CAN - en 
US Application Note
The Top 5 Questions About Insulation Testing
1. There are so many models of insulation tester, how do I choose the best one for my application?
Megger Insulation Testers aren't really application-specific; they can all get the job done! Instrument selection is based on specifications rather than application. Examine closely what is actually required of the tester. These, obviously, will be the prime considerations. But simple user preference, what the operator likes and doesn't like, is almost as important. Indeed, for instruments that are to be used in a broad, general manner, user preferences may be the prime considerations.
Your prime consideration will probably be the voltage or voltages at which you need to test. Budget level can then fix a band of models that can be individually sorted, or establish a focal point from which to work up (to accommodate absolute requirements) or down (if it turns out that a less expensive model will meet all the requirements just as well). Range of measurement values may also be a leading consideration. If proof testing is the application, there may be more latitude as to range, since an "infinity" reading is likely to be an acceptable standard for a "passed" test.
If long-term maintenance is the testing goal, then greater range is desirable, as extremely high values will want to be recorded and tracked while equipment under test is new. Power source is also a major consideration, as experience, preference, and specific job situations may make demands between battery, hand-cranked, and line-powered models.
After a few of these principle considerations have been decided upon, eligible models can be selected largely on the basis of preference. Such considerations as analog versus digital, computer capability, automated tests, and others, are not specific to an application, but rest on the likes and dislikes of the ultimate user.
2. What does the Megger® Insulation Resistance Tester measure?
Obviously, resistance, but this takes some further explanation. Some models can measure continuously from less than an Ohm to Gigohms (thousands of millions!). But the real focus of the instrument is high-voltage measurement at the high end of the resistance scale, in Megohms (millions of Ohms) and beyond. The low-resistance measurements (which are performed at only a few volts) serve as convenient extra features. Such functions can also be found on multimeters. But what cannot be found on any other type of instrument is measurement in hundreds or thousands of volts, to Megohm (or even Gigohm or Teraohm) values. Such values are characteristic of insulation since, by definition, it offers a high resistance to current flow. In order to pull a measurable amount of current through insulation, a high voltage is required, and it is this high output voltage that sets insulation testers apart from other types of instrumentation. The value thus measured is used as an indication of the quality of that insulation.
3. How does it accomplish this?
By a judicious application of Ohm's Law, which should come as no surprise. The tester outputs a high DC voltage across + and ? terminals, which are connected by test leads across the insulation that is being tested. Because there is no perfect insulation, some current will flow, though it will generally be on a Nanoamp level.
A high-accuracy measurement circuit is required in order to adequately sense this. The tester also senses its own output voltage, and with these two parameters and Ohm's Law, it calculates and displays the resistance.
Three further points need to be discussed. An insulation tester, also called a megohmmeter, always employs a DC voltage, because of its non-destructive nature. Repeat measurements can be made without degradation of the insulating material. One-time proof testing for Quality Control purposes is generally performed with AC, because regulatory agencies such as Underwriters' Laboratories (UL) recognize that AC is more stressful and more likely to expose production flaws.
It is important to have a well-regulated source of test voltage, one that does not load down to drop below the expected voltage when under load. Megger models provide a load graph that indicates that at resistance values commensurate with good insulation (typically 1 - 5 MW) and beyond, full rated output voltage is maintained. Inferior testers may load down well below the selected voltage, and will conveniently avoid providing this information! At low values indicative of deteriorating insulation, however, a Megger model’s voltage will drop, in order to prevent a rush of current that can be destructive and hazardous.
Test lead connections are typically made from circuitry to ground, but can be judiciously made in other configurations as long as the connected elements are expected to be separated by insulation (non-continuous).
4. What does the measurement tell me?
Fundamentally, how "good" the insulation is, and by extension, the overall condition of the piece of electrical equipment. Insulation is constantly stressed over its life. Dirt, moisture, oil, corrosion, vibration, electrical spikes and surges, mechanical stresses from pulling and tugging, and many other factors add up to deteriorate insulation. This can be catastrophic, as by flooding, but is more likely to be a slow, steady decline over time. The remaining life of the equipment could be in years or decades…or it could be about to break down at the next startup. Brand-new electrical equipment will typically have insulation resistance values in many millions of Ohms, but by the time that gets down to a few Megohms, the item will be approaching a critical status. Insulation measurements, especially if they’ve been repeated and recorded over time, enable the operator to determine where the test item is on the resistance-time graph, and take appropriate action.
5. How often should I test?
That depends, and is largely concerned with achieving an effective compromise between the critical nature of the equipment to be tested and the time available. Of course, a good bit of insulation testing is once-and-done. An electrician proofing a finished job or a technician performing an installation check on a new piece of equipment may be concerned only with the "here-and-now"…unless recalled in the event of a breakdown. But it would still be helpful to leave the test results where they can be accessed by the permanent maintenance crew. A maintenance schedule can be set up based on 3-month intervals or less, possibly even monthly, for critical equipment that is prone to failure and can cost big dollars in lost production and wasted materials. Less critical equipment can be tested on a 6-month interval, or yearly.
Information for this article supplied compliments of Megger