NI 5112 connection problem

[wang]

Hi

I have two boards. One is PCI 6250 and the other is NI 5112 boards, I have one test fixture with two ouput. One output

is connected to CB-68LP and 6250. The siganl was verified with no problem. The other output is  connected to

NI 5112 board to measure the signal like the scope. But when I connected my BNC cable from the test fixture output

to the NI 5112. The signal in CB-68LP and 6250 will go down from 0.61V to 0V all the time on the MAX test panel. I should expect to get

0.61V all the time. How come when I connected to BNC cable from the test fixture to NI 5112, the signal in cb-68LP will not correct.

[DFGray]

What is the input impedance on the 5112 set to?  If it is 50 ohms, it could be dragging your signal down.  Make sure it is set to 1MOhm (you can do this with the soft front panel) and check again.  If this is not the problem, you probably have a wiring issue.  My guess at that point would be you have the 6250 in differential mode.  The 5112 is always referenced single-ended (it has a hard ground) and the ground is pulling your signal down.  Good luck.

[wang]

Dear DFGray:

Thank you very much!

I  have set 1 MOhm input impedence on the 5112. The dragging signal down problem did not solve. So, I set the 6250 in differential mode and 5112  into referenced single-ended. The problem is resolved. I can get signal on 6250 test panel and 5112 on SFP software at the same time.

As you know, this is a wiring issue. 5112 (RSE measurement system) has a hard ground and this ground AIGND of 5112 is pulling

your signal down. Let say, the other signal coming into 6250 is also RSE mode, is this signal going to be share the common AIGND

inside the computer bus? If it is share, I don't understand how 5112 will pulling the 6250 signal down. Can you explain to me?

2) Is NI 5112 always RSE input configuration? I am not sure NI 5112. Could you tell me?

3) You told me to try using differential input mode for 6250 board.  I tried using different configuration for differential input mode. One

    is with bias resistor for floating point signal source. The other is without bias resistor for grounded signal source.

   In result of that, the one with bias resistor can display corrct DC voltage, and the one without bias resistor, the voltage signal

  went up and down from 10V to -2 V. In conclusion of that, I can say my signal is a floating signal source, right?

4) Then, I tried using NRSE input mode for the 6250 with different configuration of bias resistor. The problem is

   here. I am not sure the range of my bias resistor. Is it the same as differential. I noticed the specification which said

  the range is from 10Kohm < R < 100 Kohm. It is for low impedance sources such as thermocouples and singal conditioning

  module outputs according to the manual. I am not sure my input sources impedance is low or high. But I need to set 1 mega ohm.

  That should be high, right? If so, what is my valid input range?   Can you tell me? Is it alway safe to put equal resistance for AC and DC

  coupled input signal.

  Also, it did not say the range for NRSE. what is the valid range for NRSE?

5) Can you tell me how ground loop occur?

6) I tried to put 18Kohm resistor from ACH positive to AIGND and 18K ohm resistor from AISENSE to AIGND. I need to mentioned I am

   sure about if this is a valid range. If it is, I can get good signal. Can you tell me why we choose Differential, not NRSE?

[Ayman K]

Wang,

2. 5112 is single Ended configuration. This could be located in the specification document.

The following Tutorial explains Ground loops and different input configurations. This tutorial is very detailed and should be straight forward but if you have any questions about it, post them here.

Ayman K

Message Edited by Ayman K on 09-01-2005 01:40 PM

[wang]

Dear DFGray:

Thank you for your suggestion reading NI 5112 specification and tutorial regaring signal source and measurement system. I have

read it before. I am not sure so I would like to ask you the following question.

I  have set 1 MOhm input impedence on the 5112. The dragging signal down problem did not solve. So, I set the 6250 in differential mode and 5112  into referenced single-ended. The problem is resolved. I can get signal on 6250 test panel and 5112 on SFP software at the same time.

1) As you know, this is a wiring issue. 5112 (RSE measurement system) has a hard ground and this ground AIGND of 5112 is pulling

your signal down. Let say, the other signal coming into 6250 is also RSE mode, is this signal going to be share the common AIGND

inside the computer bus? If it is share, I don't understand how 5112 will pulling the 6250 signal down. Can you explain to me?

2) So, NI 5112 is always RSE input configuration for sure , no more NRSE or differential config, right?

