Noise Interactions between TC-2095 and BNC-2090A

[MSStateEngines]

Ok...first things first...my setup is as follows:

PXI-1050 (PXI/SCXI COMBO CHASSIS)
SLOT1 - PXI8110 (controller)
SLOT2 - PXI4110 (DC power supply)
SLOT3 - PXI6123 (s-series)
SLOT4 - PXI6704 (analog output)
SLOT8 - PXI6229 (m-series)
SCXI, SLOT1 - SCXI1102

BNC-2090A (X2) - hooked up to M-SERIES card
TC-2095 - hooked up to SCXI1102

LabVIEW 8.6, DAQmx 8.9

So on to business...I am relatively new to LabVIEW.  The issue I am having I can best describe as noise, though it may not be actually noise.  As the VI and equipment setup stands, everything seems to be functioning correctly, but no actual voltages have been read through the BNC ports.  No cables are plugged in at all with the exception of one thermocouple I have wired up to test the thermocouple channels.  The channels I have set up to use in the current VI are Voltages from the M-series card channels 1-20 and thermocouple channels 0-5,31.  When I have the VI running and plug the thermocouple into channels 0-4 on the TC-2095, there are no effects noticed anywhere else in the VI.  All remains steady.  However, when I plug the thermocouple into channel 5 OR 31 of the TC-2095, ALL voltages (M-Series 1-20) make a jump down that appears the same over all channels. This jump is 2V for channel 5 and 3-4V on channel 31.  When I unplug the thermocoupe , all returns to normal.  ALL jumpers on the front are set to Differential mode and ALL jumpers on the rear are in their factory configurations (ON for floating signals and ON for open thermocouple detection).  None have been touched.   I am thoroughly confused. 

I hope I made a bit of sense in trying to convey this issue.

Attached are a ZIP of my VI and all SubVIs and a screenshot taken shortly after a thermocouple is plugged into Channel 5 of the TC-2095.

Andrew

[ManLongAtNI]

Hello Andrew,

I have a couple of questions for you.

 1) Can you identify the DAQ device that is controlling the SCXI chassis?

 2) Do you have multiple M Series cards to attach the 2 BNC-2090A devices, I was curious on how you are reading 20 channels.

 3) Does this occur only when you plug into 5 and/or 31, or any connection on the TC-2095

 4)Can you quantify the "jump" down on the M Series channels

 5) When does the "jump" occur for the 5 or 31

Regards,

Glenn

Applications Engineer

National Instruments

[MSStateEngines]

Okay...here goes...

 1) M-Series Card -> Slot 8

 2) 1 M-Series Card only...32 Channels; AI0 is multiplexed for SCXI, I'm only using AI1-AI20 of the remaining 32 channels

 3) The voltages drop only when I plug into TC5 and TC31, jump back up when pulled out;  other used channels read normally with no jump, unused channels have no response either

 4) The jump appears to be around 2V on channel 5 and 3-4V on channel 31; I'm not in the lab or I'd be more precise

 5) The jump occurs simultaneously as you plug in or unplug a thermocouple in those channels...within a second or so

Hope this helps and thanks for the response.

Andrew

[MSStateEngines]

 to number 2) I guess I should add that yes, 2 separate BNC-2090A boards are hooked to the M-Series Card.  The SCXI is multiplexed through the backplane to channel AI0 (M-Series)

[JÞB]

MSS-

I have to say you are not making all that much sense.  Really a brief description of you're intended outcome would be nice even if it gets wordy- we can read code or text but not minds

Yet, I think I got enough from your post to point you in the right direction.  I'm ASSUMING, (oh- I Hate that word) you have a thermocouple that causes unanticipated changes to the voltages read by unterminated and "unrelated' channels on a PXI device when you connect it to certain channels through a BNC 2090A

AH-HA, check the channel set ups in MAX and the configuration of the BNC 2090A-  DO NOT forget to check the INTERNAL switches on the 2090A as well as the external ones.  Also, look very closely at the physical connectors (I once spent 3 days "debugging code" to find out some ape[bad person] had just jammed in the fine-pitched connector and bent a pin.)  GRRRRR.... If at first you don't succeed, find a bigger hammer  

That's all off the top of my head.  Check for set-up issues, post an intended outcome.

Good luck! 

[MSStateEngines]

This LabVIEW setup is for steady state DAQ for an engines lab. Pressure transducers, thermocouples, load cells, etc.

The "solution" of this problem would be to see no feedback or noise between channels at all.

Sorry about the confusion...I'm having a really tough time relating the issue.  It's way easier to just "see" happen...

to quote you with a couple fixes...

"you have a thermocouple (K-type, reads correctly) that causes unanticipated changes to the voltages read by unterminated and "unrelated" channels on a PXI device (M-Series connected to 2x BNC 2090A, using channels 1-20) when you connect it to certain channels through a TC-2095 (TC-2095 channels 5 or 31)"

TC-2095 is a SCXI device with 32 channels multiplexed through analog input 0 of the M-Series PXI card.  Analog input 0 is skipped and unused in any configuration files.

Also, all switches remain in their factory positions, both internally and externally on both BNC and TC panels.

Hope that helps.

Message Edited by MSStateEngines on 06-18-2009 10:58 AM

Message Edited by MSStateEngines on 06-18-2009 11:00 AM

[ManLongAtNI]

Hello Andrew,

Can you verify that when you connect a thermocouple to CH5 or CH31, the voltage would drop 2V or 4V, respectively for all 20 voltage inputs? If both are plugged in, is there a cumalitive drop? Are the other thermocouple readings effected, when one is inserted in CH5 or CH31? Have you tried inserting the thermocouples in other channels to simulate the same behaviour. What is your sampling rate?

Lastly, can you release the password to your vi, or post one that isn't password protected. I'm going to try to get a similiar setup to try to reproduce the results.

Regards,

Glenn

[MSStateEngines]

First off...password is msuace12345...sorry about that

- just checked and they do cause a cumulative drop

- the drop is not constant over all channels...with TC channel 5 the BNC channels appear to drop close to 3V on some and closer to 2V on others, but overall with each trial the drops are consistent for each bnc channel.

- TC Channels 0-4 read perfectly without affecting the BNC channels like 5 or 31

- sample rate is 1000Hz I believe

- thermocouple readings remain constant and appear unaffected when messing with TC 5 and 31

Attached are screenshots taken as a baseline before inserting TC5, inserting only TC5, only TC31, then both TC5 and 31.  The pictures are in Office 2007 format. 

Thanks,

Andrew

Message Edited by MSStateEngines on 06-18-2009 03:31 PM

[ManLongAtNI]

Hello Andrew,

What is the source impedance of your transducers? I'm concerned there may be leakage due to the high source impedances.  please review the following article to measure whether or not you may need a unity gain buffer.

Also, did we see  similar behaviour when we re-locate transducers 5,31 in other channels, like 0-4.

Regards,

 Glenn 

[Coq_Rouge]

Have you checked if the thermocouple into channel 5 OR 31 causes the a voltage that will saturate the input channel. I have seen it before in a multiplexed AD system. If one channels saturate this may cause trouble for all channels. The reason why is that then a operational amplifier is driven into saturation, the whole system will need some time to recover( the opamp and surrounding circuits). I some cases this may take some time. If the system is saturated with 1Khz intervals it may never recover.