Unity gain buffer on AI lines

[Vikash_Kumar]

Any more ideas guys? I'm still struggling with this 😞 

[Henrik_Volkers]

How does the AI1 signal look like if you are only reading AI1 (wihout multiplexing) ?

[CoffeeBreak]

Vikash_Kumar,

Based on the information provided regarding your power supply it seems that you may have a floating signal source instead of having a grounded signal source. According to the Table 1. Analog Input Connections in the Field Wiring and Noise Considerations for Analog Signals white paper you will need to add bias resistors when using a Single-Ended-Nonreferenced input configuration.

[Vikash_Kumar]

I'll try the resistors. Meanwhile here are the Multiplexed - NonMultiplexed results 

A) NRSE reading mode

A) RSE reading mode

[Vikash_Kumar]

Here is the one with the bias resistors of 90 Kohm

[Vikash_Kumar]

I was contacted over email by another application engineer. For people following this thread, I'm appending the questions and my responses to this thread.

====================  Catherine's questions  ===================================

I have a few questions. First, what value should the value of AI1 be? What is connected to AI1? Second, can you send me a wiring diagram of your system, complete with bias resistors? Finally, I saw you were using 90kOhm bias resistors. I think it would be worthwhile to test other resistances to see how this changes the results.
Best Regards, 
Catherine Bacon
Applications Engineer | National Instruments

====================  My responses to the questions ===================================

1) What is connected to AI1? :: Cylinder length sensors from SICK.com are connected to all the AI channels. Its datasheet is attached in the thread earlier if you want to take a look.

2) What value should the value of AI1 be?:: In all the plots above, the sensor connected to AI1 is untouched so it should have a constant value. The sensor connected to AI0 is made to change its output. Effectively, we are plotting -- the response of AI1, when AI0 is changing its values. In idea case, AI1 should be a flat straight line with no dependence on AI0.

3) wiring diagram of your system, complete with bias resistors

4) Finally, I saw you were using 90kOhm bias resistors. I think it would be worthwhile to test other resistances to see how this changes the results. :: I tried 11kohm, 51 khm and 90khm and there seems to be no difference at all.

 

[johnsold]

It is hard to tell from your image but it appears that you are using LM2901 for the buffer???? The LM2901 is a comparator, not an op amp! Furthermoer, its output is driven by an open collector NPN transistor with emitter grounded. The amazing thing is that you have any signal at all at the output of that buffer!

Try something like this:

Lynn

[Vikash_Kumar]

Please note, I am not using LM2901. I am using LM2902N (a low power quad OpAmp) as buffer.

Are you suggesting to try the star ground connection without the buffers?

[Catherine-B]

Hi Vikash_Kumar, 

Another method we can try to reduce the ghosting effect is to increase the interchannel delay to allow more time for settling between measuring the channels by changing the convert clock, as described in the article below. While dummy channels achieve a similar functionality, this method allows us to spread measurements out further. 

http://digital.ni.com/public.nsf/websearch/65E7445DB1AA5DC586256A410058697B

As described in the article, this can be done by using the AIConv.Rate property in the DAQmx Timing property node in LabVIEW or the corresponding C function.

To allow for maximum settling time, you should set the convert clock equal to (# of channels) * (sampling rate). So, for 80 channels and 9kHz sampling rate, you can use a convert clock rate of 720kHz. 

Best,

[johnsold]

Thank you for the clarification on the amplifier. It is almost impossible to read the part number from the image.

Yes, my suggestion is to try the star ground without the buffers. 

I still suspect that you have something connected improperly or that there is some other noise/interference source causing the problem. Do you have a motor with a variable speed drive connected to your system?

Lynn