why are my analog singals always noisy when I use NI hardware

[chandler]

I have an LVDT that outputs a 0-10VDC signal. When I use an agilent to monitor the signal it looks very clean. When I use any of my NI daq boards the signal is all over the place. What am I doing wrong? Am I scanning to fast? I am currently trying to use the lvdt signal to control a servo valve. My control is all over the place since the input signal is all over.

Message Edited by chandler on 10-20-2009 09:44 AM

[Norbert_B]

LVDT = ?

Are the settings between the NI device and the one from agilent the same?

If i had to guess: I'd say that the agilent device has onboard filters which are leveraging the measurement values....

hope this helps,

Norbert 

[aeastet]

There is a lot that can happen that can give you noise in your system. Ground loops, the way you connect your device to NI stuff (RSE, Differential etc...), shilding on your cables, the environment that your in, filtering. The list could go on and on. It has taken days sometimes to chase down a noise issue that I see. I have been able to achieve very clean signals using NI equipment. I have used most every system that they offer. Some are better than other but it all comes down to how well you put your system together and your understanding of the causes of noise.

If you can not get the noise to go away and you would like to use NI equipment then I would suggest that you use a digital LVDT. They work great and noise is much less of an issue than with and analog device. Also make sure that you are connecting your device differentially on the break out board. If you can I would isolate my signal to minimize the niose that you will see during use. Always connect the sensor shield to ground (but only on ne side of the device). Check your sample rates. What sample rate are you using with the agilent vs the NI system. If you are sampling faster with the NI system I would expect to see more niose.

That is about all we can do with the information that we have.

Message Edited by aeastet on 10-20-2009 11:13 AM

[hecmar.arreola]

Have you selected the right kind of termination in MAX (single-ended, or differential)?  Are the signal wires connected to the correct pins?  If you are connecting to a pin-out board, is the pin-out in some kind of a shielded enclosure, and is the enclosure grounded?  Is the agilent on battery power?  Is your NI devices plugged into the wall (the computer is plugged into the wall that is) or is it running on battery power?

Do you have an O-Scope that you know doesn't have any filters that you can look at the signal with?

Have you tried calling NI support on the issue?  I've found they are very useful and very knowledgeable about their hardware.

[johnsold]

Lots of good suggestions.

If you are using differential connections, check the manual.  Some devices require a bias current connection, often a resistor from each input to ground

Norbert: LVDT = Linear Variable Differential Transformer  = a linear displacement sensor.

Lynn 

[Norbert_B]

---

johnsold wrote:

[...]Norbert: LVDT = Linear Variable Differential Transformer  = a linear displacement sensor.

Lynn 

---

So the link in the questionmark is correct. I just wanted to point out that using abbrevations can lead to misunderstandings and therefore maybe wrong answers.....

Norbert 

[Davidek]

Hi

If you are wondering about the different connection types here's a diagram that explains it pretty well:

Best Regards

David

[LV_Pro]

What Agilient instrument are you using. If it is a DMM then it will be averaging the signal, may have additional filtering as well, they update their displays pretty slowly. If one of Agilients DAQ boxes, check what it is set for, filtering and averaging wise. Are the both connected "the same", i.e. are you making the measurement with both at the end of the same signal path, rather than one (Agilient) hooked up close to the LVDT, the other (NI DAQ) at the far end of cables?   Correlating the measurements of two different types of instrumentation can be very difficult, for a number of reasons. If you are sampling very fast, with no filtering, no averaging, you will see every noise "blip", whereas if you are sampling slowly, filtering heavily, etc., the reading can be very different, depending on what the signal looks like.