RoadTest: Megger Insulation Tester
Evaluation Type: Test Equipment
Did you receive all parts the manufacturer stated would be included in the package?: True
What other parts do you consider comparable to this product?: Megger has other insulation testers too as well as multi-function testers. Keysight has an insulation tester too, as does Fluke and Kewtech.
What were the biggest problems encountered?: null
An insulation resistance tester is a kind of ‘specialist’ product that is used within certain engineering fields.
As an electrical and electronics engineer, I’d not had the chance to use such a device. However a family member is a qualified electrician for commercial installations so to him this device was extremely familiar and essential for that engineering discipline. He is responsible for the design, commission and installation and testing of electrical wiring within buildings.
In the UK (and worldwide) Megger is a very well known brand; a short history of Megger is on their website. For the domestic and commercial electrical wiring installation market, Megger instruments are the market leader.
I was keen to examine the product and investigate some different use-cases. Other fields of engineering that would require insulation resistance testers would include those involved with wound components such as motors and transformers. I was also interested to see if it could be useful for antenna feed line testing, such as coax cables used for connecting television antennas. In all these cases, the point of an insulation resistance meter is to determine if the insulation (between different windings, or between a wire or winding and a metal chassis) is sufficient for correct and safe functioning of the device or system under test. The measurements are ultimately used to provide visibility, to make a determination about what has occurred or what is occurring to the state of the insulation.
Many handheld multimeters cannot measure beyond 100 megohm. Unusually, the Keysight U1282A reaches 600 megohm. An insulation resistance meter can measure into the gigohm or hundreds of gigohm range; some models reach into tera-ohms!
The reason a multimeter cannot measure such extremely high resistance is because the measurement is performed with a low voltage (often a 9V battery inside the multimeter). The multimeter circuitry provides a constant current, and it will measure the voltage across the probes in order to calculate the resistance. With a current source driven from a 9V battery, very minute voltage differences would need to be measured in order to have very high resistance measurement capability, and there is a limit to the multimeter ADC resolution. Furthermore for wiring and systems where a higher operational voltage is used than 9V, it is important to test at these higher voltages because the insulation resistance can change at higher voltages. An ordinary multimeter will not ‘see’ this and will just report an open circuit.
So, the solution is to use a higher voltage, and that is what an insulation resistance meter will do. Internally it will step up the voltage to hundreds or thousands of volts. It does mean that an insulation resistance meter should not be used on electronic circuits. Components could fail with the high voltage. Insulation resistance meters are intended for passive equivalent-circuit use-cases. This also explains why it is known as an insulation resistance meter, instead of an ohmmeter. An ohmmeter such as that found in a normal handheld multimeter is not likely to destroy unpowered electronic circuits that it is applied to. Because of the high voltages, a device like the MIT420 should only be applied to scenarios such as unpowered wiring that is disconnected from electronic circuits, or an unpowered motor or transformer. Therefore the following question and answer only applies to unpowered wiring and windings:
The short answer is ‘no’. Performed correctly, an insulation resistance test won’t cause any damage. This is because the voltage from an insulation resistance tester is lower than the typical ‘breakdown voltage’ of insulators. A different instrument is used to perform breakdown tests where the insulation is overcome and sparking can occur.
The MIT420/2 provides quite a selection of test voltages, spanning 50V to 1kV, but also a configurable custom voltage level can be chosen too. The reason for selecting different voltages is explored further below; it can help in determining the condition of the insulation!
From a bird’s eye view, a conductor looks like a very low fixed resistance, and an insulator looks like a very high fixed resistance. However if electricity is applied to a lightbulb, one soon learns that the current changes over time as the conductive filament heats up. In a similar vein, in a close-up view, an insulator cannot be modelled as just a very high fixed resistance either.
At first thought it could be expected that insulation could be modelled as a very high resistor, perhaps approaching a near-infinite value with a good insulator. However, moisture or dirt on the surface of the insulator provides a lower resistance path between the wires or windings; this is known as surface leakage and it can be modelled as another resistor in parallel with the first one.
However, two windings or wires that are separated by an insulator also behave as a capacitor. If the insulator has dielectric properties then it behaves as a higher value capacitor. So, the model begins to look like the two resistors are in parallel with a capacitor.
