Some Notes on Insulation Testers

Many products trumpet their major points while remaining silent as to the details.

Th Some 01

By Jeffrey R. Jowett

Many products trumpet their major points while remaining silent as to the details. When selecting an electrical tester, it’s a good idea to look for full disclosure, or demand it if not offered. The devil is in the details, and overlooking them can result in an instrument that does its job poorly, has a short life in field service, presents safety hazards, adds to job time, or fails to give the user valuable features. In selecting an insulation tester (megohmmeter) some of the salient points will be discussed.

A good place to start is with the load curve. Many instruments don’t even offer one. The load curve is a graph of test voltage delivered against load. The desired characteristic is fast rise time. Insulation testers require a high degree of voltage regulation. In performing a valid test, the tester should never fail to deliver the selected voltage. Let’s examine why.

Th Some 01
Click here to enlarge image

If you selected a 500 volt test, it should mean just that–not 475, not 350. Imagine an “infinite” load placed on the tester. Now suppose the load resistance could be infinitely varied. As the resistance drops, current rises. But there’s a limit, and some serious reflection will readily reveal why. The tester should have an output resistance in its circuitry that limits current. This should be specified. Typically, it’s a couple of milli-amps. Why? Just look at Ohm’s Law, with the benefit of a calculator. An insulation tester tests insulation. It does this by applying a high voltage, enough to exploit whatever flaws or deterioration exist in the insulating material, by pulling current (typically called “leakage” current) through the insulation... where it’s not supposed to go. The tester knows the applied voltage, measures the current, and voila!... calculates the resistance. Ohm’s Law and a calculator will show that if more than a few milli-amps are flowing, the material is no longer insulation. It has deteriorated. At line voltage, 5 milli-amps are typically felt as a shock. Need we say more?

Therefore, the tester limits current to values reflective of “good” insulation, and in so doing, repeat tests can be performed on the same item without damage. This is in contrast to “high-potting”, which is recommended only as a commissioning or re-commissioning test, or in troubleshooting a known fault. So what happens next? Once the tester hits its current limit, voltage decays. Does this violate our previous assertion that the tester must put out full selected voltage? That’s where the load curve comes in.

Th Some 02
Click here to enlarge image

A well-designed megohmmeter will exhibit a sharp rise against load, so that by the point where the resistance value becomes acceptable, full selected voltage has been achieved. This can reasonably well follow the familiar “One Megohm Rule” for minimum insulation acceptance, or one megohm per thousand volts. Poorly designed circuitry will produce long rise curves. Selected test voltage is reached eventually, but at much higher values than with a quality tester. The problem with this is that the area of critical interest in an insulation test is the low end of the scale. Sure, it’s reassuring to have a commissioning test that reads in the Tera-ohms. But where you’re making the really critical decisions on what to do with equipment is on the other end of the scale, where the test item may or may not be reaching failure mode. This is where you most depend on the tester’s reliability, and if a “500 volt” test is actually being performed at 250, it can result in a false sense of security where equipment is returned to service only to produce a down-time failure.

An added benefit of current limitation is that it is the means by which simple 1.5 volt throwaway batteries can produce a valid 1 kV test. Many are skeptical of this and for good reason. When battery technology was first applied to handheld instruments, problems resulted with rapid battery drain and readings that dwindled in accuracy as the batteries faded. The wary tended to stay with the proven on-board, hand-cranked generators. A mystique developed around this technology that still exists. Many developed a “feel” for the turn of the crank that they felt told them something about the load. This can’t be scientifically quantified and should not be used as a selling point... but if the shoe fits...

There are two advantages to hand cranks that can be listed. One is fundamental and the other a bit esoteric. Hand cranks eliminate human error. Unless the instrument is literally broken, it’s ready to test. No dead batteries after a 2-hour drive to the job site. And hand cranks will continue to work in extremes of cold weather well beyond the comfort zone of AAs. Beyond that, however, batteries can give fully specified and reliable service. Again, Ohm’s Law and a calculator will show that by limiting current to a few milli-amps, a handful of AAs will indeed produce a thousand volts.

