thermocouple DAQ USB 6008

[heang]

Hello

I am novice of LABVIEW and DAQ system. If now i am going to use thermocouple, will it be suitable for me to use USB 6008 to acquire and analyze signal. USB 6008 has no filters inside. Using labview, will the noise be filtered out by LABVIEW itself using FILTER vi?

if signals are going to be analyzed continuously( it will measure ambient temperature with sudden hot air is blown against thermocouple intermittently), how would u recommend me? how can i take only signals with hot air and disregard the ambient temperature( as differentiate the hot air from ambient air).

[johnsold]

Thermocouples put out microvolts to millivolts based on the temperature difference between the junctions. The USB-6008 has 12 bit resolution (differential) and the most sensitive range is +/-1 V. This means that the resolution is about 0.5 mV. This is about 10 degrees for a type J thermocouple.

The USB 6008 is really not very suitable for a thermocouple unless you have an external amplifier and cold junction compensation.

The filtering question depends on what kind of noise you expect and how large it is compared to the desired signal. Almost always it is more effective to filter earlier in a process than later.

Lynn

[heang]

hello
thanks for the prompt reply
here questions i am going to raise
1) if the the range of temperature is to be measured by thermocouple is 60 to 150 celcius, the range of input voltage should be in mV as well. What my question is the maximum senstivity of USB 6008 can only be up to  +-1V?sorry, i search thoroughly the specification, it is not written.

2) if USB 6008 is not suitable, can u recommend me the neareast application with cost effective, without having the need of using amplifier and cold junction measuring. thanks


regards

[Brian_T]

heang,

The smallest usable input range for the USB-6008 is in fact +-1V.  You can see this in this KnowledgeBase article as well as in the product manual on page 26.  As you can see in the KnowledgeBase article, the USB-6008 can only measure changes of about 9 degrees Celsius, making it not the best device for measuring thermocouples.

For a more precise alternative check out the NI-9211, which is designed specifically for thermocouple measurements.  It has 24-bit resolution and a +-80mV input range.

Regards,

Brian T
Applications Engineer

[smercurio_fc]

Duplicate discussion.

[teresadmij]

Hi to all,

 

I don't understand very well whats mean this that says johnsold:

 

"The USB-6008 has 12 bit resolution (differential) and the most sensitive range is +/-1 V. This means that the resolution is about 0.5 mV."

could someone explain to me please?

I am trying to read sensor Lm 35, through the USB6008 and get a signal with noise near .005 V. In fact when I read any signal I have this noise level, can you advise me?

Thanks, greetings.

[RavensFan]

12 bit resolution means that you can measure something and have 2^12 distinct values,  or 4096 distinct values.  In a range of +/- 1 volt, (2 volt range) the smallest increment you can measure is 2/4096 or .48 mV.  which he said is about equivalent to a 10°C range for a type J thermocouple.  So you'd need a 10°C change in temperature before you could see it.  Thus your temperature resolution is 10°C.

You say you have 5 mV noise, so about 10 times the resolution.  So your noise is going to make your signal appear to jump 100°C.

That is why you need special signal conditioning modules to be able to read thermocouples.

I'm not familiar with an LM35.  If it it has a higher sensitivity, more voltage change per temperature change, then the amount your signal jumps due to the 5 mV noise might be less.

[teresadmij]

Thanks for answer Ravens Fan,

The sensor LM 35 gives 10mV/°C, therefore  i believe that it shuld not have any problem of accuracy, however on the graph the read signal apears some small peaks of 5mV. Another signal that I'm interested it is in milivolts, therefore it is important that it has the least noise possible.

What can I do to get smaller signals with less noise through this card? may be with some digital filter or a filter VI? Could give me some advices?

thanks a lot, you have a god day.

[RavensFan]

I'd say that if you have a signal that doesn't change too rapidly, and temperature signals generally don't, then using a digital filter or VI.  Or just simply collecting 100's of data points a second and averaging them will give a good temperature reading for that 1 second period of time.

[teresadmij]

Thanks for everything Ravens Fan, i understand, but where i can find an example of digital filter? do you have one that you can send me please?

Thanks in advance, a greeting.