LVDT measuring acceleration not position

[GerdW]

Hi alaspina,

why do you ask for a VI when you already got one?

(I joined your two threads as they apparently talk about the same VI/problem.)

[alaspina]

I don't have access to the original VI, only a screenshot of its block diagram. 

[alaspina]

I do not have the original VI program, only a screenshot of it from the paper that it was published in. 

My goal is to measure readings from a double inverted pendulum design with a force that pushes the moment arm. The method I am using the calculate the displacement is an LVDT that with flexural pivots (trust me it will work) will measure the distance the rod moves through the LVDT sensor. 

I did go through online training, but believe me, I have not found ANYTHING that explains how to measure from a LVDT sensor or any displacement reading from a sensor. If you have any material, I would gladly go over it. 

[alaspina]

Thank you for the VI, but from testing I did last night the LVDT doesn't even seem to pick up the movement of the rod through the coil. Rather, it was because the rod was making contact with the inner coil, causing spikes in voltage. 

I wish I was already at the step that I could get at least some correct measurement, but I can't even get measurements from the LVDT sensor. 

[GJMABerends]

Hey alaspina,

We will get you measuring the value of the sensor, don't worry. There are two routes you can take:

First of all in the VI (unknownVI.PNG) on the left top corner there is a sequence structure. In one of these cases (1..4) there goes a purple line to the DAQmx VI read Analog 1D Wfm NChan NSamp. There should be an enum to select the correct IO where the actual signal input is located. Make sure that this IO corresponds to how you have connected your hardware. If not then there should be an error. But because you don't realy know what the software is doing I suggest the second option.

Secondly It might be easier to start from scratch with a simple DAQmx example. I have included it into this post. Try to find the signal first, make sure that you select the right IO from the enum on the left. When you have the correct signal you can start converting the data into something useful. For this you should try to understand (read the specification sheet of the sensor) what the sensor will give back to you. F.E. If it's acceleration then you can convert it back into position. 

And last, try to add names to your posts, this makes it easier to find out to who you are commenting

[alaspina]

Hey Gert-Jan, 

Thank you, I will check it tonight and post results here. I'll make sure to add names as well. 

[alaspina]

Gert-Jan, 

I used the simple DAQmx set up you shared, and I got only a constant line with no noise. 

I selected the task in IO and the physical channels (AO for my DAQ). 

Did i miss something?

[GJMABerends]

Alaspina,

Yes you missed something. If you select an AO (Analog Ouput) and you try to read that value back then you wil get the value that you wil set to the Analog Output. You can set this value with the write VI of the DAQmx pallete. 

But if you want to read a sensor value you should use an analog input. Configured as AI in that case. 

[alaspina]

Gert-Jan, 

Yep, It was set to AI by default. Tried changing to AO but nothing different. 

[alaspina]

Gert-Jan, 

Okay I noticed the selection block all the way to the left, I changed it to the corresponding pin input. 

I now see what I have been seeing; the LVDT picks up, but the only reading I get is when I make contact with inner loop, rather than moving the rod through the loop.