Over in a flash – if you get it right during manufacturing
12 January 2009
PCB testing
The flash Test (also known as High Voltage (HV), Dielectric Strength or HiPot) is one of the more controversial electrical safety tests and has provoked a lot of discussion and debate in recent years. Stewart Haile, Business Manager at electrical safety testing specialists Clare Instruments, looks at new developments in flash testing and why it’s important.
The view that flash testing is essentially a destructive test is often an area of discussion. This originates from the use of flash in type testing where the long time period does provide potential for the degrading of insulation.
However, in terms of production line testing the reduced time period and the 5mA-trip setting significantly reduces this risk and the fact remains that many manufacturers successfully test without witnessing any degrading.
Flash testing is in fact not a measurement but a procedure that aims to check that a product is safe when subjected to high voltage and that the user is not exposed to danger.
The test is designed to detect that gaps or clearances between conductive parts and earth are sufficient and that damage in the form of pin holes / cracks in insulation and other protection devices have not occurred during manufacture or through wear and tear.
The test involves applying a high voltage to the product to check the insulation between the live conductors and all exposed metal surfaces. For Class I equipment the high voltage is applied between conductors and earth while for Class II equipment, the high voltage is applied between the conductors and the outer surface of the product. The flash test is necessary and designed to ensure the safety of the product. In particular, recent legislation like the LVD (Low Voltage directive) and issues of product liability have increased the need for manufacturers to demonstrate due diligence.
So what’s new? Although there have been few noticeable alterations in the flash testing requirements of most standards in recent years, changes in the technical specification of electrical and electronic products prompted by the EMC directive has required modifications to the flash test. For example, EMC considerations have required the introduction of circuit devices on the supply input to prevent emissions back into the mains. (Filter Circuits including X & Y Capacitors etc…)
These devices often take the form of resistive or capacitive or a combination of the two components. . When tested using an AC flash test, these circuits will often prove problematic as the capacitive part can induce leakage currents in excess of the capacity of the flash tester or demonstrate leakage greater to that required by the electrical safety standard. Situations such as these have led to a substantial increase in the use of DC flash testers, which are unaffected by this capacitive effect.
It’s been highlighted that AC flash testing could corrupt sensitive electronic components. Here, solutions include:
•Using a DC flash test. The voltage needs to match the specified peak AC voltage which is achieved by multiplying, the specified AC voltage by 1.414. A discharged facility following application to ensure that no residual voltage remains is required
• A “soft” DC flash test. This involves ramping up to the required voltage. In some instances this test can benefit from ramp down as well. This involves a slow ramp up from zero to the required value then holds for a timed period before ramping down to zero and discharging the unit under test
•The advent of EMC measures has increased the use of suppression devices etc, but this can cause problems for flash testing. While it should be noted that most designers of such components have upgraded their products to meet the specified flash tests, there are possible solutions.
It is important to remember that Flash Testing, together with earth bond testing (Class I) and insulation resistance measurement, are probably the three core tests for electrical safety testing. In addition, a load leakage test maybe specified also within the electrical safety standard.
Flash testing and insulation testing appear to be, on first examination, very similar. However there are fundamental differences: Flash testing is designed to detect gaps or clearances between conductive parts and earth, pin holes in insulation and other degrading as a result of production processes and /or wear and tear. Insulation resistance testing provides a quantitative measurement of the high quality of insulation.
If a wire was positioned 1/2mm from exposed metal, an insulation test – conducted in dry air – could well provide a pass reading. However, a flash test is more likely to detect this situation as dangerous. Similarly, if insulation is somehow contaminated, a flash test would produce a pass, but an insulation test would highlight a deficiency.
For example, the normal minimum insulation resistance value for Class I appliances is 2 M Ohm, with a 1500V flash test, the current would be 0.75mA and would not be detected by the 5mA and would not be detected by the 5mA trip which has to accommodate the capacitate losses that occur.
Obviously a DC flash test with a leakage meter can provide insulation resistance monitoring as the capacitate component is overlooked after the initial in rush of current.
Test identification & traceability
The flash test can be regarded as a negative test on the basis that a good product will often not provide a measurable flow of leakage. With the development of modern instrumentation using microprocessor control and data logging software it is now possible to produce fully traceable records of testing undertaken.
Understanding why flash testing is necessary is important but being able to prove that electrical or electronic products comply with the various standards is also vital – particularly if there is a subsequent failure or fault identified. The only effective means of demonstrating that a product has been tested is through proper documentation.
