simultaneo​us strain gage and thermocouple measuremen​t problem

[JG001]

I may have been in a wrong forum with this question. Hoping someone might give me some input to this problem:

I have a PCI-6221 connected to a SC-2345 (white with back connection). I also have a SCC-SG01 and a SCC-TC02 module. I’ve connected a strain gage to the SG01 module and a K-type thermocouple to the TC02 module. I can see a nice dynamic signal out of the SG01 module when I bend the strain gage, so I know my gage measurement is working fine.

Things turn ugly when I plug in the TC02 module and try to read a temperature along with my strain measurement. As long as the SG01 module is plugged in, the TC02 module reads a completely wrong temperature (My lab is not set at 190 deg C!) with a lot of noise (temperature reading jumps between 170 and 190 C). But as soon as I physically take the SG01 module off the SC2345, the temperature reading comes back to a more realistic level (25 C) and becomes very stable. I don't see any difference with the SG01 module when I plug/unplug the TC02 module.

After a lot of digging around in these forums I am thinking of two things:

1)      My thermocouple wires may not be connected properly. Currently I’m just connecting the red wire to the TC02 negative and the yellow wire to the TC02 positive. I may need two 10K resistors connecting the TC+/TC- ports to GND.

2)      This could be due to "ghosting".

I will try adding the resistors first and then try moving the temperature channel before the strain gage channel tomorrow to see if it is due to ghosting. But in the mean time if someone has any thoughts in helping me figure this problem out, I would really appreciate it. Thanks.

John

[Kyle_A.]

Hello,

There are some configurations with the TC01/TC02 modules which could cause the behavior you are seeing.  Fortunately, putting the TC01/TC02 module first in your scan list will correct the issue.  Another possible workaround is to scan your CJC channel before scanning your temperature on these modules.

[JG001]

I am in an uncomfortable position: the setup works now, and I’m not sure what I did. It started to work when I used DAQAssist to create two channels to read the data.

The TC02 channel is still AFTER the voltage channel but it works. No more cross-talk or ghosting.

When I first posted this topic, I was using the create channel, start channel, read channel, clear channel sequence. But I only connected the task in, error in, task out and error out nodes for the TC02’s create channel vi. I forgot to configure the other nodes (like thermocouple type, min, max etc.).

[Kyle_A.]

Did you create two channels to read the data and then measure from two different thermocouples?  Creating two channels effectively implements the second workaround suggestion I posted which is scanning the CJC (or in your case another channel) to eliminate the noise from the previous measurement. 

Configuring the other properties will certainly help attain a more accurate measurement, but properties like thermocouple type, min, and max are not going to eliminate noise concerns and ghosting.  I say this so that you understand the reason you are seeing a correct setup is the addition of a second channel into your scan list.

[JG001]

Oddly, no!

Please see attached screen caps of my setup and VI. There is only one voltage channel and one temperature channel in my setup. Yes, I've been trying to figure out what happened to make things work correctly all of a sudden as well.

[Hwy-Star]

I have an SC2345 connected to a DAQCard-6062 that has had severe noise problems no matter what modules I use.  I currently have 2 SCC-AI03's in slot 1 and 2 connected to 2 0-5v transducers.  An SCC-FT01 that has a 250 Ohm resistor for measuring a 4-20 mA pressure transducer in slot 2 and an SC-TC01 in slot 5.  The temperature measurement sppears to be reading the right nominal value but has +/- 20 DegF noise and the pressures have a 20 PSI offset with +/- 2 PSI noise.  This has been a typical noise floor regardless of what I plug in or try to measure.  The fact that you are telling users to adjust their scan order to help eliminate noise problems indicates to me that there are serious design issues with this product.  Sampling order should not be a concern with signal conditioning.  Especially for a product as expensive as National Instruments products are.  This Signal conditioner has caused nothing but noise problems since I purchased it.  To the point that it is almost better to use the cable and signal breakout card than the expensive signal conditioning box.  I thought the problem might be the unshielded wiring supplied with the panelettes.  I am about to go to the trouble of rewiring the panelettes with shielded cable.  But I would like to know if the NI guys have an answer for this inherent noise issue with the SC2345 in general before I go to the trouble.  Or should I just throw it away and buy something else?