Some thoughts on Flash Testing

Ray Jefferis is a Senior Technical Manager at ASTA BEAB in Guildford, UK. He has over 25 years of experience in the electrotechnical sector, in engineering, testing and certification.

Most certification schemes insist upon 100% flash testing of all products. The purpose of this paper is to explore the merits (or otherwise) of this requirement. There is no doubt that flash testing conducted correctly during the production process can be of benefit and assist in ensuring that a product has been assembled correctly, that certain components will not break down and the product will remain safe. The requirements for 100% flash testing can however be misconstrued and in many production environments, the tests being applied are of little or no benefit.

What is Flash Testing?

The flash test is similar to the 'electric strength' test described within safety standards relating to electrical products and checks the strength of the electrical insulation of the product between live parts and earthed metal or live parts and accessible insulated material or unearthed metal. Whereas standards require the electric strength voltage to be applied for up to several minutes, a flash test applies a similar voltage level, but for a period of one to two seconds. The voltage is normally only applied between live parts and strategic insulated parts of the appliance or it's earth and is used as a confidence tool that 'the product' has been put together correctly and that the insulation system has not been compromised.

Class I Product

This is probably the easiest product to flash test because the voltage is applied between live and neutral connected together and earth. This will ensure that no breakdown is occurring between basic insulation and the earth plane, perhaps due to a wire becoming trapped. A Class I product will normally require 100% flash testing due to its construction where basic insulation is allowed to come into contact with the earth plane. However, this test may be being applied where analysis of the construction of the product reveals that failure of this test was either impossible, or if a failure occurred it would not be shown up by the test anyway.

One aspect of a Class I product that is often missed is that the product can have Class II areas that are not flash tested at all. In many Class I products it can be the Class II areas where there is a greater risk of undetected problems. For example, a trapped wire between two different earthed metal parts may result in a functional failure that is immediately obvious during functional testing of the appliance. Whereas a trapped wire between two plastic mouldings could result in live wire strands protruding from the product that would not be apparent unless a flash test is conducted at this point or, worse still, somebody received an electric shock!

Class II Product

Class II product, and Class II situations in Class I product, are normally the most difficult on which to conduct a meaningful flash test. It is the construction of this type of product that needs careful evaluation before investment in the production test apparatus is made, so conducted tests are not meaningless. Take a normal vacuum cleaner for example, many of these products basically consist of a plastic box in which is mounted the vacuum motor and an on/off switch. The motor itself has normally already been 100% flash tested prior to assembly into the final product to check the basic electric strength of the motor. In many cases the only accessible metal on a vacuum cleaner will be the screws that hold the plastic box together. Applying a flash test between these screws and live and neutral supply conductors connected together could be meaningless because the screw nearest to live parts could be several centimetres away and could never come into contact with the internal wiring or other live parts. Many products are tested pointlessly in this way adding cost and delay into the production process with no benefit, detracting from the reason for conducting the test in the first place.

Continuing with the vacuum cleaner example, the risk areas are more likely to be where the different parts of the plastic box are joined and internal wiring could become trapped causing basic insulation or live parts to become accessible. It is in these areas of Class II product where flash testing can be of benefit. Rather than apply the test voltage to a fixing screw, the voltage can be applied along the join in the product case using a method, for example, such as conductive foam.

Many companies have recognised that, especially with Class II products, it is far more cost effective to design the product in such a way that, should a wire become trapped or some other assembly defect occur, it is either readily obvious, or the product just cannot be put together. In fact, some companies spend considerable amounts of money during the development phase to alter the design of a product to ensure that flash testing is not required on every product.

However, does this mean flash testing is not required at all under these circumstances? No, but in these cases it could be recommended to conduct flash testing on sample products, for example one per batch as part of the end-of-line quality audit. In this way, product assembly is quicker and confidence in the electric strength of the complete appliance is maintained during the complete product life cycle.