Digging deeper, as mentioned, it can be seen that some insulators will be dielectric, i.e. they will contain molecules that can be aligned when a voltage is applied across it. This allow additional charge to be stored, and this is how capacitors can have an increased capacitance compared to an air spaced capacitor. But, there is also a delay before the molecules fully align (or un-align when discharging), and this is known as dielectric relaxation; it can be modelled as an RC circuit, i.e. a resistance in series with a capacitance. The effect of this dielectric relaxation phenomenon is known as dielectric absorption; when a capacitor is charged, there is some additional small current that continues to flow until full alignment, because that alignment process takes energy.
All of a sudden, the model of an insulator between wires or windings now looks a lot more complicated!
To compound this, there are other factors that could cause a variation in the measured resistance. In particular, temperature and humidity can affect the reading. As temperature is increased, the measured resistance decreases (almost logarithmically if temperature is on a linear axis) with the slope value depending on the insulation material.
The MIT420 can apply a high voltage (configurable, up to 1kV) to the item under test. Typically one connection would be made to one conductor, and the other connection would be made to a second insulated conductor or to the chassis or ground, depending on the equipment under test.
A measurement could change over time because of the capacitances in the model as discussed earlier. The resistance that is measured could also be a function of the temperature and humidity. This all leads to the question, what use is a precise measurement if there are so many variables?
The answer is, that often the precise measurement isn’t as important as the trend over time, under the same temperature and humidity conditions ideally. This would reveal any degradation over time.
Is an insulation resistance measurement useful as a one-off measurement, where historical measurements have not been recorded? Again the answer is ‘yes’, but it can be even more useful to examine the slope of resistance over a short period of time. The test is known as a ‘dielectric absorption ratio (DAR)’ or ‘polarization index (PI)’ measurement. They are both very similar, and essentially measure the resistance at one time, keep the voltage applied so that the molecules in the dielectric continue to align, and then another measurement is taken. The ratio between the measurements indicates the slope. A shallow slope result where the second measurement barely increased, could indicate that there is leakage caused by moisture perhaps.
The PI test usually takes a measurement after one minute and after 10 minutes. The DAR test is shorter and takes a measurement after 30 seconds and after 60 seconds. Such dielectric related tests are useful, but are not always applicable. For instance if the measured resistance is of the order of many gigohms such that the measurement is at the limits of the ADC within the test instrument, then even slight humidity and temperature changes can cause enough variation to produce significant inaccuracies in the ratio measurements.
Another factor is the voltage used for the measurement; there are guidelines for which voltage to test at, but generally it shouldn’t matter significantly; the measured resistance may decrease slightly with an increase in voltage, but if there is a significant difference then it could be indicative of faults with the insulation such as moisture or dirt within cracks in the insulation.
In a commercial/industrial scenario electricians will keep a record of the insulation resistance each time the equipment is tested (for instance yearly). It provides for good preventative maintenance to have this record and see if the condition of the insulation is changing over time.
The Megger company (and other companies) also produce multi-function testers. They are needed to perform the tests necessary for residential and office and industrial electrical installations. They provide a multitude of functions including loop testing and RCD testing as well as insulation resistance testing. The MIT420/2 has some differences however. The measurement range is massively different; the MIT420/2 will measure all the way up to 200 gigohm whereas a multi-function tester will only reach 1 gigohm. The MIT420/2 is almost half the weight and is lower cost. It is ideal for first fix teams that are measuring the insulation resistance and confirming that there are no issues with the cabling before the second fix team does the extra tests. It is also more affordable for home and automotive use to test things like iron elements or fan/heater components.
Over the decades different materials and construction techniques have been used in products such as motors and transformers. This means that insulation resistance, and dielectric absorption differ when examining older products compared to newer products. In a used device where there is some dirt and moisture a value of the order of gigaohms or higher could be good, but a value of the order of tens of megohm doesn’t necessarily mean the product is faulty either. The standards (in particular IEEE Std. 43) provide some guidelines in the absence of specific detail from the manufacturer of the device in question. As an example, for older (pre-1970) equipment, one guideline is that the minimum insulation resistance (in megohms) should be equal to the rated voltage (in kV) between connections on the device, plus 1. So, an old device rated at 1kV should have a minimum insulation resistance of 2 megohm. The same 1970 device may well measure at 100 times this value or more however; it depends on the condition of it. As mentioned earlier, the trend is more useful, using historical measurements.
Measurements that take into account the dielectric absorption (i.e. PI and DAR measurements) are also helpful because they can reveal if surface leakage is having a significant impact on insulation resistance, by observing the slope of insulation resistance change.
The MIT402/2 looks like a large handheld multimeter (it is about the same size as Keysight’s U1282A multimeter). The plastic has had a 2-shot injection moulding process with a softer black grip surround. It certainly feels ready for a construction site or industrial scenario.
