Bad Values seen when first terminal of SCXI-1102 is an open value

[labviewjunkie]

I am having some issues with sampling thermocouple channels connected to my 12-slot SCXI-1101 chassis.  Attached below is background on the hardware I'm using and full detail of the issue I'm seeing.

Hardware:

The first 6 slots of the chassis are populated with the SCXI-1102 Thermocouple Modules using the 1303 connector block.  I am using a pcie-6320 daq card that controls the chassis through slot 1.  I have a SCXI-1100 connected to a 1300 terminal block in Slot 7 and I have another SCXI-1102 connected to a 1303 connector block in Slot 8.

Issue:

I created a daqmx task which uses all 24 thermocouples of my SCXI-1102 (Slot1), 4 analog voltage readings from Slot 7, and 2 analog voltage readings from Slot 8.  I have been sampling these values and displaying them on my front panel in LabView.  All of my temperature readings look normal if there are no open thermocouples, but if I unplug the first channel connected to the first terminal of the SCXI-1102 all of my thermocouple values change by about .4 or .5 deg F.

This same thing happens if I create a task using any of the SCXI-1102 modules connected to this chassis and try unplugging the first channel connected to the SCXI-1102 to make it open.  I am using the built-in CJC which I have set for all of my thermocouples within MAX.  All of my channels have been made into global channels within MAX and my task which I have been testing with has been created in MAX and comprises a small subset of these global channels.

[scmave]

I'm a little unsure about what causes the readings to change. Can you clarify when you mean when you unplug the first channel connected to the first terminal? Does that mean the first thermocouple has only one connection to the SCXI-1102? Or are you completely removing the thermocouple from the SCXI card?

[labviewjunkie]

The thermocouples are terminated to a panel that houses the female T-thermocouple plug-ins.  I had the male thermocouples plugged into this panel at each jack and were all hanging in an outside ambient at 70 deg F.  I believe I have a working solution now after reviewing more examples and trying some different methods out.  I have shown all of my findings below.

Issue:

If I read 24 of these TC's while they were plugged into the female jacks on my first SCXI-1102 in slot 1 along with 4 analog channels in Slot 7 and 2 analog channels in slot 8 in a single daqmx task, I was having no issues as long as every thermocouple was plugged in.  If the 1st TC on the card was pulled out, then other TC's that I was sampling would change by about .4 or .5 deg F.

SCXI DAQ Sampling Trial 1:  I configured all of the TCs within MAX to be T thermocouples with a range of -40 deg F to 310 deg F using the built-in CJC.  Also, I scaled all of my analog readings to the appropriate engineering units to read Volts, Watts, Amps, and a 0-5 V signal off of the SCXI-1100 (Slot 7) and to read analog pressure readings that were appropriately scaled reading off of Slot 8 within MAX.  I configured the task to be a finite task acquiring 100 samples at 40 HZ.  When acquiring the data the analog measurements read fine, but if I removed the first thermocouple other TC's were influenced by the .4 or .5 deg F.

SCXI DAQ Sampling Trial 2: I programmatically configured my daq task.  I configured all of the TC's as analog voltages and used the following example as a guideline to manually scale the TC's: http://www.ni.com/example/29460/en/.  I also scaled the other analog measurements as well programmatically and read them into this single daqmx task.  This method worked if sampling the tc's only, but when sampling the other analog channels with the TC's in a single task the TC readings were jumping around about .3 or .4 deg.

SCXI DAQ Sampling Trial 3 (Working Solution):  I programmatically configured my daq task.  I prepended my analog global channels created in max to the create daqmx task function and then added my analog voltages.  I split this array of waveforms read and used the following example as a guideline to manually scale the TC's: http://www.ni.com/example/29460/en/.  The task is configured for 100 samples at 40 HZ and seems to be functioning fine now.  If I remove any TC while sampling there are no measurement changes and all of my scaled analog readings read properly as well.

[Kodester]

Hi LabVIEWJunkie,

Glad to hear you have a working solution. Do you have any idea why your 3rd solution isn't seeing the same .4/.5 degree offset?