Insulation Resistance (IR) is maybe the most common motor test. It is also one that has more components of currents involved than some users realize. An insulation resistance test can be done with a hand-crank meter to measure megohms in its most basic form, or with a more advanced electronic device that may measure and display voltage and leakage current, and calculate megohms, DAR and PI ratios, as well as plot graphs of megohms over a period of 10 minutes or more.
In an IR or megohm test the voltage applied and the total leakage current are measured between the windings and the motor frame/ground. Ohms law is applied to calculate the resistance in megohms.
Where R is resistance in megohms, V is voltage applied in Volts, and I is the total resulting current in micro Ampere (µA).
A temperature correction factor may be applied to correct the megohm measurement at present temperature to what it would be at a standard temperature of 40°C per IEEE 43 and ANSI/EASA standards.
In a megohm measurement of a used DUT (device under test) that has not been cleaned, the leakage current is often mostly surface current running in the “dirt” on the outside of the windings. Dirt, in this case, includes particles, oil, grease, moisture etc. The ground insulation certainly can be weak with conduction current flowing through the insulation to ground, but this is often dwarfed by the surface currents. Therefore, the insulation resistance test or megohm measurement is sometimes referred to as the “dirt test” as the megohms tend to drop with increasing amounts of dirt.
With new motors, a megohm measurement is often not interesting, other than to check that there are no direct shorts to ground. Users will often go directly to a Hipot test.
Currents involved in a megohm measurement – not all are “bad”
- IC– Capacitive current: The DUT has capacitance, and a capacitive inrush current brings the potential in the DUT up to the test voltage by “charging up” the DUT. This current will drop quickly and reaches zero within a few seconds after the test voltage is reached. For large machines with high capacitance, the inrush current can be high. Total leakage current failure limits must be set high enough to avoid tripping the limit during this initial phase of the test. For more information on the subject of capacitive inrush current and how to avoid tripping a limit, see Hipot Test.
- IA – Absorption current: This is the current it takes to polarize the insulation. In random wound motors built after about 1980, this current also goes to zero or very near zero within 30 seconds to 1 min, often much quicker. For form wound motors this can take much longer because of the layers of insulation used with different properties. The change in absorption current over time is what is used to calculate DAR and PI ratios in an insulation resistance test.
- IG – Conductance current: This is the current flowing between the copper conductors and Ground through the bulk of the insulation. This current is usually zero if the motor is new or undamaged. As the motor insulation ages and cracks or is damaged, conductance current may start to flow depending on the test voltage applied. It tends to accelerate with increasing voltage. This current is sometimes referred to as leakage current, or as part of the leakage current.
- IL – Surface Leakage current: According to IEEE 43 this is the current flowing in the dirt and humidity on the surface of the windings to ground. It is called surface conduction current in other standards. The dirtier the DUT is, the higher this current is, and the lower the megohm result will be per Ohms law. After 1 minute with a random wound DUT, or 5-10 minutes with a form wound DUT, this is typically the only current remaining unless the insulation is weak or dama
There may be an increase in the surface leakage current on machines where a stress-control coating has been applied to the end-windings.
- IT – Total current: The total “measured current” is the sum of the 4 currents, and is what is measured by the insulation tester, motor tester or motor analyzer. From the information above one can see that in many cases the total current equals or is very close to the surface leakage current at the end of the insulation resistance test. This gives the operator a good measure of how “dirty” or contaminated the DUT is. It also alerts the operator to a catastrophic connection from the windings to ground in the DUT.
To find out if the leakage current is mainly a surface current or also contains the more damaging conductance current through the volume of the insulation to ground, one must do a step voltage test or ramp test to a test voltage one is comfortable with given the level of megohms measured. See information below on minimum megohm levels. Click here for information on DC Hipot testing. Note that these tests can be done to voltages lower than the normal DC Hipot test voltages to find out if there is conduction current.