The MIT402/2 comes in a hard carry case along with a CDROM, calibration certificate and a 1-page A4 quick-start guide which is worth holding on to and possibly laminating. Another 1-page leaflet indicated that the unit comes inclusive with a 3-year warranty if registered within one month of purchase. There are test leads/attachments supplied too, which are detailed below.
The carry case is thoughtfully designed; it looks tough, hinged (including hinged clips) instead of thinned plastic, and has an internal pocket for storing the CDROM and paperwork. There is lots of room for the test leads and there is a small recess next to the handle, where the pointy tips for the test leads could be stored.
The MIT420/2 has an IP54 rating and that is really useful; earlier non-IP54 devices often have a thin layer of dust trapped between the transparent window material and the liquid crystal display from work in dusty construction environments such as the example below which is from an older generation discontinued tester from Megger.
The liquid crystal display quality on the MIT420/2 is quite reasonable, it seems on par with previous Megger models. For a clear view it does have to be viewed pretty much face-on however. Contrast is rapidly lost when viewed from above, but it is still legible, especially when the backlight is switched on. The backlight is white and consistent across the entire large display. The buttons and dial are not illuminated. The well-spaced buttons have a hard rubber type of feel and have a nice click action. However the software debounce seems excessively slow; I’m used to pressing buttons quickly, but that doesn’t work with the MIT420/2. I have to consciously slow down and be more deliberate with the time I hold buttons pressed for.
The unit is constructed really well, I could not find any fault with it. The shape is nice, it won’t easily slip even when wearing protective gloves.
There is a wide hole on the rear of the enclosure for passing a strap or loop of cloth material, although no lanyard/belt strap or anything similar was supplied. A 1-inch wide strap up to about 3mm thick would fit comfortably.
The unit comes with three test leads (category III 1kV and category IV 600V rated) and some attachments. I can honestly state that I’ve never seen such nice test leads supplied with an instrument before! They seem near-faultless. The cable (1.2m long) is very lightweight and soft. The probe handles have a slightly oval cross-section that is comfortable to hold (there is a gray rubber grippy area too) and the other end has fully shrouded banana plugs of course. As well as being ergonomic it all looks really stylish too. I can’t over-state how awesome they are.
The tip end is interchangeable. Two options are supplied; normal pointy probe ends, and also a couple of alligator clips. All these attachments have the same ratings as the test leads.
The interface is based on 4mm banana plugs and sockets, so you could use the alligator clips with normal test leads too. All of the sharp pointy attachments come with a tiny protective boot which you will lose.
When connected to the supplied test leads, there is a very strong attachment between the alligator clip and the test lead. It won’t come off accidentally.
The test leads are improved from previous generation test leads supplied with Megger instruments; the example in the photo below does not have separable probe tips and the alligator clips push on top of the entire probe which might seem convenient but has the drawback that the combination is easier to damage if high current passes through them accidentally; they get damaged at the interface between the probe tip and the alligator clip. The newer ones supplied with the MIT420/2 would not have that problem and are more robust.
In summary I was so impressed with the two main test leads, I want to buy another for one of my multimeters too, and I wish Megger would manufacture normal banana patch cables too, for use with other test equipment.
A third test lead is also supplied; it is known as a SP5 Remote Switched Probe. It basically extends the ‘Test’ button function from the front panel of the meter onto the probe handle. The cable is a lot thicker (and far less supple) but it is what it is, and could vastly improve safety in certain circumstances where one cannot easily press the button on the main unit. The design and ergonomics are again superb with this remote switched probe. It is Category II rated until a supplied extension is added to the tip, and this allows it to reach the same ratings as the other test leads. There doesn’t seem to be a standard for remote switched probes yet, which is a shame; I can’t use the Keysight U1282A remote switched probe with the Megger or vice-versa. However, the supplied probe is backward compatible with other Megger testers.
There are industry-specific techniques that are followed as good practice when using insulation testers. These are designed to provide accuracy, repeatability and safety while working with the equipment. An example of a technique is to try to perform measurements at the same temperature as the historical measurements. In some industries this is not possible, and a formula or look-up tables can be used.
The safety aspect is important because some of the tests run for long periods and if the insulation and wiring forms a high capacitance then a lot of energy can be stored. It doesn’t need to directly kill; a slight shock can cause one to become startled and fall off a ladder or bump one’s head against a sharp corner in a wiring closet and pass out close to electrical wiring. Protective gear and procedures are important. Incidentally, this story highlights what a dangerous business it can be.
