Differential signal connections

[denslen]

Ok, so I'm still new to using a DAQ unit. I have an NI USB-6343 hardware unit and I am trying to measure multiple signals on a battery pack. My maximum voltage is 3.65 volts. In trying to maximize the differential channel count, I have sequentially added channels where the first 6 channels used differential inputs AI 0+ & AI 0-, AI 1+ & AI 1-,.....AI 5+ & AI 5-, then I added 8 thermocouples starting with physical channel ai6 through ai13. The first two thermocouples were attached to AI 6+ & AI 6- and to AI 7+ & AI 7-. After that the remaining thermocouples  are associated with physical channels ai8 through ai13. These become the real connections to AI 8, AI 9, AI 10, AI 11, AI 12 & AI 13 all of which are referenced to AI GND, which looks like a single ended measurement to me. Since the positive connections are also shared with the differential voltages I've already connected, I have the feeling I'm not doing myself any favors by having one end of a thermocouple(positive) being tied to the negative input of one of my other differential voltages.  

I went ahead and made the connections as described and the measurements look correct. Since the themocouple signals were still a bit noisy, I decided to reread the NI KB article on wiring analog connections. This lead me to think I should use a resistor from the negative Differential input to AI GND and planned to do this for all of the channels. The connections I described above do not allow me to add a resistor from the negative input to AI GND if one end of the thermocouple (negative) is already tied to AI GND.

Can anyone further enlighten me on this subject? If I separate the "single ended" thermcouples to individual channels I come up one short of what I need. I think I can get away with single ended measurements (NRSE?) for my battery voltages which would then give me the channel count I need.

Any suggestions are welcome and thank you in advance for your help. 

[crossrulz]

You are not going to get the measurements you want.  AI8 is the negative of AI0.  You really should not reuse it.  All you will do is cause an offset and get really inaccurate results.

The USB-6343 can handle 16 differential channels.  You only listed 13, so it sounds like you have enough channels (with 6 RSE channels to spare).

[denslen]

I suppose my real question here is how does the 6343 hardware actually handle the connection scheme? Is there an underlying muxing scheme being implemented that is transparent to the user?

I originally had the system wired up so that each voltage and thermocouple were connected to individual differential pairs. Now I need to add more voltage measurements. I went back in to edit the task in MAX to reassign the physical channels in the fashion I described to allow for more differential channels to be added. I just picked up in the physical channel selection process where I left off with the last thermocouple. This is where MAX forced me to select AI 16 as the first available differential pair. Is there a method to this that is better, ie: go back and start over with the process where I would add all of the voltage channels first and then add the thermocouples? Thermocouples first and then adding the voltage channels?

I find it interesting that MAX is able to distinguish that the next available physical channel can't be chosen to be a differential channel pair, but would allow for a differential measurement in certain cases but would allow for the configuration I described. Also, why doesn't MAX force the issue as you add channels. In other words, I add the 6 voltage channels first, all configured for a differential measurement. I then add 8 thermocouples and select sequential physical channels. The first two channels are added as differential channels AI 6 and AI 7 (physical connection terminals 23, 24 & 26, 27) and then the remainder are automatically chosen to be the connections AI 8 (terminal 2) referenced to AI GND (terminal 3) through AI 13 (terminal 21) referenced to AI GND (terminal 22) instead of continuing with the remaining differential pair connections AI 16 through AI 20.

I'm willing to accept the channel limitation if I have to but I also want to mention that even though you state that it won't work and that I will "cause an offset and get really inaccurate results", I seem to be getting the measurements I would expect. I still have some doubt though and wonder about the temperature measurements for the "cross channel" connections (AI 8 being shared) being slightly higher than the two differential thermocouples.

Any further help on this subject would be appreciated. Understanding this better would have saved me time and trouble hooking up all of the connections only to find out it may not work.

Thanks.

[crossrulz]

I recommend setting up all of your differential measurements and then add the RSE measurements with what is left.

I'm not sure I fully understand your system.  This might be one of those I just need to see it in order to figure it out.

[_Tim]

There are some useful guides on ni.com that explain the fundamentals of data acquisition and working with signals. They might prove helpful as you set up your system.

http://www.ni.com/white-paper/4523/en