The requirement for the electrical safety testing of medical equipment is essential to ensure that the apparatus is safe to operators and patients alike. Several tests are used to assess the integrity of insulation in electrical appliances.

Electrical currents are a necessary part of medical electrical devices; however, faulty or excessive currents can cause a serious hazard to the patient, operator, or medical device. As a result, the electrical industry and international community have implemented stringent procedures and requirements to ensure the safe and effective operation of medical devices. These procedures are embodied in the IEC 60601 standard for medical electronic devices.

The risk of unacceptably high electrical fault currents can be minimized through design criteria—i.e., through effective levels of electrical insulation and isolation between the operator or patient and live parts or potentially live parts in a fault condition. Such insulation can be achieved through physical spacing (creepage and clearance), dielectric materials (e.g., PTFE, PVC, short-fiber polymer paper), or proper component selection. These methods can help achieve the highest possible level of insulation while ensuring that the device operates properly.

The effectiveness of electrical insulation is tested through electric leakage measurements (results in mA or μA) while the level of isolation is often tested using a dielectric or insulation test. During a dielectric, or hipot test, a potentially high voltage (up to
4000 V ac) is applied across different parts of the electronic design to stress the dielectrics and identify total breakdown. Although repeating the dielectric test during routine testing could degrade the dielectrics, some manufacturers repeat this test during the manufacturing process as a final indication of quality. In this case, acceptable levels of dielectrics are then displayed as an amount of leakage current (typically in mAs or μAs).

An insulation resistance test applies a lower dc voltage, typically 250–
500 V dc, across different parts of the electronic design. The results are displayed in megaohms (MΩ). Unlike the dielectric testing at high voltage, conventional insulation resistance measurements have been the traditional method of completing preventive inspections of insulation levels in medical (and nonmedical) devices across a variety of guidelines and local standards. Although a 500-V dc insulation test is not specified within more recent releases of IEC 60601 (i.e., after 1989), the insulation test is now an optional part of the recently published standard for routine testing of medical devices—IEC 62353—since the voltage levels no longer affect the dielectrics.

It is important to note the differences between these two standards. IEC 60601 is an international standard that stipulates the minimum design requirements for safety and performance of medical electronic devices; IEC 62353 describes the tests that establish whether the electrical safety of a medical electronic device is still acceptable once in use or after a repair.

Despite the traditional merits of a 500-V dc insulation test to verify the level of insulation, this method can be problematic in some circumstances. For example, it may cause damage to the equipment under test, and it may also fail to indicate the true state of the insulation when presented with an alternating voltage. An alternative leakage test within IEC 62353 applies a typical line voltage (120 V) and frequency (60 Hz) as the insulation test source rather than dc. Both have their relative merits and place in periodic testing, provided that manufacturers understand the different limitations of each test (see Table I).

This article provides an overview of insulation testing and presents an alternative method based on IEC 62353. This method has been used across a number of local standards for both medical and nonmedical equipment and is now being adopted by the IEC standard. It is modified to meet the medical equipment requirements in line with IEC 60601 (i.e., the pass-fail limits of the alternative test represent similar thresholds to IEC 60601 tests).

 

Insulation Resistance

According to IEC 62353, insulation resistance is normally checked by applying 500 V dc between

• Input (live conductors, phase and neutral, connected together) and enclosure (protective ground in Class 1) (see Figure 1).

• Output (applied parts) and enclosure (protective ground in Class 1) (see Figure 2, p. 11).

• Input (phase and neutral) and output (applied parts) for floating type applied parts (BF and CF) (see
Figure 3).

The resistance is measured and compared with the minimum acceptable value to assess pass or fail conditions, which can vary greatly depending on design and test voltage variations. In other words, the thresholds vary depending on design and voltage. Note that the test voltage between different testers varies from 300 to 600 V dc even though testers claim to perform the test at 500 V dc. The test voltage will drop with the test load, depending on the output power of the 500-V dc supply. Voltage greater or less than 500 V dc will affect the reading.

With all measurements of insulation resistance, the device being tested must have the power switch (on-off switch) set to on before performing the test; otherwise, the test voltage does not pass beyond the mains switch, in which case only the insulation of the mains cord will be tested. (During the test, the power cord is connected to the tester and not mains.) However, the insulation resistance test does not power up the device. This could be seen as an advantage (reducing the time taken to test and eliminating the danger of moving hazardous parts), but extra care should be taken to ensure that the equipment switch is in the on position to complete a meaningful test.

In addition, devices fitted with electronic mains switches or residual current device plugs (also referred to as ground fault indicators) cannot be tested in this manner because it is not possible to close the mains switch (because mains must be present). In some cases, sensitive electronic devices may be damaged by 500 V.

