Group Actions
Engagement
  • Author Author: Jan Cumps
  • Date Created: Date Created
  • Views 5222 views
  • Likes 14 likes
  • Comments 37 comments
Related
Recommended

DC Voltage Measurement Accuracy: AD584KH voltage reference

Jan Cumps
DC Voltage Measurement Accuracy: AD584KH voltage reference

Follow up on  DC Voltage Measurement Accuracy: Measuring DC voltage with high accuracy on a DMM6500 . 
I bought a little voltage reference based on Analog Device's AD584 .

image

The reference is the AD584KH. The kit supports the 4 programmable reference values. It comes with a label showing measurements on a HP 3458A

image

image

image source: analog devices datasheet

Calibration values (25° / 49% humidity):

  • 2.49974
  • 5.00203
  • 7.50113
  • 10.00277

Calibrated by http://www.haoqielec.com/wap/en/aboutus.html

Checking Keithley DMM6500

I let the reference and DMM warm up for an hour. Room temperature was 26°. Humidity unknown

image

I used Keithley's TSP script "Measuring DC voltage with high accuracy". I used that in the previous post too.
The Test Script Builder log gives a higher resolution than the 6500 display.
I let the reference settle for 5 minutes on each value before running the script and getting the measurement

CAL (HP 3458A DMM6500 round
2.49974 2.4997806153 2.49978
5.00203 5.0021082424 5.00211
7.50113 7.5012037198 7.50120
10.00277 10.002870735 10.00287

Thoughts and remarks? Post them in the comments

  • I got this board in the post today!

    image

    However, I've noticed a couple of points. I don't think it was calibrated with a HP3458A, because when I measured it, both my measurements, and the ones you observed, were slightly higher than the calibration sticker. It seems like too much of a coincidence, considering we got them from the same AliExpress seller, within the space of a week or two.

    I measured at just under 25.0 deg C, at around 24.7 deg C, and you measured above, at 26 deg C.

    These are the measurements I got on my board:

    Label              My measurement
    -------            ---------------
    2.49954            2.49963
    4.999501           4.99964
    7.39875            7.49899
    9.99823            9.99854

    The other point is obvious from the photo above, that the person doing the calibration must have had a late night! There are two mistakes. The second value has 6 digits after the decimal place whereas the others have 5 digits, and furthermore, there was clearly a typo on the third line, which states 7.39875, it should have been 7.49875 (i.e. the label is not directly printed from a measurement result, it is manually typed). Still, not bad for the price, and I shall stick my own measurements on it, and will check a few meters I have lying around!

  • in reply to ggabe

    > By the histogram of measuring AD584?

    Definitely not by the histogram as displayed in the application (although I don't know its behind-the-scenes data representation).

  • in reply to ggabe

    yeah, the PC Tools software isn't documented. I'd like to see that formula too.
    This is all I have from Renesas:
    image

    We 'd have to open our schoolbooks to refresh our signal-to-noise ratio calculations ...

  • in reply to Jan Cumps

    I’m curious how the Renesas software came up with ENOB. By the histogram of measuring AD584?