Tracking megohm measurements over time
Megohm measurements may be tracked over time to help determine when a motor or generator should be reconditioned. This can be done automatically with the iTIG II motor analyzer. Especially for larger DUTs, other insulation resistance tests including DAR or PI tests are used in reconditioning assessments along with DC Hipot Step Voltage tests or Ramp tests, Surge tests and Partial Discharge measurement.
Standards and temperature compensation
ANSI/AR100-2015 and IEEE 43-2013 make the same recommendations as follows.
It is recommended that machines with low Insulation Resistance test readings not be subjected to high-voltage testing.
NOTE: The limits above are for windings at a temperature of 40°C. Since the windings usually are not at this temperature when tested, the megohm test results should be temperature compensated. Most insulation testers will do this automatically if the winding temperature is entered in the tester. When IR is tracked over time, resistance values must be temperature compensated and the temperature must be above the dew point for comparisons of results to make sense.
The most common temperature compensating formula says that the insulation resistance drops by a factor of 50% for every 10°C increase in temperature. From this formula, it is very clear that the insulation properties drop precipitously as the temperature rises, and we all know what happens to over-heated motors. IR of 10,000 megohms (10 Giga Ohms) at 20°C (~68°F) drop to 2,500 megohms at 40°C, and to 39 megohms at 100°C.
There are several temperature compensation formulas, and the one above may be the most conservative. Different types of insulation systems in form wound motors will have different temperature characteristics. These can only be obtained from the manufacturer of the motor.
The bottom line is that temperature has a significant effect, and must be taken into consideration.
Limitations of interpretation
Question: How much better is test number 1 than test number 2?
Answer: Who knows, maybe no better at all? A difference of 0.01µA could be the result of a number of things such as temperature, changes in other environmental conditions, electrical noise, instability in the voltage or current etc.
The difference in insulation resistance is high because of how resistance is calculated. But, what physically changed is the current, and the change in current in the example above is extremely small. Some insulation testers display leakage current to the 3rd or even 4th decimal with a resolution as low as 1nA or 1pA, and will calculate and display Terra-Ohm (TΩ). For such insulation testers, the accuracy in the last digit(s) is not specified or is poor, and for good reason. It is too dependent on factors other than the leakage current it is supposed to measure, and significant operation errors may come into play. See cautionary note from IEEE 43 under PI below.
Other advice and tips from IEEE 43-2013
- Before starting a test, the winding insulation should be discharged to avoid measurement errors.
- For machines with a stress-control coating applied to the end-windings, there may be an increase in the surface leakage current and thereby lower megohms than expected.
- For winding temperatures below the dew point, it is impossible to predict the effect of condensation on the surface. Therefore, a correction to 40 °C for trend analysis will introduce significant errors.
- For directly water-cooled windings, the water should be removed and the internal circuit thoroughly dried. The winding manufacturer may have provided a means of measuring results of the insulation resistance test without need for the coolant water to be drained.
- A minimum discharge time of four times the voltage application duration is recommended.
Electrom Instruments note: Measurement instruments discharge the DUT through a resistor in the instrument. When the DUT voltage is low (for example less than 100V), connecting the winding directly to ground with the instrument ground lead or a shorting stick or jumper will almost immediately complete the discharge. The exception to this is residual absorption discharge if one exists. However, this discharge will also speed up, but can take some time. If a DUT with absorption charges is to be handled right after a test, keep it connected directly to ground.
- Absorption discharge: This decay may take more than 30 min depending on the insulation type and physical size of the DUT.
- A signiﬁcant decrease in insulation resistance (increase in measured current) with an increase in applied voltage may be an indication of insulation problems in an insulation resistance test.
- For tests conducted under similar conditions, a steady increase in the IR with age, (decrease in absorption current) may indicate decomposition of the bonding materials, especially when the insulation materials are of the thermoplastic type.
- When a low PI occurs at temperatures above 60°C, a second measurement below 40°C and above the dew point is recommended as a check.
- PI can be used to indicate when the drying process of insulation may be terminated. This would be when the PI has exceeded the recommended minimum.
- If the IRvalue (at 40 °C) is greater than 5000 MΩ, the PI may be ambiguous and can be disregarded.