How to fix waveform by using DAQ?

[JasonRivers]

I get voltage waveform by using DAQ assistant. But waveform will be move. So, how to fix waveform like the picture?

[cbutcher]

A couple of questions, with different possible courses of action:

• You want to make a fixed length acquisition of data, and display all of it after the acquisition finishes. This isn't always what people want, but if it is what you're looking for, place the graph/chart outside of your data acquisition loop, and it will be updated only once when the loop finishes. This is sort-of like a single acquisition mode on an oscilloscope, except that it won't show anything until the whole window is complete.
• You want to continuously acquire data, and continuously display the most up to date data, but you don't want the position of the results to move - this is more tricky, and a little less easy to understand the rationalisation for (why does it matter if the waveform moves?). However, you can accomplish it by setting the difference between the Xmin and Xmax values to be a multiple of the period, and setting the chart update mode to be either scope or sweep. Sweep mode will display a moving vertical line showing the position of the most recent data.
• You want to continuously acquire data, but only show the first T seconds, and then be able to move the position of the graph independantly of the acquisition. To do this, simply disable autoscaling on the X axis.

If you tell us why you want the waveform to be fixed, we can perhaps give more specific guidance as to how to best achieve it. 

[JasonRivers]

I will use write to measurement file to auto save waveform data. Then, using this data subtract other waveform data. But there is a phase shift problem when using write to measurement file. So, subtraction result is different. 

[cbutcher]

So you're acquiring data from two separate sources, and you want to keep them in phase with each other? Or at least, to have some starting point where they are in phase?

How are you acquiring these data - by which I mean, from what sources and with what sort of hardware?

I expect there are several solutions to your problem, but I'm not certain trying to "fix the graph" is going to directly help you.

Perhaps you could instead record a mean crossing time as T0? (edit - I mean the time at which the waveform crosses the mean value of the same waveform)

[JasonRivers]

I use CT and VT to get voltage and current. My situation is : 

• A device open, using write to measurement file save current waveform data
• B device open, using write to measurement file save current waveform data
• A and B device open in same time, getting waveform for reference 

Using read to measurement file read two data and using add function add two data. If phase shift correct, it become A + B current waveform. However waveform will be move and cannot keep same starting point. So, I cannot get correct waveform.

[cbutcher]

I still have no real information about your hardware, but it sounds like you just need to set up a trigger for a known point on the curve (i.e. 0 volts crossing time, or mean voltage crossing time) and then trigger the waveforms at that point.

If you give some more information it might be possible to give more specific thoughts.

[niarena]

or after acquisition, you can read the data, get to the point from where you want to start to display it, discard the previous values and plot the waveform from there.