4070 measurement time mystery.

[mperu]

Hi all,

I have a DMM PCI 4070 model that I evaluate under linux (Redhat distribution). Using the example code (found in "/usr/local/natinst/nidmm/examples/c/Single_Point/DC_Volts") for single point measurement of DC Volts I have added some code to check how long it takes to get a result and I find an average value of 290ms.

The documentation says that the maximum aperture time is 100ms for this model of DMM in DC volt measurement mode, could someone explain me why I get such a long time (3 times the maximum aperture). I have attempted to disable auto zero feature and I have got an average measurement duration of 220ms. If I disable auto zero and ADC calibration I can get 160ms.

If I put the resolution to "0.0001" instead of "0.00001" the average value of the measurement duration falls down to 8ms. Why is there such a gap between the two higher resolutions?

The only modifications that I have done to the example code is to build a loop of measurements (100 in my case) around the "niDMM_Read" code and some "gettimeofday" calls before and after the measurement calls.

Thank you for clarifying my mind.

Regards,

Michel Peru

[johan G.]

Hi,
It is right that this card is prettty slow.
Did you make your test with a simple code like the one provide in the LabVIEW examples
to check if it is due to the DMM drivers or something else.

[Mark_S1]

Michel,
    I'm not sure where the help is located under Linux, but I believe these subjects should be there.
    In the Windows help file there is a subject called "NI 4070/4071/4072 DMM Measurement Cycle" that explains the different steps in taking a measurement.  Aperture time is only part of it.
    Also in the Windows help file there is another subject called "DMM Measurement Defaults" that lists the default aperture time and settling time for the different functions and resolutions.  The difference in times between "0.0001" and "0.00001" is because the first is the same as 5 1/2 digits and the second is the same as 6 1/2 digits.
    You can, if you want, override the default time using the aperture time and settling time properties (see the "Configuring Measurement Timing" subject in the help).  You will need to verify that changing those times still gives you the accuracy/resolution you need for your application.  For example, the 100ms aperture time was selected in part to provide powerline noise rejection.

Good luck,
Mark

[wildcat_600]

Michael,

The aperture time for a 6.5 digit measurement is 100ms.  However there is still Auto Zero, ADC Cal, and Settling Time that add to the measurement period.  By turning off Auto Zero and ADC Cal you were able to decrease the measurement period so that it is closer to the aperture time.  However settling time is still calculated in.  One way to reduce the settling time is to take a Multipoint measurement (without triggering or switching).  This measurement reduces the settling time because the driver knows that you are measuring the same channel over and over so very little settling time is needed after the first point.  You can also manually set the settling time by using the niDMM Property Node.  Just remember that if you do manually change settings, such as Settling Time, after setting the resolution that the resolution you set may no longer be what you get from your measurement.  This is because when you set resolution the driver calculates normal settings that you would need to achieve that resolution.  You may be able to change these settings and still achieve the desired speed and resolution, but it depends on your measurement and environment.

Here is a good article that might explain any other questions you have about the measurement cycle.  Adjusting the Reading Rate of a DMM

Brian P.

DMM Product Support Engineer

[mperu]

Hi,

Many thanks to all of you for the valuable help.

This help and the testings that I have performed have clarified my mind about the 4070 DMM.

Have good time,

Michel