From a technical point of view, all that is required to perform the insulation test is to connect the test leads, turn the dial to the red area to select a desired test voltage and then hold down the yellow ‘test’ button for as long as you wish to run the test. While the button is held down, the display will indicate the measured resistance. The value will change over time if there are any significant dielectric absorption effects. Incidentally the simulated analog display is really nice; the dial has a logarithmic scale that spans five decades so that measurements are not always swamped into one end of the dial. It took a few minutes to get used to, but now I really love it for quickly seeing the approximate amount of insulation resistance really quickly.
From a practical point of view, before the test is run, one may wish to confirm that the wiring is unpowered and connected correctly. It would also be useful to check the capacitance. All of these things can be done with the additional functionality present in the MIT420/2. These accoutrements comprise of voltmeter, ohmmeter and capacitance meter. Just to be clear, the MIT420/2 will never replace a multimeter, but will complement it. The voltmeter/ohmmeter and capacitance meter features in the MIT420/2 are designed for the purpose of preparing for insulation resistance tests.
Having said that, the voltmeter is actually quite good, offering capabilities not seen in some general-purpose multimeters. Cleverly it measures a combined ‘AC+DC RMS’ by default, and it can be changed to measure RMS or DC by pressing a button.
Typically multimeters will measure either AC or DC. The AC measurement (see the blue shaded area near the position marked ‘A’ in the diagram above) could be an RMS measurement (it depends on the multimeter) that is supposed to represent the voltage that if it were DC would produce the same heating effect, but if the DC component is not utilized then it will not be a very useful measurement. The combined AC+DC RMS mode that the MIT420/2 offers will take into account the DC offset too (see the shaded area marked ‘B’). This means that the value on the display will represent the equivalent DC voltage that would produce the same heating effect as the combined AC+DC RMS measurement. For those interested, the formula (for a sine-wave AC signal) is:
Even when DC or AC is selected, the graphical dial will show the combined AC+DC RMS value, so a ‘dual display’ type of feature is implemented for maximum visibility into the measured signal. The frequency of the AC signal is also displayed.
In summary the voltmeter feature is excellent and the user interface conveniently places the function just before the insulation resistance modes, so that it is a quick thing to confirm no voltage is present before the insulation resistance test is executed. The voltmeter accuracy (+-2% for a sinewave) does not allow for the device to replace a multimeter for product design, but for installation and test before commissioning a machine or site wiring, it is more than adequate, and the excellent AC+DC RMS capability makes it really worthwhile.
The ohmmeter functions up to 1 megohm and is useful to confirm there is no short circuit before an insulation test is performed. It is also useful for measuring very small resistances (in steps of 0.01 ohm with an accuracy of 3% plus 2 digits) that is useful for electrical wiring installations. The main use-case would be to measure the resistance of the Live and Earth wires (shorted at one end) in order to determine what the fault current would be. In the UK it is called an R1+R2 test. This allows the proper determination if a circuit breaker will operate reliably in time (as mentioned before, this is not a replacement for a multimeter in a design lab; a high current is used to perform the low resistance tests which should not be used on electronic circuits). It is possible to null out the resistance of any attached test leads (Press and release the yellow Test button while in ohmmeter mode) and the nulling action persists even if the meter is powered off, until the Test button is pressed and released again.
The capacitance feature has a range that is quite restricted from an electronics engineer’s point of view but is highly targeted for measuring the capacitance in an electrical wiring installation.
Once it has been verified that no other appliances or attachments are connected to the equipment or wiring under test, the voltmeter, ohmmeter and capacitance meter functions can be used for further reassurance before selecting the insulation resistance test mode (the meter is protected and will not perform the insulation resistance measurement if any significant voltage is detected, but one should always verify first) and then pressing the ‘Test’ button!
Since it can be tiresome to keep the button held for long periods, there is a lock button too. A timed test is also possible (it is set to 1 minute by default but this is configurable) where the voltage is applied and after the set time, the test ends and the final insulation resistance measurement is presented on the display. The dielectric absorption tests that were described earlier (PI and DAR) can also be enabled. In all cases, the test runs autonomously and there is no need to hold a button down for a long period.
In summary the MIT420/2 provides an awesome set of functionality for obtaining insulation resistance measurements and ratios to get visibility into the condition of the insulation.
I wanted to use the insulation resistance tester to characterize the health of a telephone ringer. This is an electromechanical component that was used in old telephone sets.
