Ni 9213 connection to T type Thermocouples

[bobby_y]

Hello
I am using NI 9213 connected to 16 type T thermocouples. The thermocouples have shielded sleevings and they are about 50 feet long. I am only connecting the + and - side of the thermocouples to the 9213 module. Everything works fine and the readings are good . But once in a while one or
Two of the channels get overloaded. The only way to clear the error is to disconnect the bad
Channel from the NI 9213 and then connect it back. Then the reading is fine again. This happens once in a while.
The com pins on the NI 9213 are not connected and the thermocouple Sheilds are not connected
To anything either.
Do you think I need to connect the COM to all the thermocouple Sheilds.
This is a major problem.
Please help
Thank you
Bobby

[MarkCG]

Usually a common mode issue. Are they floating or grounded TC? (grounded TCs have the bead tip welded to the protective metal sheath). What are you measuring with them? Is it a motor, plasma or some kind of high electrical noise device?

[bobby_y]

Hello,

The type of thermocouple is

Rugged Pipe Plug Thermocouple Probe with 1/4 or 1/8NPT Fitting.

The length is about 50 feet. We measure AIR Flow temperature through a pipe. The system is installed inside an airplane.

So there is some vibration although the airplane is on ground. As you can see the thermocouple has only 2 leads that we

use to connect to the NI 9213. So the COM lines on the NI 9213 are not connected.

Thanks

Bobby

[MarkCG]

If you look at the NI 9214 TC module manual it does talk about what situation to use the COM terminals in. The TCs are already interanlly connected to the COM terminal so all you have to do connect com to your ground reference.

"The NI 9214 also has two common terminals, COM, that are
internally connected to the isolated ground reference of the
module. A connection to COM is not necessary for most
applications. Specifically, leave COM unconnected for
configurations in which all thermocouples are floating, all
thermocouples are referenced to the same common-mode voltage,
or one thermocouple is referenced to a common-mode voltage and
all others are floating.
However, for configurations in which two or more thermocouples
are referenced to different common-mode voltages, connecting
COM according to the following guidelines improves
common-mode rejection performance. Connect COM to chassis
ground if the common-mode to chassis voltages are smaller than
±1.2 V. Otherwise, connect COM to a valid common-mode voltage
reference that is within ±1.2 V of the common-mode voltages of all
the thermocouple inputs."

The test as you described doesn't seem to have big sources of electrical noise so perhaps we can set that hypothesis aside.  It may be difference in ground levels at the two TC locations. Or it may be that vibrations are causing a momentary open TC error.

 There is a way to find this out- if you are usign the scan engine variable nodes then look at the error when the TCs are reading bad. There are separate error codes for open TC and Common mode voltage out of range. 

http://zone.ni.com/reference/en-XX/help/373197B-01/criodevicehelp/crio-9214_rt/

If you can figure out what error code is being thrown that will help figure out the issue you are having.

[bobby_y]

Hello,

We finally discovered the problem. It was the NI9213 connector. The Thermocouple wires got loose inside.

I think NI need to change the connector type.

Thanks

Bobby