Flash testers, such as the HAL from Clare, automate the testing process on the production line and retain results in an internal memory for later downloading or print-out. Documentation through automated flash testing minimises liability and provides effective proof that a product is safe at the end of the manufacturing process.
What should the test conditions be? Two distinct forms of testing are usually recognised - type testing and production line testing. The former varies according to the relevant product specific standard, but normally for Class I equipment they require between 1000-1500V applied for one minute with a trip level of 100mA. For Class II equipment, voltages are usually higher at between 2500 and 4200V, but with similar timing and trip settings.
Production line flash testing requires special considerations in terms of practicalities such as time and high voltage levels. Here, the need to have faster but equally rigorous tests is recognised by applying 10% over voltage, but reducing the test time to a few seconds. Thus a type test, with a voltage rating of 1250V, would be carried out at 1375 on the production line with trip levels of 5mA.
One cautionary note is that it is possible to get a seemingly satisfactory result when the equipment is switched off or not properly connected. Obviously there is a need to ensure the equipment under test has the power switch on and the unit under test is properly connected.
Even for experienced operators this can be a stumbling block. In production line situation, such problems are emphasised, because of the greater throughput of products.
Solutions include a simple continuity test, applied on live and neutral, built in to the test programme prior to the flash test and the detection of captive leakage that occurs whenever an AC flash is applied. If no leakage is detected a warning is initiated. Regular fault simulation at the test connection point should also be applied, at each shift change or at each start of a production day is highly recommended.
The test itself is not quantitative and fail is recorded if a breakdown of insulation or a flash over between components occurs. Most testers indicate pass or fail via a warning light and/or sound which activates when 5mA / greater of leakage occurs.
The best way to maximise productivity is to reduce the time taken to apply all of the safety and functionality tests. By using an integrated test station and enclosure connection to the device under test, only one test sequence is required. As establishing a connection is often the most time consuming part of production line testing, combining up to four or five tests at an integrated test station can significantly reduce time and cost.
Type tests often call for high levels of high voltages to be applied for up to one minute. However, in reality this is not practical in production facilities. This is because at most factories a one minute test would reduce productivity, while the call for a 100mA-trip level can cause a potentially lethal scenario. Lastly, voltage levels and test procedures realistically demand a skilled operator.
Safe test areas
Production line flash testing requires special considerations in terms of practicalities such as time and high voltage levels. Effectively designed test instruments mean that the required operator skill level can be reduced and the use of high trip levels can be protected by safe systems with only qualified operators having access.
With the integration of EC electrical safety testing standards in EN50191, specific safety conditions have been specified for all locations where electrical testing is carried out. For example, the use of test enclosures on the production line is advisable to minimise the safe working area around the points where flash tests are to be applied.
In Class II equipment the absence of an earth requires protection via primary and secondary insulation. The first practicality to mention is that flash testing of Class II equipment involves much higher voltage levels, typically between 2500 and 4200V. A common problem, particularly on new equipment, is that you can have failure on the primary insulation that is undetectable by a flash test on the outer surface, which tests the secondary insulation only. In testing regimes BOTH these need to tested.
In order to test the primary protection one needs to find a method of accessing the primary insulation. This is essentially a contradiction in terms, since this connection needs to be inaccessible metal! However, experience shows the following options as feasible:
•Test of the primary insulation prior to final assembly. However, there has to be a check that on assembly no degrading of this protection takes place, e.g. screws penetrating the insulation
•Design the product with an access that can be permanently sealed after testing. This is often an element that product designers fail to anticipate
•Design test jigs and probes that allow access through the enclosure; ensuring that the integrity of the product on terms of the relevant standard finger tests is maintained.
Testing also needs to be carried out on the secondary protection. Standards generally define the product be wrapped in aluminium foil so that high voltages can be applied to all the outer surfaces.
This test may be practical for laboratory situations but is impractical for production scenarios, both because of complexity in test setup and time and also in that the outer surfaces of the product can easily be marked.
Clare’s pioneering use of conductive foam in a special jig to create a nest or envelop around the outer surface of the product enables test voltages to be applied. Although not quoted in standards, this methodology is recommended by standard authorities.
These include some standards which allow the disconnection of such components for safety testing - although this is often impractical for production line testing. Another approach is setting higher trip levels e.g. 10 or 15mA.
The use of this option should always be accompanied by the use of safety precautions such as key locked switches, so that authorisation is gained before carrying out this test.
Ideally, the safest solution in such circumstances is to conduct the test with the item housed in an enclosure with appropriate interlocks. Our experience has shown that these need not be too complicated or expensive and provide maximum operator safety.
Contact Details and Archive...
