DMM 4071 accuracy @ 4.5 digits

[Brandon_G.]

Sorry about that John,

Attached is a screenshot and the raw data.  The raw data is actually a .lvm file but you can view it as a text file as well.

Thanks!

[BKendall]

I am using the 4071 to measure current and have a couple of questions on how to calculate accuracy.

I am using the 4071 in digitizer mode @ 10Ks/S. I believe this gives me about a 5 bit resolution.

The signal I am measuring is arround 3 microamps. I am trying to determine accuracy on various ranges.

Questions:

1) Does the discussion in this forum referencing voltage measurement apply equally to current measurements?

2) I calculate that based on 10 Ks/S i am getting about 5 bit resolution. Is this correct?

3) The chart on page 4 of the spec says the values apply to 6.5 digit measurement. So what are the numbers for 5 bit accuracy?

Thanks.

[Michael5000Watts]

1) Yes, this discussion also applies to current measurements.

2) It is important to note that the 4 1/2, 5 1/2, 6 1/2, 7 1/2 digit resolution specifications listed are benchmarks that we guarantee certain specifications based on calibration timeframe. Any point between these points are not specified and must be approximated. There is no formula to calculate the accuracy in between these points. 

That being said, below you will find an excerpt from page 1 of the NI 4071 specifications. Based on this table and graph, you will be able to achieve approximately 5 digits of resolution with a digitizer sampling rate of 10 kS/s. Keep in mind that these resolutions are in digits, where the resolution in bits is higher. 

3) The chart that you are referring to is specific to 6 1/2 digit resolution, which is why we must consider the additional noise errors for current. Notice that the Additional Noise Errors for Current table does not include an entry for 6 1/2 digits, because the DC Current table is based off of 6 1/2 digit resolution. 

I hope this explanation helps with your questions.

[BKendall]

Thanks for the information. So this is the calculation as I understand it.

I'm measureing a current of about 5 µA with a range setting of 100 µA. I am sampling at 10Ks/S.

So I assume I will have about 5 digits of reolution. Tcalculate accuracy I:

1) Take the numbers from the 2yr cal column which are 100+20.

2) I am allowing for a 10 degC swing so I take the nubers from the temp coef column of 10+.5 times 10 = 100+5.

3) I add the additional noise of 30 (times the range) (0+30).

So I have 200 (x reading) +55 (x range) = 200 e-6 * 5 µA + 55e-6 * 100 µA = 0.0065 µA.

Does this look right?

Barry

[Michael5000Watts]

Hi Barry,

Based on the information that you have provided, I agree with your calculations.

[BKendall]

Thanks for your help.

Barry

[Erro11]

“Notice that both scales are logarithmic. 0.003PLC is less than 1ppm of the range, but for simplicity I am going to use 1. “

The scales may be logarithmic but 0.003PLC doesn’t appear to indicate a noise ppm less than 1. It seems to be approximately 6. (See Below)

So 6 x 6 would then appear to be 36 ppm of range; correct or not correct?

[Erro11]

if 6 is the correct conversion from rms to pk to pk, the 6 x 36 ppm of range would be 216 ppm of range (pk to pk).

But isn't the conversion from rms to pk to pk

?                                                                                                                                                                                                                                                                                                                                                                         

[Erro11]

Appearantly NI has change the aperture time specifications in the help of LabVIEW 2018.

The aperture time has changed from 50 us in 2012 to 20 us in 2015. (More noise.)