I have a very simple set up with two copper cylinders (30 mm diameter by 30 mm high) between which I place a thin layer (fiberglass 0.25 mm thick or mica 0.08 mm thick).
Then I plug the leads of a Megger S1-1054/2 to the two cylinders and apply DC voltages of 200, 500, 1000 and 2000 V. After having run a series of these tests I started to suspect that the resistance measured by the Megger was ever-increasing.
For example: I apply 4 minutes 500 DC V and get X Mohms and another subsequent 4 minutes at 1000 DC V and get a slight decrease. Fine. Then I go back to another 4 minutes at 500 DC V and get something like 2X. And if I run another 4 minutes at 500 DC V it continues increasing without any stabilization.
I decided to launch a 40 minutes test at 500 V to a new sample. The result was an almost perfectly linear ever-increasing resistance over time for which I find no theoretical reason.
I appreciate if someone can shed some light.
Many thanks!
Hello Ricardo de Paz ,
The answer to your question is twofold: 1) your observation is an artifact of the megger, and 2) it is also caused by the sample itself.
First, if you look in any book on meggers (see below), they all mention that the reading of the megger is time dependent, i.e., that you have to wait an appreciable time for the final reading. The book by Coombs even mentions that some meggers have built-in timers to only allow a reading to be taken after a specified time has elapsed. I am sure the manual that came with your instrument says as much. The reason for this time dependence is poorly explained, but the hand waving explanation is that the sample's shunt capacitance and the high sample resistance coupled with the high internal resistance of the megger itself leads to long RC time constants.
Second, the sample itself may exhibit very nonideal behavior such as dielectric absorption and voltage coefficient of resistance and/or capacitance. These nonideal behaviors are rarely discussed in textbooks, manuals, or reference books, which means that they are basically invisible to the technologist; the book by Coombs, for example, mentions what he calls "absorption current" on p. 26.25 without ever defining it. But these effects are very real and significant. For example, dielecrtic absorption can substantially increase the time it takes to discharge or charge a capacitor. If you are still interested, then let me know and I will provide you with some references that discuss these nonideal behaviors.
[1] H. Buckingham, E. M. Price; Principles of Electrical Measurements; Philosophical Library, Inc.; 1957; pp. 77-78.
[2] Clyde F. Coombs, Jr.; Electronic Instrument Handbook, 3rd Edition; McGraw-Hill; 2000; pp. 26.25-26.26.
Regards,
Tom Cuff
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running