Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

environment problem with NI USB 6218 daq

Hi. Has anybody come across a problem with the Multifunction USB NI-621X series DAQ. Our voltage measurement changes when we put a hand or part of our body close to the unit or, when we move it from where it is sitting on the floor. The floor is constructed using metal panels sitting on metal pillars (typical modern office space construction) and covered with carpet tiles.

 

Is there a static or a capacitance problem going on here? Anyone who has any suggestions or experience white said equipment/environment I/we would be greatly appreciated.

 

Antoni.

0 Kudos
Message 1 of 5
(3,248 Views)

What is the device you are measuring voltage from?

It sounds like you have a high impedance output stage and your body is disrupting the local EMF field from the power mains or it is indeed a capacitive coupling issue.

Is the input signal floating or ground referenced?

 

If you have an oscilloscope, using it to monitor the voltage from your device instead of the DAQ card may give you some insight as to cause and effect of your voltage drift.

~~~~~~~~~~~~~~~~~~~~~~~~~~
"It’s the questions that drive us.”
~~~~~~~~~~~~~~~~~~~~~~~~~~
0 Kudos
Message 2 of 5
(3,238 Views)

Hi AnalogKid2DigitalMan. We are protyping a gait mat to track foot fall in the Lab initially. The mat is constucted out of a neoprene rubber base, the switch layer is a matrix of switches constructed using two thick paper sheets with foil segments separated with a perforated foam layer. And the top protective layer is a sheet of closed cell foam. The mat is divided into columns of 16 switches, to each switch a resistor is connected. The resistors increase in value to the power of two starting from 1Kohm and so on to form a parallel resistor ladder. Each switch ladder is connected to a leg of a  balanced wheatstone bridge circuit using 51Kohm resistors. When we use a fluke dmm to measure the output voltage everything is as we would expect.

 

Antoni.

0 Kudos
Message 3 of 5
(3,235 Views)

Antoni:

 

Cool application. Since the Fluke measures OK, I would set up your analog input as a differential (this mimics the Fluke as being a floating differential device). The 6218 is a bit different from cards I have worked with before in terms of differential inputs, refer to pg 4-36 of the manual. AI0 would be the + input, AI8 would be the - input for the differential pair.

http://www.ni.com/pdf/manuals/371931e.pdf

 

Also, what is the range of the voltage from the circuit? You may want to set your card range to the appropriate levels.

 

Another approach (digitally) would be to multiplex/demultiplex and scan the contacts at a high rate, but this would require a totally different approach in your electronics.

 

-AK2DM

 

Message Edited by AnalogKid2DigitalMan on 10-17-2008 09:54 AM
~~~~~~~~~~~~~~~~~~~~~~~~~~
"It’s the questions that drive us.”
~~~~~~~~~~~~~~~~~~~~~~~~~~
Message 4 of 5
(3,230 Views)
Sorry I forgot to mention that we have set the inputs to differential and the voltage range is from minus 5 millivolts to around 9.7 volts. Unfortunately when we ground all the negative inputs to a common point it affects all bridge outputs simultaneously.
However I have found as result of read the section you prescribed a document by NI on Inaccurate Data Because of Instrumentation Amplifier settling Time. This is  something I will try as I'm sure our voltage output impedance is very high by using 51 kohm resistors in our wheatstone bridge but, I'm not sure if this accounts for a possible capacitance problem in our cables or somewhere though. So i'm going to try and rectify the impedance by using a unity gain amp to see if that will help stabilize  the low voltage readings.

Antoni.
Message 5 of 5
(3,185 Views)