3) You told me to try using differential input mode for 6250 board.  I tried using different configuration for differential input mode. One

    is with bias resistor for floating point signal source. The other is without bias resistor for grounded signal source.

   In result of that, the one with bias resistor can display corrct DC voltage, and the one without bias resistor, the voltage signal

  went up and down from 10V to -2 V. In conclusion of that, I can say my signal is a floating signal source, right?

4) Then, I tried using NRSE input mode for the 6250 with different configuration of bias resistor. The problem is

   here. I am not sure the range of my bias resistor. Is it the same as differential. I noticed the specification which said

  the range is from 10Kohm < R < 100 Kohm. It is for low impedance sources such as thermocouples and singal conditioning

  module outputs according to the manual. I am not sure my input sources impedance is low or high. But I need to set 1 mega ohm.

  That should be high, right? If so, what is my valid input range?   Can you tell me? Is it alway safe to put equal resistance for AC and DC

  coupled input signal.

  Also, it did not say the range for NRSE. what is the valid range for NRSE?

5) Can you tell me how ground loop occur? Is it only occur in floating point system?

6) I tried to put 18Kohm resistor from ACH positive to AIGND and 18K ohm resistor from AISENSE to AIGND. I need to mentioned I am

   sure about if this is a valid range. If it is, I can get good signal. Can you tell me why we choose Differential, not NRSE?

[DFGray]

In brief, ground loops occur when the reference/ground used by the source and the detector are different.  Using differential mode without the bias resistor is a good example of this.  Oscilloscopes, in general, have RSE inputs.  The 5112 is no different.  That is all it has.  The 6250 has more input types.  In this case, you can use the 6250 in either RSE or differential mode.  If in RSE, make sure the ground of the 5112 and the ground of the 6250 are the same wire.  Switching them could easily kill your signal (and is easily done on the terminal block).

The exact resistors used by your system depend on the output impedance of the sensor and the input impedance of the detector(s).  Without this information, picking resistor values in a coherent fashion is not possible.  The resistors act like a divider network.  How true you are to the original signal depends on all the impedances.  If you have them, you can easily figure out what the voltage transform is between the signal generator and your input.  I know what the input impedances to the 5112 and 6250 are.  I don't know what the output impedance of your device is.  Voltage division is also not a bad thing, if you know what it is.  It's pretty easy to multiply your results by a scaling factor.

[wang]

Dear DFGray:

Thanks for your answers.

I still have a trouble calculating the voltage between ACH0 and ACH8 for differential input of 6250 boards.

I read the specification of 6260 regarding input impedence ACH0 to AI GND and ACH8 to AI GND. It said 10Gohm.

It is very big . In order to match this impedence, my floating souce of my test is also 10Gohm. Therefore,

I assure my floating source is a very high impendence source, right? If so, I read the tutorial about the ground loop.

It said the bias resistor need to be the same value for this case. So, I put 18 Kohm for positvie terminal and

18Kohm for negative terminal. I also put 18Kohm resistor in between ACH0 and ACH8 for my DC current source

ouput of my test fixture. Lastly, I got 0.31V on my test panel. I tried to calculate this value. I don't know

the formula to calcute. Since there are bias resistor in this case. How am I going to calculate.

If I did anything wrong in the step, ple notify me. Thank you!

[DFGray]

The 10GOhm impedance of the 6250 allows you to effectively ignore it, provided your bias resistors are in the 100kOhm or less range.  However, without knowing the output impedance of your source, you can't calculate voltage division accurately.  You want the output impedance of your source to be as low as possible.  If you are using some sort of sensor, it should be listed on the specification sheet.  If you are using a voltage follower or some other amplifier, it should be on that specification sheet.  If you are using a home-brewed amplifier, you should be able to calculate it based on the circuit design and components used.  If the source output impedance is low enough, you can also effectively ignore it, as well.  But until you know it, you won't know if you can ignore it or not.  Find out what your output impedance is, then do your calculation.

[wang]

Dear DFGray:

Have you already seem my output impedence of the diagram I sent it to you?

I tried taking off one resistor across ACH0 and ACH8 and I just use two bias resistors connected to ACH0 and ACH8.

I got the voltage. Can you tell me how to calculate the voltage I got using bias resistors? I am not very sure.

Many thanks if you can reply to me promply

[DFGray]

I have never received a diagram from you.  Please try again and I will respond when I can.