QUESTION: SE 2012 Data View Display reverts back to time graph every time

[Braincramp]

Hello,

I have SigExp 2012.

If I try to add a data view, and make it a graphic style gage, thermometer, etc., as soon as I right click it, add my TC channel, the data view immediately reverts back to a time graph.

Additionally, even if I use the time graph, after I add the signal, it will only let me view the channel as a table, a time graph or a waveform chart.  These are the only choices.

If I use a time graph or a waveform strip chart, it does read the signal correctly, but I cannot configure the data view the way I want it.

Is this a bug or am I doing something wrong?

Thank you!

[FuManWho]

Hi Braincamp,

Does this happen with any other type of input? Such as Voltage or current. If you do not have these modules, you can simulate them:
http://www.ni.com/tutorial/3698/en/

Have you also tried a repair of your software?

http://digital.ni.com/public.nsf/allkb/AC6ED75D3D93375686256E8E00245F0D

Regards,

Paolo F.
National Instruments
Applications Engineer

[Braincramp]

Paulo,

Yes, it happens with any/all inputs.

If I start a SE project from scratch, insert a new data view and use a graphic indicator (gage, thermometer, etc.), select the DAQ device & channel, the gage dissapears and reverts back to time graph.

If I right click the time graph to "display as" the only choices are table, strip chart or time graph.  I cannot make it a thermometer or a gage or any type of graphic.

I have not tried a virtual signal.

I will try repairing the software, per your suggestion, or it is some kind of bug.

---

Does this happen with any other type of input? Such as Voltage or current. If you do not have these modules, you can simulate them:
http://www.ni.com/tutorial/3698/en/

---

[FuManWho]

Hi Braincramp,

I did some testing and I found what reproduced this behavior and I believed that this is expected. The values that you need in order to use those inputs, such as a thermometer, need to be 1 sample(on demand). This is because if you have multiple data points, the data display would not know which data point to choose. This also occurs if you are attempting to put multiple channels of data on the same data view.

Paolo F.
National Instruments
Applications Engineer

[OKors · Accepted solution]

This happens when you try to change the view of a signal in its raw "waveform" format.
You must convert your signals into scalar format. To do this simply add an Amplitudes and Levels step (under: Analysis> Time-Domain Measurements). Drag that newly converted DC signal into a new chart and right click the chart to change the view.

Honestly I'm not sure why raw waveform signals are inherently limiting in its viewing functionality but in cases like this, it is.

Alternatively; depending on which devices you are using and the order in which you add signals to your DAQmx Acquire step, some signals will default to scalar without the need to convert.

see the attached screenshot and you'll see how the icons are different between waveform and scalar signals.

Hope that helps

[Braincramp]

That is it!  Thank you.

I would have never thougt of the problem being an incompatible data type, but it really is that simple.  So simple, I would have never figured it out.

In Labview, you have to change a scalar (orange or blue line) data type to a waveform (brown thick line) on purpose.  Here, the waveform data type is default.  That makes no sense.

Temperature and DC voltage from aqcuire signal is always scalar by default in Labview, which is what I am measuring from my DAQ device (cDAQ-9219)

Paolo brings up another issue:

If I want to monitor and record a chamber temperature for several days, what is the difference between continuous samples, and looping "get one" or "get n"  samples?

In Labview, you have the choice of single, n samples or continuous when setting up a channel "acquire signal".

If I select continuous, I can still connect the data type to a thermometer graphic indicator as it is a scalar data type by default.

Odd.

[OKors]

Glad that worked for you.

As far as timing settings go, Signal Express does a pretty good job of explaining the differences if you enable context help and hover your mouse over "acquisition mode".
Basically; 1 Sample (on demand) does just that, hit run and you'll get one sample and acquisition stops. Same for N Samples, only this time it will return a finite number of samples, in both modes samples returned are user-defined in the settings. Continuous Samples will acquire data until you stop.
Each of the above settings will affect how your data is being captured in a graph.

Sounds like you want continuous and maybe set your sample rate really low.
I did some thermal stuff a while back where I needed to measure temps in a thermotron chamber over the weekend unattended but only acquiring data once every five minutes or so. The rate (Hz) ended up being 0.003333333 for 1 sample/ 5 mins.

However, all that being said, scaler signals are fine for viewing but bear in mind when converting from a waveform to scalar there's some smoothing function going on, so your actual sample rate (i.e. samples logged to file) will be different than what your settings are. Example: if you're logging the actual waveform data, 1 second of 1kHz data and 100 samples to read will yield 1000 samples in your data file. Wheras if you log the scalar data, those same settings will yield only 10 samples on your file.
This can bite you if your customers expect 1kHz sampled data but you're only giving them 10. However if you anticipate this rule you can adjust the rates of scalar data to yield the desired acquisition rate; theres just a little bit of math to do before hand.

[Braincramp]

OKors,

I don't think I have ever logged temperature faster than 10Hz... Ever.

I did once have a PWSRN* tell me that I MUST sample at 10kHz, on a 64 channel vehicle set-up (all TC's), as I might "miss seeing" a thermal anomaly or temperature spike.

I just laughed and nodded, grunted and agreed with them, logged at 10Hz, and they never knew the difference.

Nyquist Theory generally doesn't apply to analog temperature measurements.

---

However, all that being said, scalar signals are fine for viewing but bear in mind when converting from a waveform to scalar there's some smoothing function going on, so your actual sample rate (i.e. samples logged to file) will be different than what your settings are. Example: if you're logging the actual waveform data, 1 second of 1kHz data and 100 samples to read will yield 1000 samples in your data file. Wheras if you log the scalar data, those same settings will yield only 10 samples on your file.
This can bite you if your customers expect 1kHz sampled data but you're only giving them 10. However if you anticipate this rule you can adjust the rates of scalar data to yield the desired acquisition rate; theres just a little bit of math to do before hand. 

---

Next, I am using a cDAQ-9219 which has a hardware maximum near 2Hz sample rate if using TC, and a bit faster if reading voltage.

9219 is the universal input module with built in signal conditioning that can read any input you give it.

Handy little module, but slow scan speeds.  It's perfect for mixed channel types, up to 4 channels, and it fits in your pocket.

So, I am logging TC at 1.5Hz anyway just for the chamber condition to add to a test report.  Not rocket science.  Slow & low resolution is fine.

*PWSRN = Person who shall remain nameless