Analoge inputs ELVIS - "crosstalk"

[marin268]

Greetings,

I am a student and I am trying to use NI ELVIS analog inputs to measure 3 different signals at the same time. I'm using the DAQ assistant and am getting confusing results.

For testing purposes I connected 2 function generators (sine function on one, and triangular function on the other) and one DC power supply. 

On the charts I can clearly see that the peeks of the sine functions are not the same, and the DC reading has either a sine or triangular form of a lesser AC component amplitude than the amplitude of the AC signals but of course with the added DC value.
Let's say the AC component on the DC signal looks like a triangular function. If I disconnect the triangular function, then it gets an AC component of a sine function and the peeks of the sine function become equal. If i connect back the sine function, and the AC component of the DC signal looks like a sine function again and peeks of the sine functions again are not the same.

The problem also exists in the oscilloscope application.
When using oscilloscope inputs, there is no such problem. but my assignment is to measure 3 signals, and there are only 2 oscilloscope inputs.
Can anyone please help me or at least explain to me why is this happening?

Best regards,
Marin Filipec

[Bob_Schor]

It sounds like you might have mis-wired the inputs to your Elvis, or not properly grounded something (the usual recipe for crosstalk).  Your Signal Generator should have a Ground and Signal out.  To look at three channels, you should configure differential inputs, with Signal Generator Ground going to the low Analog Input for Channels 0, 1, and 2, and the high (signal) pins going to the high Analog Inputs for these channels.

I don't know how the pins on the Elvis are arranged, but on the USB 6009 that our students use, AI0 uses pins 0 and 4 (0 for high, 4 for low), AI1 uses pins 1 and 5, etc.  Plus we tell MAX to record differentially.

Bob Schor

[marin268]

Thx for your reply.
The settings and wiring is exactly as you mentioned. Could it be that ELVIS is faulty? Or is there any other reason for this error?
I will try to photograph and screenshot everything and post it here as soon as I am able.

Marin Filipec

[A-Guz-Man]

Hello @marin268

That would be of great help!
If you still consider it's a crosstalk issue make sure to review this material: http://digital.ni.com/public.nsf/allkb/B9BCDFD960C06B9186256A37007490CD
Which also redirect you to these: http://www.ni.com/white-paper/3344/en/

Let me know if it helps

[marin268]

Here are the results. As seen the signals taken with MyDAQ do not show crosstalk, while the signals taken with ELVIS II+ do show some of crosstalk.
One of the problems is in the picture of ELVIS measurements where AI4 is not connected, and we can see that the AI4 has a DC component without any crosstalk while AI2, the connected input, has cross talk.

The settings were of course the same for MyDAQ and ELVIS, with only the sampling rate being reduced for measuring with MyDAQ.

For measuring two signals with MyDAQ the setup is as in the picture "MyDAQ connections". The upper MyDAq is of no use here, and was just connected for the later MyDAQ measuring. 
For measuring 3 signals from ELVIS I used the function generator, MyDAQ function generator and the DC power supply.

[Bob_Schor]

Based on your last Elvis picture, it looks to me as though you have wired AI0, AI2, and AI4 (differential), with all other AI channels unwired, thus presenting a high input impendance to the A/D circuitry, ideal for recording 60Hz power noise.

Bob Schor

[marin268]

Could you please elaborate?

Could you guide ne through a way of solving this issue?

Best regards,

Marin Filipec

[Bob_Schor]

The picture shows a pair of wires (Blue, Blue/White) going into Breadboard Rows 1 and 2, which are shown as Elvis AI0 + (pin 1) and AI0 - (pin 2).  Nothing goes into Rows 3 and 4 (Elvis AI1).  Orange and Orange/White go into Breadboard Rows 5 and 6 (Elvis AI2), an Orange wire goes into Row 9 (Elvis AI4+) and a Green wire goes into Row 10 (Elvis AI4-).

Is this what you intended?  Note that nothing connects AI1 and AI3, so if your program is reading them, I would expect "noise".

Bob Schor

[marin268]

AI1 and AI3 are not connected and are not read.

In the pictures we connected ports AI0,2,4 and I have shown you result of reading ports AI0,2,4. we did not read the not connected ports (AI1,3).

the same problem occurred when connecting to ports AI0,1,2.and reading ports AI0,1,2.
If you, or anyone else has the equipment involved, could you connect some signals to your ELVIS's AI ports and post back the results, so we could maybe find what is the problem with this setup or is our ELVIS faulty.

Best regards,
Marin Filipec

[Bob_Schor]

Sorry.  The year that NI released the Elvis II, our wonderful NI Sales Person convinced our Department of Biomedical Engineering to purchase a dozen Elvis I's, which were already obsolete and had no software for the then-current version of LabVIEW.  They have been useless from Day 1 ...

Bob Schor