There are still some considerations worth examining, however. Look for some qualifications of actual battery performance. What is expected battery life? If an average number of tests is specified, for how long per test? If one manufacturer considers an “average” test 10 seconds and another 30, the lower number of tests may actually be the better-designed unit. Continuity tests consume more power, but some testers offer current selection or other means to limit drain. Watch out for backlight. These are a major drain on battery life. Look for a tester that has been engineered with a “power saving” backlight, or one that shuts off after a few seconds (that’s probably all you’ll need to actually observe the reading). And finally, watch out for testers that drift as batteries become low. This can result in a number of bad readings before the low battery is actually detected. A well-designed tester will give full accuracy until the battery limit is reached and then cut off. If no mention is given, ask.

A word about measurement range is in order. You pay a lot for extended range, but do you really need to measure into Tera-ohms? The answer may be yes! Newer cross-linked polymeric materials are rewriting the book on insulation response, and IEEE is actively building a data base in order to better understand what to expect from these materials. However, just because the insulation material may range into Tera-ohms doesn’t mean the test must. It depends on the purpose the test is intended to achieve. If long-term maintenance records are being kept, then yes, it’s a good idea to be able to measure all the way to the limits of the insulation. This will facilitate the establishment of a time line of deterioration, something a series of “infinity” readings cannot do. But if a pass-fail test is being performed to assure basic operational capability, a much less expensive tester will do. Nothing is “bad” at 1 Gigohm!

If you intend to use a guard terminal, be sure to check the error. The guard terminal is a third terminal that not all testers have. Even if they do, it is often ignored, like the lower gears on an automatic transmission. The guard is a shunt circuit that can be used to remove one or more parallel resistances from the measurement. It is used to “sectionalize” an item under test rather than having to test it as a whole. As an example, it can be used to eliminate surface leakage traveling through dirt and moisture from one alligator clip to the other, so that the actual effect of leakage through the insulating material is better understood. Being in parallel, the guard circuit is competing with the test circuit for the limited current supply. If poorly engineered with insufficient internal resistance, it can pull down the test voltage as indicated earlier and produce erroneous results. Be sure to look for a specification that indicates percentage error into a given load. It can be as much as 80 percent! And that you don’t want.

There are also two universal standards that profoundly influence the quality of a tester, not just insulation but electrical in general. These are the IEC61010 and IP ratings. The former relates to safety, specifying protection from arc flash and arc blast. The latter determines how well the instrument is protected from the environment. Anybody can call an instrument “safe” or “water resistant”, but these organizations (the International Electrotechnical Commission and Institute of Petroleum) force them to put a number on it.

Arc flash and arc blast are potentially lethal explosions of a test instrument when connected to a live circuit while an “event” occurs (that is to say, a high voltage spike caused by some disturbance on the line). If the spike causes the instrument to internally arc, the superheated plasma can be deadly. The IEC rating indicates how well the instrument is designed to withstand these threats. It should include a category (“CAT rating”) and maximum voltage. The “CAT rating” indicates the environment in which it can be safely used–how far “downstream” of the voltage source it must be–and the voltage rating is the maximum operating voltage of the line.

IP (often referred to as “Ingress Protection”) consists of two numbers which indicate the level of protection that the design and sealing of the casework affords against invasion of dirt, dust, water, moisture and similar foreign material. Ranging from 0 to 6, IP rating follows the simple guideline of the higher the better.

But watch out for weasel-wording! Become familiar with what the standards are actually specifying and what information should be contained. Some products specify an IP rating that applies to only part of the instrument. Is it IP54 with the lid open, or closed? Does the rating apply to the battery compartment? Similarly, the IEC rating may apply only to particular test configurations while other functions have a lower rating that only appears in the fine print.

Taking the time to go past the banner declarations can give a considerably different picture of what you are getting. Diligence in selection can guard against accidents, yield substantially improved performance, and give you more for your money.

About the Author:
Jeffrey R. Jowett is Sr. Applications Engineer, Megger, Norristown, PA. Jowett received his B.S. from Ursinus College, Collegeville, PA in biology and chemistry. He has written numerous trade journal articles, conducted training of distributors’ sales staffs and customers, and given seminars, training sessions and talks to various electrical societies [including Nat’l Electrical Testing Ass’n (NETA) and Nat’l Joint Apprenticeship Training Committee (NJATC), BICSI], and to those open to the general industry public, such as Electric West, Carilec. Member IEEE, and committee for the revision of Standard 81, Recommended Guide for Measuring Ground Resistance and Potential Gradients in the Earth.

More in Home