Recently I got a DL9118 multi-function tester and one of its capabilities is, of course, measuring Zs. But I wondered how it worked in terms of compensating for test lead impedance (no mention in the manual) and after inquiring it was confirmed it has factory cal for the supplied leads.
Next thing was I repeated a test at a CU and got results that differed by over 10% which bothered me, so later I ran some tests using a 13A socket (safer and easier to do) but measuring the prospective short circuit current so I could use both the high-current test without tripping the RCD, and the low-current test that works for Zs on RCD circuits. Results are below for 10 or 11 tests triggered by switching the socket off/on using its own DP switch:
First run uncertainty was +/-11% using high current mode, 2nd run using low current was +/-6%, and then trying high current mode again it was +/-2%.
So what I get from this is the instrument can be fairly stable, but cycling this (little used) socket showed a reduction in variability. What is odd is it did not go towards the low peak values. Any ideas?
But what it says to me is the contact resistance when probing might well be the biggest factor in accuracy for low values of Zs. Sure it is small compared to what is needed to satisfy the installation is correct, but interesting (to me, at least).
Anyone else seen variability that concerned them?
Next thing was I repeated a test at a CU and got results that differed by over 10% which bothered me, so later I ran some tests using a 13A socket (safer and easier to do) but measuring the prospective short circuit current so I could use both the high-current test without tripping the RCD, and the low-current test that works for Zs on RCD circuits. Results are below for 10 or 11 tests triggered by switching the socket off/on using its own DP switch:
Run | High current Zs | Low current Zs | High current Zs |
1 | 0.21 | 0.22 | 0.23 |
2 | 0.22 | 0.24 | 0.23 |
3 | 0.24 | 0.21 | 0.23 |
4 | 0.24 | 0.24 | 0.23 |
5 | 0.23 | 0.24 | 0.23 |
6 | 0.20 | 0.24 | 0.23 |
7 | 0.21 | 0.24 | 0.23 |
8 | 0.25 | 0.24 | 0.24 |
9 | 0.21 | 0.23 | 0.23 |
10 | 0.22 | 0.23 | 0.23 |
11 | | 0.24 | |
Ave | 0.223 | 0.234 | 0.231 |
RMS | 0.016 | 0.010 | 0.003 |
Min | 0.20 | 0.21 | 0.23 |
Max | 0.25 | 0.24 | 0.24 |
Max-Min | 0.05 | 0.03 | 0.01 |
First run uncertainty was +/-11% using high current mode, 2nd run using low current was +/-6%, and then trying high current mode again it was +/-2%.
So what I get from this is the instrument can be fairly stable, but cycling this (little used) socket showed a reduction in variability. What is odd is it did not go towards the low peak values. Any ideas?
But what it says to me is the contact resistance when probing might well be the biggest factor in accuracy for low values of Zs. Sure it is small compared to what is needed to satisfy the installation is correct, but interesting (to me, at least).
Anyone else seen variability that concerned them?