Connecting sensors with floating grounds and different output ranges to NI 9205

[sharmaa]

I have to connect three sensors to the same NI 9205 which is in NI cDAQ 9172. The first is an Lucas Schaevitz Inc. LVDT +- 15 VDC excitation and 0 to +-10VDC output signal. The common of the excitation and the common of the signal must be connected. The second is Ellison Sensor Int. GS 4002 pressure transducer with 13 - 30 VDC excitation and 0 to 10 VDC output signal. The third is Data Instruments Inc Model AB pressure transducer 5 VDC excitation and 0 - 100 mVDC output signal.

All the exication voltages share a common ground as I have connected all the excitation grounds together. But apart from the LVDT whose signal common is same as the excitation common, the other two have floating grounds.

I have two questions. Can I have some channels in RSE mode and some in differential mode in the same NI 9205? I personally think it might not work but I put up the question just to be sure and which leads to my second question.

If I want to connect the above three sensors in differential mode how would I do it?

[GarPar]

Hi sharmaa-

     Yes, you can connect signals to the 9205 in both differential and RSE mode at the same time. 

     In terms of how to connect your signals in differential mode, the 9205 couples two analog input channels in a differential pair, as displayed on page 12 of the NI 9205 Operating Instructions and Specifications.

     Additionally, pages 11-16 spell out more information about how to connect your signals to the 9205 and will be a great reference for your application.

     I hope this helps.  Best of luck to you!

[sharmaa]

Thanks,

I was wondering, are the channels isolated when being used in differential mode? Because the ground/common of my output signals (from the sensors) might be at different potential (floating ground)?

[GarPar]

Hi sharmaa-

     Each analog input channel must remain within +/- 10.4 Volts of common (page 18 of the above referenced manual).

     Hope you had a great weekend!

[sharmaa]

I have a question.

I have connected a LVDT in differential mode to Ch 01. The full scale voltage of the LVDT is +-10 VDC. It has 0 VDC output at say 0 inch and -10VDC for say - 0.25 inch and 10VDC for 0.25 inch. And I have connected a pressure transducer whose full scale output is 0 - 100 mVDC as shown in Figure 5 of http://zone.ni.com/devzone/cda/tut/p/id/7113 (DIFF, Floating Signal Source). The LVDT does not have a register as shown in the figure because the input signal is referenced to building ground.

The difference in potential between the COM port in NI9205 (as shown in Figure 2 of http://www.ni.com/pdf/manuals/374188d.pdf) and the negative of Ch 01 is 2.50 VDC. So I was wondering how was I able to measure 10 VDC at 0.25 inch? At 0.25 inch wouldn't the potential have been 10 + 2.50 = 12.50 VDC? Or to put it in other way shouldn't the voltage have been 10 VDC way before reaching 0.25 inch and stayed at 10 VDC thereafter? Or shouldn't the NI 9205 have been damaged because of voltage overrun (not considering the 30 V voltage overrun for one channel)?

I had also connected a DMM to the ouput of the LVDT and both the DMM and NI 9205 were giving me the same reading (well there was a small difference of about 50 to 100 mVDC) all the way from 0.25 and -0.25 inch.

[GarPar]

Sharmaa-

     Your measurement was a differential measurement, correct?  If so, then all the 9205 looks at is the positive and negative leads of your transducer and calculates the difference between them (hence the term "differential").  The 9205 ignores the voltage on the COM signal, so it does not 'see' 12.5 volts.  All it is going to see and report is 10V.   

     The 9205 won't necessarily be damaged by having a voltage reference > 10.4 Volts above common.  That spec is the maximum working voltage.  What this means is that if your signals extend beyond +/- 10.4 V with respect to COM, we can't guarantee that the readings will be correct.  You may experience some unexpected behavior such as crosstalk or other phenomena.  We only guarantee measurements within the maximum working voltage.  Of you were to extend your measurements beyond +/- 30V, you can expect some damage to occur.

     I hope this answers your questions.  Have a great day!

[sharmaa]

Gary,

Thank you for your patience. I am a civil engineer so it might take me quite a while to fully understand what would be obvious to people working in electrical, electronics and computers. So what is the use of COM in differential mode?

What I understand when I look at 

a)  http://forums.ni.com/t5/Real-Time-Measurement-and/Incorrect-reading-with-Ni-9205-in-differential-mod...

and

b)  Figure 5 in http://zone.ni.com/devzone/cda/tut/p/id/3344

is 1) If I have a building ground referenced signal then I do not need to wire anything to the COM. I take it that COM and AI GND in Figure 5 are the same. Since the measurement system ground and the signal ground in a building ground referenced signal are the same so no wiring required to COM. Also since common mode voltage is rejected, I guess that is why I see only 10 V and not 12.5 V. Is that correct? If not then what is it that I am not understanding?

2) But if we have a non referenced signal (floating ground) then we need to wire the negative of the signal to the COM as in Figure 5 of http://zone.ni.com/devzone/cda/tut/p/id/7113   (DIFF, Floating Signal Source). And in this case the signal should be within the +- 10.4 VDC range from the COM. Is that correct?

[Seth_B.]

COM and AIGND on this module are the same.  For this module, even though COM isn't referenced in a differential measurement, it is till required because it is used as a reference for the control circuitry of many of the front end components, including the multiplexer and the ADC.  If it isn't connected, these components will not be properly referenced and it can influence your measurements.  If you know that your input signal will not exceed ±30V of ground, then you can just wire COM to ground (the ±10.4V limit is with regards to the + and - terminal of a channel).  If your measurements are outside this range, you should use bias resistors to connect your + and - terminals to ground, with values of between 10kOhm and 100kOhm.

[sharmaa]

I have a few ground referenced signals (output +-10V) from LVDTs, a few floating ground signals from pressure transducers whose full scale output voltages are 0 - 10 V, 0 - 5 V and 0 - 100 mV. The difference between ground and the full scale voltages of these transducers are well within +-30V.

So I could just connect these sensors in differential mode in the same NI 9205 module, connect the ground to the COM port and not worry about the bias resistors?

[Seth_B.]

For the floating sensors, since you will have the COM terminal tied to a ground, you'll still want the bias resistors.  This is because while the + and - terminals of the floating source may always be within 10V of each other, there is nothing preventing the two terminals from building up a common mode voltage of greater than 30V.  The bias resistors will ensure that the floating signal doesn't float too high without interfering with your signals.

In this situation, the resistors are there not to make sure the front end of the module doesn't float, but to make sure the sensors only float within the permissible range.