PXI-6225 SCB-68 RSE differences

[punkmonkey]

Should there be any differences between connector 0 and connector 1 in an RSE measurement.

I am collecting data across 30 thermocouples. I have 2 SCB-68 boxes wired with 15 thermocouples apiece exactly the same.  I have them set up as RSE and I found great posting on how to read the CJC and calculate the temperature from the voltages. My first 15 channels on connector 0 behave exactly like they should. Connector 1 thermocouples are putting out jibberish. When I pull up the channels as a NIMax task and just look at the voltages, there is a noticible difference (more noise and an offset) between connector 0 and connector 1. If I swap cable or boxes, the difference is still on the connector 1 side. I've also swapped out PXI-6225 cards and it is still occuring.

I am assuming that the CJC in connector 0 box is close enough for connector 1 to do the calculation against.

[Speedy_Badger]

Hello punkmonkey,

I have a few questions and suggestions regarding the noise that you are running into. Let's tackle the questions first:

1. What type of cable are you using for each SCB-68 and connector?

2. You mentioned that connector 1 thermocouples are putting out gibberish, and then went on to describe the presence of noise and offset. Are you receiving signals that are changing along with the temperature or are the signals constant and completely discernable?

3. Have you tried activating the onboard CJC sensor on each SCB-68? Each SCB-68 has an onboard integrated-circuit temperature sensor which can be activated and use to regulate your thermocouple measurements.

4. Are the thermocouples connected to connector 1 in a different environment than the thermocouples connected to connector 0?

The PCI/PXI-6225 has two connectors that are not identical. Connector 0 contains pins pertaining to both analog and digital signals. We therefore recommend that you use the SCH68-68-EPM cable which is configured with the proper shielding for both types of signals. Connector 1 of the PCI/PXI-6225 on the other hand consists of pins entirely assign to analog signals. As such, we recommend that you use the SCH68-68 cable which does not contain the shielding of the EPM cable. It is certainly possible that if you are using an EPM shielded cable for connector 1 that you will experience noise and inaccuracy in your analog signals.

I would also suggest activating the CJC on each SCB-68, or directly wiring the CJC from one SCB-68 to the other. I am assuming that the document you read on thermocouple measurements describes how to do both of these tasks, but I will go ahead and summarize the steps again for your convenience.

1. The SCB-68 manual tells you how to configure the dip switch on the SCB-68 to configure channel 0 to read from the temperature sensor. You can reference this on page 2-4 of the following document:

http://www.ni.com/pdf/manuals/320745b.pdf

2. Once the SCB-68 has been configured, open up Measurement and Automation Explorer. Under Devices and Interfaces, right click on the PCI/PXI-6225 and choose Properties. In the new window that opens up, click on the accessory tab. For each connector, select the SCB-68 from the drop down menu. Once selected, for each connector choose configure and enable the Temperature reference.

3. You can now choose the Built-in CJC source for each SCB-68.

Let me know if these suggestions help at all or if answering these questions brings forth any new information and I will try to assist you further if necessary.

Regards,

[punkmonkey]

 have a few questions and suggestions regarding the noise that you are running into. Let's tackle the questions first:

1. What type of cable are you using for each SCB-68 and connector? Both are SCH68-68-EPM.

 It is certainly possible that if you are using an EPM shielded cable for connector 1 that you will experience noise and inaccuracy in your analog signals. Wouldn't shielding help keep extraneous noise down?

2. You mentioned that connector 1 thermocouples are putting out gibberish, and then went on to describe the presence of noise and offset. Are you receiving signals that are changing along with the temperature or are the signals constant and completely discernable? Need to verify better.

3. Have you tried activating the onboard CJC sensor on each SCB-68? Each SCB-68 has an onboard integrated-circuit temperature sensor which can be activated and use to regulate your thermocouple measurements? Not yet. Hoping I wouldn't need to do this. It would probably be a better measurement though if I read the CJC on connector 1 for connector 1:).

4. Are the thermocouples connected to connector 1 in a different environment than the thermocouples connected to connector 0? Same environment. In fact they are side by side.

Thanx for the response.

[Speedy_Badger]

Hello punkmonkey,

You are correct that shielding would in fact help keep extraneous noise down. It is not a question about whether or not the shielding is helpful; it is how the shielding is configured. The SCH68-68-EPM cable is designed to satisfy the needs of an M-series DAQ board. This means that it is providing the proper shielding for both digital and analog lines, including separation between the two types. For the analog signal lines, each channel is routed to an individually shielded, twisted pair wire where as the digital lines are not. Since connector 0 of the NI-6225 is the same as a standard M-series device, this cable is recommended for use.

Connector 1 of the NI-6225 on the other hand is not a standard M-series connector in that it only contains analog lines. Since the SCH68-68-EPM has some of its lines designed specifically for digital signals, these lines will then be used for the analog lines of connector 1 and will not be the best solution. The SCH68-68 however is designed specifically for analog signal lines, providing individually shielded twisted pair wires for each analog channel. This is why the SCH68-68 cable is recommended for connector 1.

It is a very subtle difference but an important one when considering the potential for noise on a line. If at all possible I would recommend attempting to use an SCH-68-68 for connector 1 to see if this improves the situation. I would also recommend attempting to use the respective CJC for each SCB-68 as it is certainly possible that the issue is related to the temperature compensation.

Let me know if either of these suggestions improves the situation.

Regards,

[punkmonkey]

Will do. Must by new cable.

Thanx.

[punkmonkey]

Got the cable and set up the system. The properties of the CJC configuration tells me that enabling the temperature reference for connector one is not allowed. How do I read the CJC in the connector one box?

Thanx.