Logging Data using Queues

[Jack_W]

Hi,

I want to log measurement data using queues.

The queues seem to work fine with the measured voltage signal. However, in addition to the measured voltage , I want to store another column which is let's say constant 5 for the whole time the code is running.

I tried to do this as in the VI posted, but I only get the 5 every now and than in the file, all other times, it's 0. This makes a little bit of sense, as I think during one loop iteration, more than one measurement value is created (?). Not sure if that is the reason..

I hope the problem is clear

Thanks for any hints on how to solve the issue.

Regards

jack

[SteveChandler]

The number of columns in a 2D array is the same for each row. The row with the smaller number of columns will be padded with the default value. If the measurement array has five elements then the constant '5' array will be 5,0,0,0,0.

Put this in your upper loop.

[SteveChandler]

Just to be clear the build array with 1,2,3,4 is obviously for illustration and you don't need it

[Jack_W]

Thanks for the hint, that solved the issue!

However, in my actual program, not the stripped down version I posted, it almost works fine, but in my measurement file I get (randomly) rows of 0 for all columns:

99999.000000    18.007537    0.039616    0.000000
99999.000000    18.007537    0.033613    0.000000
99999.000000    18.007537    0.032973    0.000000
99999.000000    18.007537    0.044816    0.000000
99999.000000    18.007537    0.036829    0.000000
99999.000000    18.007537    0.045051    0.000000
0.000000    0.000000    0.000000    0.000000
99999.000000    18.007537    0.039986    0.000000
99999.000000    18.007537    0.057114    0.000000
99999.000000    18.007537    0.032430    0.000000
99999.000000    18.007537    0.039107    0.000000
99999.000000    18.007537    0.037108    0.000000
99999.000000    18.007537    0.034194    0.000000.

Not a huge problem, but not very nice...

Any hints? Thank you for the help.

[SteveChandler]

I'm not sure what is going on. Is the 99999.000000 value the constant, the one that was 5 in your example? If so you could put your enqueue in a case structure and only enqueue if that is not zero.

[Jack_W]

Yes the 9999 is fine, all the values are as they are supposed to be, just the row of zeros is confusing, it occurs at random points during the measurement, lets say every 20-30 values.

Strange. I could use a case structure, but I'd rather find the cause for the problem as I might "miss" one set of data.

Hm...

[Stephan_D]

I think that the problem with the zeros in your measurement file is caused by the following reason:
You use the "DAQmx Read" inside a while loop with no timing. It is possible that you want to read new samples, but there aren't any new samples. In this case, you will receive an array with zero entries from the DAQmx Read VI. In this case, you add an constant to this empty array, which will probably cause this behavior.
I would try the following solution: If the array returned by the DAQmx Read VI is zero, don't enqueue the data. If the size of the array is bigger that 1, enqueue the data like now. Also, you should add a little timing in the while loop, like 10ms or so. That should solve the problem.

[Jack_W]

Thanks for the hint.

I will give it a try and post the results. Right now, I use a "simulated device", maybe that causes a problem? Timing the producing loop with 10ms did not solve the issue.

Regards