PCB007 Magazine


Issue link: http://iconnect007.uberflip.com/i/1299286

Contents of this Issue


Page 85 of 103

86 PCB007 MAGAZINE I OCTOBER 2020 tion states not less than 72 hours by commit- tee agreement. This was to revert to 168 hours, but evidence is now available that flux resi- dues may lie dormant for beyond 500 hours in service, and hence there is a need for a three- week or four-week test. Today, there are two different developing technology regimes: (1) high voltage (~1,000 V) electronics for electric vehicles and (2) low voltages and fine-pitch devices (~2 V and <100-µm feature size) in the medical and space industries. The existing test is not appropriate for these technologies. For example, in the high-voltage testing standard ISO PAS 19295:2016(E), electric components or circuits are required to operate with a maximum work- ing voltage between 30 V AC (rms) and 1,000 V AC (rms) or between 60 V DC and 1,500 V DC. An impetus for a new test has been created by the removal of the ROSE test from Rev G of J-STD-001. There is a desire to qualify clean- ing efficacy underneath bottom terminated components (BTC), using a modified SIR test, along with a new test vehicle, that can take ad- vantage of low-cost test vehicles and use SIR patterns underneath the BTC to evaluate clean- ing efficacy. Following on from a current HDPUG proj- ect on corrosion, a method will be produced to look at pitting and crevice corrosion through solder mask. Here a modification of the SIR technique is proposed to evaluate solder mask integrity and use a new test vehicle. Aim The aim here is to develop new SIR stan- dards to cover the low and very high voltag- es, and validate the developed approach with a Gauge R&R study, all under the auspices of IPC. The work will build on the approach in but will tailor the approach appropri- ately for the two technology areas. will also be updated from the current 2007 version, and incorporation of the new IPC B53 test pat- tern, which incorporates a 200-µm SIR pitch pattern. The new standard will also look at test duration, which will include a minimum of one option to test for at least 500 hours and possibly beyond 1,000 hours. The cleaning efficacy SIR evaluation, and the corrosion of solder masks, will follow a simi- lar path, with the development of test vehicles, dummy components, and a test protocol. The IPC SIR committee 5-32b will lead the develop- ment of the documents and organise a round robin trial with Gauge R&R validation. Funding The standards will be written as now with- in 5-32b by voluntary work. Production of the samples might be funded by IPC, and then the intercomparison work by the collaborators will take place at their expense. Methodology For these, we need a consensus on the track and gap for the patterns. Our current point is 25 V/mm, with the 200-µm (B53) and 500-µm (B24) patterns. For low-voltage applications, it is envisaged that the test voltage of 2V and ~50- µm track spacing, and this will lead to 40 V/mm. With high-voltage testing, an anticipated field strength of 500 V/mm is expected, hence with a 1,000 V test, the feature spacing will be 1 mm. It has been demonstrated that electrochemi- cal processes will not always scale with feature size, or SIR pattern pitch, and the applied volt- age. Thus, careful consideration must be given to the applied test, and the conditions of the test must be applicable to the use case. If not, the produced data can be valueless. Therefore, new test coupons will be required, and the in- put from the wider industry is essential to de- fine the requirements. New material systems are known to have a long incubation period before the onset of cor- rosion; periods of up to 500 hours have been noted. Testing at >1,000 V may generate fail- ure modes that occur over relatively long dis- tances and hence may take even longer times, and test durations of over 1,000 hours may be required. Proposals for the cleaning effica- cy SIR evaluation and the corrosion of solder masks will be brought forward. Validation When the committee has agreed on the test methodology, the method and chosen test ve-

Articles in this issue

Archives of this issue

view archives of PCB007 Magazine - PCB007-Oct2020