Printed Circuit Board Failures Causes and Cures Webinar

Transcription

1 Printed Circuit Board Failures Causes and Cures Webinar Bob Willis bobwillis.co.uk Your Delegate Webinar Control Panel Open and close your panel Full screen view Submit text questions during or at the end

2 Bob Willis Bob Willis has been involved with the introduction and implementation of lead-free process technology for the last seven years. He received A SOLDERTEC/Tin Technology Global Lead-Free Award for his contribution to the industry, helping implementation of the technology. Bob has been a monthly contributor to Global SMT magazine for the last six years. He was responsible for coordination and introduction of the First series of hands-on lead-free training workshops in Europe for Cookson Electronics during These events were run in France, Italy and the UK and involved lead-free theory, hands-on paste printing, reflow, wave and hand soldering exercises. Each non commercial event provided the first opportunity for engineers to get first hand experience in the use of lead-free production processes and money raised from the events was presented to local charity. More recently he co-ordinated the SMART Group Lead-Free Hands On Experience at Nepcon Electronics This gave the opportunity for over 150 engineers to process four different PCB solder finishes, with two different lead-free pastes through convection and vapour phase reflow. He also organised Lead-Free Experience 2, in He has also run training workshops with research groups like ITTF, SIN TEF, N PL & IVF in Europe. Bob has organised and run three lead-free production lines at international exhibitions Productronica, Hanover Fair and Nepcon Electronics in Germany and England to provide an insight to the practical use of lead-free soldering on BGA Ball Grid Array, CSP Chip Scale Package, 0210 chip and through hole intrusive reflow connectors. This resulted in many technical papers being published in Germany, USA and the United Kingdom. Bob also defined the process and assisted with the set-up and running of the first Simultaneous Double Sided Lead-Free Reflow process using tin/silver/copper for reflow of through hole and surface mount products. Bob also had the pleasure of contributing a small section to the first Lead-Free Soldering text book Environment - Friendly Electronics: Lead-Free Technology written by Jennie Hwang in The section provided examples of the type of lead-free defects companies may experience in production. Further illustrations of lead-free joints have been featured in here most recent publication Implementing Lead-Free Electronics He has helped produce booklets on x-ray inspection and lead-free defects with DAGE Industries, Balver Zinn and SMART Group Mr Willis led the SMART Group Lead-Free Mission to Japan and with this team produced a report and organised several conference presentations on their findings. The mission was supported by the DTI and visited many companies in Japan as well as presenting a seminar in Tokyo at the British Embassy to over 60 technologists and senior managers of many of Japans leading producers. Bob was responsible for the Lead-Free Assembly & Soldering "CookBook" CD-ROM concept in 1999, the world s first interactive training resource. He implemented the concept and produced the interactive CD in partnership with the National Physical Laboratory (NPL), drawing on the many resources available in the industry including valuable work from NPL and the DTI. This incorporated many interviews with leading engineers involved with lead-free research and process introduction; the CD-ROM is now in its 3rd edition.

3 Bob Willis currently operates a training and consultancy business based in England. Bob is a member of the SMART Technical Committee. Although a specialist for companies implementing Surface Mount Technology Mr Willis prov ides training and consultancy in most areas of electronic manufacture. He has w orked w ith the GEC Technical Directorate as Surface Mount Co-Coordinator for both the Marconi and GEC group of companies and prior to that he w as Senior Process Control Engineer w ith Marconi Communication Systems, where he had w orked since his apprenticeship. Follow ing his time w ith GEC he became Technical Director of an electronics contract manufacturing company where he formed a successful training and consultancy division. As a process engineer, he w as involved in all aspects of electronic production and assembly involved in setting up production processes and ev aluating materials; this also inv olv ed obtaining company approval on a w ide range of Marconi's processes and products including printed circuit board manufacture. During the period w ith Marconi, ex perience was gained in methods and equipment for env ironmental testing of components, printed boards and assemblies w ith an interest dev eloped in many areas of defect analy sis. Over the last 15 y ears he has been inv olv ed in all aspects of surface mounted assembly, both at production and quality lev el and during that time has been inv olv ed in training staff and other engineers in many aspects of modern production. Ov er the past few y ears Mr. Willis has trav elled in the United States, Japan, China, New Zealand, Australia and the Far East looking at areas of electronics and lecturing on electronic assembly. Mr. Willis w as presented w ith the Paul Eisler aw ard by the IMF (Institute of Metal Finishing) for the best technical paper during their technical programmes. He has conducted SMT Training programs for Tex as Instruments and is currently course leader for Reflow and Wav e Soldering Workshops in the United Kingdom. Mr Willis is an IEE Registered Trainer and has been responsible for training courses run by the PCIF originally one of Europe's largest printed circuit associations. Bob has conducted w orkshops with all the major organisations and ex hibition organisers World Wide and is know n for being an entertaining presenter and the only presenter to use unique process v ideo clips during his w orkshops to demonstrate each point made. Find out more at: Bobwillis.co.uk Mr. Willis w as Chairman of the SMART Group, European Surface Mount Trade Association from and has been elected Honorary Life President and currently holds the position of SMART Group Technical Director, he also w orks on BSI Standards Working Parties. He is a Fellow of the Institute Circuit Technology, an NVQ Assessor, Member of the Institute of Quality Assurance and Society of Env ironmental Test Engineers. Bob Willis currently w rites regular features for AMT Ireland, Asian Electronics Engineer and Circuits Assembly the US magazine. He also is responsible for w riting each of the SMART Group Charity Technology reports, which are sold in Europe and America by the SMTA to raise money for w orthy causes. Bob ran the SMART Group PPM Monitoring Project in the United Kingdom supported by the Department of Trade and Industry. He w as coordinator of the LEADOUT Project for SMART Group. LEADOUT w as one of the largest EU funded projects, currently he is coordinating European projects TestPEP, ubga and ChipCheck Printed Circuit Board Failures Causes & Cures Webinar Through hole plating failures CAF contamination shorts PCB Delamination Nickel/Gold Black pad & Black Tar Inner layer separation Solder mask cracking Outgassing Need for product specifications Auditing a PCB supplier