(Image source: snippet from classicrotaryphones.com)
It consists of a solenoid with several taps on the winding, and a dome and small hammer. Inside a telephone it was connected to the line wiring, via a capacitor to pass only the AC ringing signal. This signal energized and de-energized the solenoid rapidly and caused the bell to ring. This is from an old Bell Systems telephone.
A faulty solenoid could cause an excessive load to be applied to the telephone line.
To test this, I first connected a normal multimeter to it. One test lead was connected to any of the winding leads, and the other test lead was connected to the metal chassis of the ringer. An open circuit was observed; it was not possible to measure the very high resistance between the coil and the metal chassis.
Next, the multimeter was removed and the MIT420/2 was used in its place. I set it to a 250V test voltage, which I hoped it could handle despite the normal phone line AC ringing voltage being lower (100V or so).
When I pressed the test button, the measured insulation resistance was about 68 megohms. Next I tried a DAR measurement. As described earlier this takes a minute to perform and the final ratio was 1.1; the solenoid isn’t very large so looking at the equivalent circuit model described earlier, it was likely that the leakage value was changing. The ringer had been stored in a cold shed and I’d brought it indoors for the test. It was interesting to see this! After a few hours I repeated the test and the insulation resistance had risen to 204 megohms and the DAR had reduced to 1.0. Another important test was to increase the voltage of the test and see if the measured insulation resistance value would change significantly. It did! At 500V the measurement was 188 megohms, and at 1kV it dropped to 102 megohms. The insulation is therefore quite poor, which is not a surprise (the ringer is more than half a century old and has not been stored in the best conditions) but likely still adequate at the lower expected telephone ringing voltage. So, the ringer will still provide many years of good service : ) A nice testament of the quality from Bell Systems.
In summary it was great that the insulation tester provided visibility that there was initially an issue (leakage from moisture) and also highlighted the ageing of the insulation but provided some reassurance that at telephone ringing voltage levels the solenoid should continue to function. It was great to see that for work on vintage coils, transformers and wiring, the insulation tester is really useful.
There is a built-in memory capability in the MIT420/2 and usually I’m pessimistic about ease-of-use with such functions. Often one needs a user manual in order to figure out how to record and view the data. I was really surprised that it was easy-to-use on the Megger. The button with the floppy disk symbol is pressed at any time and the measurement is captured into memory; an incrementing identifier is also briefly displayed. To retrieve the data, rotate the control to the blue ‘folder’ symbol and the identifier is shown. The left and right buttons allow selection of the desired identifier, and then the ‘OK’ button toggles between the identifier and its associated display screenshot. If only other products had such an easy-to-use data storage system : ) Deleting is also intuitive and won’t be described; there is a bin symbol on the rotary control. The device also has an easy-to-use configuration menu. It is obtained by rotating the dial to the spanner position. I kept the defaults but set the backlight timer off (I’d rather manually switch it off when desired).
The MIT420/2 provides a really rich set of functionality. A very decent voltmeter, and an ohmmeter for fault path impedance testing is built-in and this extends the usefulness of the base insulation resistance function. It provides comprehensive industry-standard and user-configurable settings for the insulation resistance measurement and ratio measurement techniques. The build quality is excellent, it feels rugged, and usability is excellent too. The buttons and rotary control are very intuitive and the backlight is excellent. The test probes will be used frequently, with a multimeter too.
As a newcomer to insulation resistance testers, I’m very happy that I discovered this new technique towards determining the health of wiring and of wound products, and it has given me the know-how to explore further in future with car parts and antenna cables. I’m told there is a very healthy business in testing and repairing vehicle alternators : )
In terms of improvements, I thought that the software button debounce was set a bit slow (but probably tuned for those wearing gloves), and that it would be more intuitive to have actions on button-press rather than on button-release (this occurred for a few features) but the well laid out user interface and easy-to-use features, configuration and data logging capabilities were really nice to see.
Also, the icing on the cake would have been to have temperature measurement capabilities built-in too, although a separate temperature meter or logger isn’t expensive (and some multimeters have it built-in).
Finally, it was really pleasing to see that the traditional AVO and Megger quality still appears to exist in their designs. Good quality and safety features can help save lives when it comes to electrical work, but the dangers are huge; always hire a qualified electrician.
One of the nice features of the Keysight was the ability to limit the current applied during testing.
I found it allowed testing of 'delicate' items without fear of damage.
Excellent road test report Shabaz.
Great detailed review, interesting to read about others ideas and views about the same tester.
An educational and detailed roadtest, top stuff Shabaz