Signal Source Type and Input Configuration

[chemph]

Hello,

I read the tutorials for Field Wiring and Noise Considerations but I am not sure how to connect my sensor in which configuration.

Because of my measurement takes place in noisy environment and my cables are longer than 3 meters, I think I should make a differential measurement. (By the way, is this cable ok? http://www.lappkabel.de/index.php?id=750147)

The question is what grounding type the sensor is, in which position the FS/GS switch at connection box should be and which software configuration (differential, pseudodifferential) I should change.

The sensor is a hydraulic pressure transmitter which is bolted into a metal hydraulic pipe so I think the sensor is connected to building ground.

Maybe you can help me and tell me which configurations I should use.

Another question refers to this picture: http://zone.ni.com/cms/images/devzone/tut/a/af64be30612.gif

When using a floating signal source with Diff configuration, why must AI(+) and AI(-) connected to GND? Is it possible to measure a DIFF signal without any GND connections at Device or Signal?

Greetings from Hannover (Germany)

[johnsold]

1. That cable should be fine.

2. Is the sensor circuitry internally connected to the metallic pipe fitting or is it isolated?  Is it a powered or passive sensor? Is the power source grounded?

3. The note in the upper left box (Floating/Diff) explains why the resistors are required.  The input amplifier in the DAQ device requires a small amount of current, called a bias current, to keep the internal transistors operating in the correct mode.  This current is usually microamperes or less, but if no path to ground exists for the current, the amplifier will not work correctly.

What type of DAQ device are you using? What type of sensors?

Lynn

[chemph]

I am using this sensor: http://www.sensor-technik.de/images/stories/pdf/download/a09_datenblatt_en.pdf

The sensor is powered but I do not know whether the power source is grounded or not.

Device: PCIe - 6343 with BNC-2120

The resistors shown in the upper left box are not built in in the BNC-2120 which I can switch on or off? There is a 4k99 resistor shown.

I do not understand what is wrong if in upper left box the resistors are missing and no GND signal is connected.

[johnsold]

It looks like pin 2 of the sensor and the negative side of the power source should be connected to AI Ground.  Unless you have the current output sensor, a single ended connection to the DAQ device should work fine.

The 4k99 resistor will take care of the bias currents.  If you are using the voltage output sensor, I think you will use the GS position of the switch. If you are using the current output sensor, you would need to add the burden resistor shown on the sensor data sheet. 249 ohms is often used, producing 5 V at 20 mA.

If the resistors are missing in the upper left box, the amplifier may not work correctly.  Since it appears that you do not have a strong electronics background, I am not sure how to best explain it to you.  For the internal transitors in the amplifier to operate correctly a small amount of current must be able to flow in the inputs.  To allow the user to get tha maximum possible versatility from the amplifier the manufacturer of the amplifier does not include those resistors inside the integrated circuit and NI does not put them in the DAQ device.

Lynn

[chemph]

Thank you for answering.

It's correct that an external resistor is not required when using the BNC 2120 Box with Diff configuration and Floating Signal? I found this "Analog Input Configuration" picture with a little difference in a manual from an other measurement device with missing resistor from AI(+) to AI GND but the other resistor from AI(-) to GND is still there.

When using a sensor with current output I have to change a single ended configuration? I thought ni recommends Diff configuration vor noisy environment and long cables. What happens when connecting AI(+) to one contact of 249 ohms resistor and AI(-) to the other in Diff configuration? To be sure that the power supply is not connected to ground, I can realise it by using a DC/DC converter.

[johnsold]

It appears from the data sheet that the BNC-2120 has the internal resistors so you do not need an external resistor.  As long as your signal source has a DC path through it, the second resistor will not be needed.

A sensor with current output must still have a complete circuit for the current to flow.  With the BNC-2120 the internal resistor provides a path to ground.

Whether a differential connection will improve performance depends on the presence of common mode interference.  Try it both ways to see which works best.

Lynn

[chemph]

ok I think I have to try different configurations with my sensor even if I want to understand the correct technical way and not want to choose one way because of the mesaurement result looks better 😉

I tried a measurement assembly with a 9V monobloc battery and my BNC-2120 connection box. I connected the battery differentiel to ai0, the switch is in FS and I see in MAX a clean 9,4V voltage chart. I extected this.

Now I tried to simulate a grounded sensor. For this, I connected the negative pole of battery to a metal fitting of my heater (connected to building ground). It should be the same like a preassure sensor in a metal hydraulic pipe... The result is shown in the picture. I expected the result have to be better in GS (I connected a GS signal...) mode because in FS mode I have a ground loop (or not?). But it is the other way round... What do I make wrong?

[Suse]

Hi,

try the Modus "continuous".

Best regards

Suse

[chemph]

Hi,

there is no difference. Should be any difference? I can not imagine because it is only the "time / sample amount setting" or not?

chemph

[johnsold]

chemph,

You have discovered the secret of grounds! Not all grounds are created equal.

Even though your pipe may be connected to building ground and your computer and DAQ devices are connected to power system ground which should be connected to the building ground somewhere, the voltage at the two grounds is not the same.  This occurs because the resistance of the grounded conductors is not identically zero and because some currents are flowing in parts of those grounded conductors.  By Ohm's Law current times resistance produces a voltage drop.

If you ground the battery to a ground connection at the BNC-2120 only, you probably see very little noise.  

Building grounds and power system grounds (at least in the United States) are intended to assure electrical safety, not measurement integrity. The intention is that when a fault occurs between an energized line and ground enough current will flow to trip a circuit breaker or blow a fuse.  The National Electric Code allows significant voltages bewteen grounding conductors in normal operation. I do not recall the exact value, but I think it may be hundreds of millivolts..  I think voltages up to 10 V are permitted under fault conditions.

Lynn