Class III Product

Class III product, and Class III situations in Class II products, are also often overlooked during production testing. This is mainly due to the mistaken idea that a low voltage output poses no safety hazard and therefore no flash testing is required. This is incorrect because many Class III outputs can deliver large currents, and should short circuits occur, there could be a significant risk of fire. A simple analogy is that of an arc welding plant that only produces 1 or 2 volts but a large enough current that can easily melt stainless steel.

Further Considerations

It is always advisable that each product designed is assessed and a production line safety-testing regime developed that is most meaningful and cost effective. Many companies do not change tests from one product to another, leading to some products being inadequately tested and others having safety tests conducted that are completely unnecessary.

The trip current setting of the flash tester should be reviewed for each product design, the trip levels are normally adjustable either by a potentiometer or a position switch on the test set. The level that should be chosen is that just above the actual current expected or measured for the particular product design. This will vary considerably between Class I and Class II product; especially where RFI filters are used between live parts and the earth of Class I products, or between live parts and unearthed metal of Class II products. Having the trip level set correctly will ensure that any product that is tending to draw a higher leakage current will become readily apparent.

It is not unusual to find that a factory is applying the flash voltage with the trip current set to 30mA, whereas the product itself may only draw 1mA leakage current. This is normally because the test has been misunderstood and setting to the higher trip level has been mistakenly considered to be the worst case condition - which it is not. This means that faults on the product may not be picked up by the flash test. These faults may not necessarily be safety defects, but rather be caused by incorrect components being fitted etc. Another issue that is often overlooked is the ramp time of the flash test voltage. On the production line the flash voltage is only applied for one or two seconds, whereas the relevant safety standard may require the voltage to be applied for several minutes. Sometimes there is a conflict between the time it takes for the voltage to rise to the test voltage and the overall test time. It is not unusual to find that the flash tester resets before the flash voltage reaches the required level, hence the full flash voltage is never applied to the product.

Daily Check

A daily check should be conducted on the flash tester and its associated special leads and fixtures to ensure they are basically functional. Although referred to as the 'daily check', in larger product plants this may need to be conducted more than once per day. This is to minimise the number of products that may be at risk if the daily check is found to fail the next time it is conducted. Similarly, smaller production plants may not produce product every day and therefore no daily check needs to be conducted. Providing the records are clear, then this is not a problem.

The check is normally achieved by using either a test box fitted with two resistors or, for automated production lines where the flash voltage is applied automatically to the product whilst mounted on the production line, modified products containing the resistors. One resistor value is chosen that gives a just pass at the flash voltage and trip current and the other resistor value chosen to give a just fail. In this way basic confidence in both the voltage and trip level of the flash tester is maintained. It is important that all test leads etc. used during this process also form part of the daily check as it is not unusual to find the daily check only being conducted on the basic test set and then finding the test leads being used are open circuit. This means that any flash testing conducted is meaningless, as the flash voltage is not being applied to the product. Sometimes also a short circuit is used to check the trip level, but this is pointless, only indicating that the trip works but not if the actual level or if the voltage is correct. The purpose of using the resistors is to check functionality of the trip level and has several advantages:

a) They provide confidence that trip level and voltage is correct
b) They ensure that a faulty test set is detected if the trip level has drifted to a higher value
c) They ensure good product is not rejected if the trip level has drifted lower

Where different products are produced in the same factory it may be necessary to have different values of test resistor. This is because each product may have a different standing leakage current and require a different trip level, or it may be of a different insulation class and need an alternative test voltage. The resistor values must be chosen to test both the voltage and trip levels at the relevant values for the particular product. This ensures the trip level and/or voltage has been set correctly before the start of production of the new product.

In conclusion, the flash testing conducted at many factories is a going-through-the-motions exercise rather than an effective and beneficial test. To get the most out of this flash testing, each product design has to be evaluated on it's own merits and the right test regime implemented and maintained accordingly - making the tests cost effective, worthwhile and meaningful.