Although a 500-V dc insulation test is typically quick and safe to perform, it usually does not provide a true indication of the effectiveness of the insulation in modern medical devices or the expected leakage values that may be experienced during normal or typical operation. This is due to the increased use of switch-mode power supplies that may indicate very high dc insulation resistances (i.e.,
>100 MΩ). But when measured with an ac voltage, these supplies could indicate high leakage. This is due to the greater influence of capacitive and inductive leakage experienced in these devices rather than resistive leakage, as in a heating element.

Infinity readings are common when performing dc insulation tests, but these readings do not indicate whether the unit was actually switched on or off. This renders the test results meaningless from a safety point of view.

It is a matter of debate as to whether a 50-MΩ (higher) result is safer than a 10-MΩ (lower) result, considering that the equipment has been exposed to a voltage at which it was not designed to operate. Furthermore, the 50-MΩ device might have been designed to measure 100 MΩ and therefore has lost 50% of its insulation level. This could lead to higher leakage currents and unsafe conditions.

Finally, in some electrical equipment, components connected to the live or neutral conductors for electromagnetic compatibility (EMC) filtering or surge protection can significantly influence the measurement, indicating an erroneous failure of the test. On the plus side, the insulation resistance test is relatively quick and easy to perform, which is probably why it is the most widely used.

 

Alternative Leakage

To verify the effectiveness of insulation while maintaining the speed and safety of a traditional insulation test, an alternative leakage method is contained in IEC 62353. The alternative leakage test is similar in setup to the dielectric strength test (high voltage), a dc insulation test, and the IEC 60601 earth-enclosure leakage test in the open phase or neutral single-fault condition. As with the IEC 60601 leakage test, the alternative leakage test is done at mains potential and frequency, thus representing operational conditions unlike the 500-V dc insulation test and the dielectric tests. This effective and safe method involves the application of a test voltage between the input and output of a medical device.

 

Equipment Leakage (Alternative Method)

The equipment leakage test is performed by placing a current-limited ac voltage (120 V, 60 Hz) between the mains input (live conductors, phase and neutral, connected together, and protective earth in Class 1) against output (applied parts) including the enclosure as shown in Figures 4 and 5.

The results measured with the alternative test can be compared with the IEC 60601 leakage measurements performed with open neutral single-fault condition. During a comparison between IEC 60601 earth leakage and the IEC 62353 alternative method, results showed a consistent relationship between the two measurements. The alternative method is roughly twice the expected leakage during normal conditions, similar to the IEC 60601 open neutral measurements.

 

Applied Part Leakage (Alternative Method)

Applicable to floating applied parts (BF and CF) only, the applied part leakage test is performed by placing the test voltage (120 V, 60 Hz) between the output (applied parts only) and enclosure (protective earth in Class 1) and input (phase and neutral) together. This is shown in Figures 6 and 7.

The test voltage is at mains potential and frequency of 60 Hz, which means the leakage paths will be similar to those present when the equipment is in operation. This setup avoids the problems associated with EMC filtering or surge protection affecting dc insulation tests. It also provides a more accurate reading of the true insulation, taking into account capacitive and inductive elements. However, the alternative leakage test still has some limitations because any electronic switches present will not be on (as with insulation testing), and relays or other active circuitry that may affect measurements may not be activated.

Any variation in leakage is easily noted, and the test rarely yields infinity readings as with the dc insulation tests. This method appears to be far more reliable in ascertaining the safety of the equipment under test. (In fact, an infinity reading probably tells the manufacturer that the unit is not switched on or that it has active circuits to power up the device. This would not be indicated in a dc insulation test because this test often gives infinity readings.)

Due to the fact that the equipment under test is not powered up, the alternative method is a safe and quick method of verifying the effectiveness of the insulation and thus the expected safety. Because both mains phases are shorted together during the test, no mains reversal needs to be performed, which saves time. Coupled with the more-accurate and realistic data, this makes for a safer and reliable test method.

Importantly, by stipulating various test methods and pass-fail limits, IEC 62353 provides the basis for consistent data collection and the development of formal preventive maintenance procedures. Because IEC 62353 also stipulates that a comparison is made between previous and current test results, it is more obvious when a device is likely to fail.

Conclusion

The onus remains on medical device manufacturers to determine the appropriate tests for their equipment. But IEC 62353 also affects medical service companies as well as clinical and electrobiomedical engineering, medical physics, and other technical departments in clinical settings. In particular, with the introduction of this standard, care should be taken in the specification and selection of medical safety analyzers to ensure that they can be used to test in accordance with the IEC 62353 requirements, and that they are capable of performing accurate and repeatable test routines.

 

John Backes is a divisional manager for Rigel Medical (Peterlee, County Durham, UK). He can be reached at johnb@rigelmedical.com.