05-06-2011 10:26 AM
I've just recently started using the Cursors function to determine specific values within a recorded signal.
Problem is when I use the Zoom tool (located within the magnifying glass icon under the timescale) to zoom in on a specific peak the cursors malfunction. When you grab a cursor handle and drag it the cursor intersection no longer follows the signal trace in the graph.
To recreate:
I'm pretty confident this is recreateable since the same behavior exists on both our machines in the lab as well as the 2011 Beta software at my desk.
Solved! Go to Solution.
05-09-2011 02:45 PM - edited 05-09-2011 02:46 PM
Unfortunately, I was never able to duplicate the behavior you're seeing. I sampled a sine wave, and no matter how I zoomed or when I displayed the cursors, etc., I was always able to keep the crosshairs following the graphed signal.
Perhaps it's something specific with the signal you're trying to measure. Could you provide a screen shot of it? Also, at what resolution are you trying to zoom to? Are you sure the cursors don't follow the graph, or are they simply not able to define the peak you're interested in because of a too-high resolution?
05-09-2011 05:33 PM
Ive attached some screen shots of the graph and the steps along the way. Also included is the project file.
I'm feeding an analog +/-10V signal through the feedthrough panel then scaling it as a load cell.
The signal I'm viewing in the graph is dbl signal from the Amplitudes and Levels step.
05-09-2011 05:34 PM - edited 05-09-2011 05:35 PM
More attachments
05-10-2011 02:01 PM
I changed your project files to acquire an analog input for the hardware I have and I still cannot replicate this issue you're seeing. Can you replicate this problem with other graphs? With other log files? With other SignalExpress projects? Or is it only with this project and this graph in this particular log file? Which version of SignalExpress are you using?
Also, does this same behavior occur with other zoom methods, or just the one you have outlined (top-left zoom icon)? After zooming in, if you right-click on the graph and disable the crosshairs, then right-click again and re-enable them, do they lock to the graph then?
05-10-2011 04:16 PM - edited 05-10-2011 04:19 PM
I created a completely new project (still using the same signal as an input). Here's some new results:
If I take the signal straight from the DAQmx Acquire step and add it to the graph I can view the cursors as expected. I can also zoom in using the zoom tool shown earlier and the cursors intersects follow along nicely.
If I pass the signal through an Amplitudes and Levels step and drag the DC signal into the graph this is where the problems start. Cursors are normal but once I zoom they no longer follow the signal. This is true regarless of which zoom tool I use.
I should also mention, when zoomed in and dragging the cursors drag handle right and left across the graph I can see the crosshairs following "something". It's not the signal line in its current zoomed scale, it's something else; as if it's following the signal in its original scale, even though the whole graph and signal line are zoomed in and scaled much larger- the cursor's crosshairs are not 'zooming'.
Hope that made sense.
When viewing signals I only use scope/strip charts, XY graphs, or the Large Display for a numerical readout. Since I cannot use the native signal from the DAQmx Acquire step to view as a scope chart or XY graph I pass it through the A&L step. As such, I will likely never use just the raw signal from the DAQmx Acquire step. It will always be turned into something I can view as scope/strip or XY graph; hence the need for the A&L Step and where the issue seems to generate.
The crosshairs do not lock onto the graph if I disable/re-enable them.
Check my signature for current hardware/software versions. Also attached is a screenshot of the MAX's device drivers list.
Thanks for your help too btw.
05-11-2011 12:22 PM
OKors, I still cannot duplicate this issue you're seeing. I've tried all of the things you've listed in your posts, but still no luck. Could you list out the exact step it takes to cause the problem you're observing? If this is a bug you've found, it could be very specific to what you're doing. In addition, you may try repairing SignalExpress 2010, as part of its installation could have been corrupted in some way. You can run a repair by going to the Control Panel » Programs and Features » National Instruments Software » Uninstall/Change » NI LabVIEW SignalExpress 2010.
05-11-2011 02:24 PM
Okay, I've discovered a few step details that may determine why you have not been able to replicate the issue on your end.
The main step which seems to make or break the cursor issue is whether or not you view the graph as a strip chart or scope chart. Changing the Update Mode to Scope Chart is where the issue seems start.
This time I simply created a 0.5hz Sine wave (Create Analog Signal step) and passed it through the Amplitudes and Levels step. Here are the detailed steps following the creation of the sine wave:
If what I've outlined above does not cause the issue try this:
With the actions detailed from above I am able to recreate the issue on both our lab PC's running identical hardware and software, as well as the cDAQ at my desk running Signal Express 2011 Beta.
05-12-2011 06:48 PM
Thanks for providing those steps! It is greatly appreciated and very beneficial. While I wasn't able to duplicate this issue exactly as you have described, I did get to a point where I noticed strange behavior with the cursors. Depending on where and how I zoomed into the graph, I could get the cursors to "think" they were following a signal, when in fact this was the incorrect signal. What it actually was was the larger zoomed out version of the signal that the cursors were trying to follow. So, bottom line as you described it, when you zoom in to a particular level, the cursors don't follow the zoomed in signal.
Also as you mentioned, this behavior is only viewable with a scope or sweep chart, as I could not duplicate it with a strip chart. This is most certainly a bug and I've filed it so to R&D. Unfortunately, there's no workaround at this time, except to use a strip chart. I'm sorry if this causes any inconvenience, but hopefully R&D will be able to resolve the issue quickly. I will try to post with any updates I have.
05-13-2011 09:17 AM
Rory,
Thanks a lot for your patience while trying to isolate this issue. I'm glad you were able so see some sort of stange behavior.
You describe exactly what I'm seeing on my end. If you zoom in far enough the cursors "think" it's following the signal when in fact its a ghosted signal left over from a previously zoomed resolution.
The workaround is to change the graph to strip chart momentarily while I use the cursors, then back again to view the live data.
This is not much of an inconvenience but I'm glad we can document it as a bug.
Thanks!