Measuring open circuits

[Lama]

Hi all,

Any suggestions on how to solve the following problem would be greatly appreciated.

I am making DMM resistance measurements, 2 wire, routing the measurement points through two switches (configured as multiplexers) and comparing the results against a set of expected results, using the Comparison Express VI, set to equal within tolerance.

This works well until I measure open circuits. I need to confirm that some points are open circuit, some points are directly connected and other points have a resistance between them, of up to 15 Kohm.

My open circuits get reported as ‘NaN’ (Not a Number), which doesn’t work with the Comparison function. I tried fitting a 1 Mohm resistor across the DMM terminals to bring it within the DMM measurement range, but the only ways that I have found to get a number reported back, rather than NaN is either to allow the DMM to autorange, but this looses synchronisation with the switches and so isn’t acceptable, or to set the DMM range and resolution for 1 Mohm, but that results in too poor a resolution for the low resistance measurements.

Is there a simple way to convert my NaN readings to a large number before passing them to the Comparison VI?

Or is there a better way of approaching this?

Thanks,

Mark

[Jorge_Amaral]

i can do you a small help. it's show you how to detect NaN event. And show what to do in this case.

See the example attached

[Lama]

Thanks Jorge,

I wasn't aware of the 'Not a Number' comparison vi, what you suggested was exactly what I needed.

Lama

[johnsold]

Lama,

If you know the expected range of values for each part of the test, can you programmatically set the DMM range? Set it to 200 or 2000 ohms for the short circuit measurements, 20000 or 200000 ohms for the resistance measurements and 20 megohms for the open circuit test. Setting the range probably takes about the same time as switching the multiplexers and avoids the autoranging issue.

You probably still need the NaN test which Jorge mentioned since an open could occur on any test if the device is defective or the mutliplexer fails to connect properly.

Lynn

[mikeporter]

The thing I'm wondering about is more basic: Why is the measurement code returning a NaN in the first place? Open circuit is a valid measurement condition and so should return a valid number.

Mike...

[johnsold]

Mike,

Good point. I assumed that the overrange result from the DMM was being converted to NaN by the driver.

Lynn

[Lama]

Unfortunatly I have a further complication.

Because my readings are generated via switches running through scan lists, my results are in the form of a 1 D Array, but neither the Select VI, nor the Search and Replace VI will accept an array as an input. Nor can I convert an array to a string.

Do I have to use the index array function in a loop to go through each element of the array to lok for, and change my NaN's?

Thanks.

Lama

[johnsold]

Lama,

Use the autoindexng feature of the for loop. Look at the help files and the LabVIEW tutorials for more information about these elementary LV capabilities.

Lynn