire will have a much slower response to a temperature change. Many thermocouples may be enclosed in a sheath material. For example below is another example of a Type K thermocouple but the junction is enclosed in a sheath material.

Type K TC.png
The material the sheath is made out of, the thickness of the sheath and the insulating material inside the sheath all effect the thermocouple response time. If I take a thermocouple at room ambient temperature and shove it into a flame all thermocouples will not respond to the sudden temperature change the same. So while I can make a pretty plot it matters not how fast I sample (sample rate) the temperature as the numbers are all garbage if I am sampling faster than the thermocouple can respond to a changing temperature. This is why I gave you a link to an example of a thermocouple response curve. Thermocouple response is not instantaneous.

Now again, do you have a data sheet for your thermocouple and don't tell us it is a Type K? The reason thermocouples are not sampled fast is they have slow response curves. What do you expect to gain using a fast sample rate?

Ron
 
Last edited:

Reloadron

Joined Jan 15, 2015
7,501
Hi All,

Thanks for the reply. I am using K type thermocouple and wants to measure liquid temperature with thermocouple.
I mean i want to see the change in temperature within 20ms. if temperature changes within 20ms then thermocouple amplifier should be able to amplify it with in that time period. i wants to see these results on y labview interface.

Thanks
Siddharth
Thinking further about this. I have used Inor Temperature Transmitters in the past but this sort of unit has a dozen manufacturers and even more distributors. Temperature Transmitters take the mV signal from your thermocouple and convert it into a usable signal. For example a Type K thermocouple signal can be converted to a specific range and scaled to a 4 - 20 mA output. That 4 to 20 mA is passed through a 500 Ohm resistor giving you 2 to 10 volts which you run into an A/D converter. Using a 250 Ohm resistor you could get 1 to 5 Volts. Either way the speed of the A/D conversion (Samples Per Second) is a function of the A/D converter. Then you can run that signal or look at it in any software solution you wish, I covered this in post #3.

If you want to use the AD8495 then by all means use it. Just make sure, as was mentioned, you work within the performance curve.
The AD8494/AD8495/AD8496/AD8497 are precision instrumentation amplifiers with thermocouple cold junction compensators on an integrated circuit. They produce a high level (5 mV/°C) output directly from a thermocouple signal by combining an ice point reference with a precalibrated amplifier. They can be used as standalone thermometers or as switched output setpoint controllers using either a fixed or remote setpoint control.
You still need to run the amplified signal into a A to D and the sample rate of the A/D needs to meet your requirements. Normally liquid temperatures do not change very fast but your A/D sample rate is what will be plotting your curve.

Ron
 

Thread Starter

sidring

Joined Mar 7, 2017
7
Hi All,
Thanks a lot for fast responses. @Reloadron i will get back to you once i get the original thermocouple which i have to use but currently i am using a random thermocouple and dont know about the specification of it.

Thanks
Siddharth
 
Top