4 PCB Manufacturing Reference Books PCB Manufacturing Reference Books

5 IPC Specification Documents IPC 600 F IPC Specification Documents Specifications for solderable surface finish

6 Printed Circuit Board Defects Printed Circuit Board Defects

7 Plated Through Hole Manufacture Dr Paul Eisler credited with making the first PCB Printed Boards were commonly used after World War II During s single sided boards were used Through Hole Plating developed during late 1960 s Multi Layers boards were developed during the 1970 s Plated Through Hole Manufacture PCB PCB PCB

8 Plated Through Hole Manufacture PCB PCB PCB Plated Through Hole Manufacture Blind via hole

9 Plated Through Hole Manufacture Buried via hole Initial Supplier Audit Requirements Supply a full package of information electronically or hard copy five days before any visit, this should include: You should obtain general information on the company, location, map and contact details List of production equipment and capability for the products being proposed for fabrication List of production process capabilities and processes available onsite, any off site services should be confirmed

10 Initial Supplier Audit Requirements Design requirement/customer DFM Report for new builds Fabrication flow path for each type of circuit being considered List or Inspection/process checks conducted for the build List of training modules production and quality staff attend Copy of technology roadmap and its introduction time scales, good to ask for one a couple of years before the date of visit Supplier Audit Report Procure a sample batch of the new circuit board design Assess the quality of the samples to relevant standards and run through assembly process as part of a NPI build Produce a report to compare suppliers strengths & weaknesses Alternatively highlight issues at the end of the visit

11 Traditional Sulphur Free Tissue Paper Outer packaging to provide protection to the boards and tissue paper interleaved between boards to prevent surface damage to solder mask, legend ink and solder pad surfaces. These techniques are still used but vacuum sealed is the most common standard offered Pink Poly or Other Packaging

12 Moisture Barrier Bag Bags available with and without zip lock. Also the barrier material can be provided in roll format with a sealing unit to produce custom sized bags PCB Mechanical Inspection

13 PCB Mechanical Inspection Measurement of bow and twist on a flat surface PCB Mechanical Inspection Copper through hole measurement using eddy current principle, typical through hole thickness would be 25-30um of copper

14 Optical Inspection Systems Optical Inspection Systems

15 Adhesion Testing (Tape Test) Adhesion Testing (Tape Test) Tape testing is all about the bond to the base material, normally solder mask and not necessarily a test of the coating material s performance. Different solder masks formulation may give varying results ASTM D3359

16 Solder Mask Thickness Solder Mask Thickness Agree specification with supplier for mask thickness and variation on the surface of the panel

17 Solder Mask Undercutting Printed Board Markings Date of PCB manufacture will be marked on the surface of the board in the legend, formed in the solder mask or on the copper surface

18 Multilayer PCB Manufacture Microsection example of through hole copper plating and tin/lead plating on the knee of the hole The original copper laminate foil is also visible Multilayer PCB Manufacture Microsection of surface conductor copper foil, copper plating and tin/lead plating

19 Multilayer PCB Manufacture Microsection of inner layer conductor foil copper plating connection in the through hole Multilayer PCB Manufacture Microsection showing blind and buried via holes with connection to the inner layer

20 Plated Through Hole Manufacture Plated Through Hole Manufacture Copper metallisation on through hole prior to copper plating Copper through hole metallisation inspection using the back light test for coverage

21 X-Ray Inspection X-Ray Non Destructive Inspection

22 X-Ray Non Destructive Inspection Blind Via PCB Examination In Process Close up of the PCB surface copper after drilling and prior to metallisation to make the interconnection to the capture pad

23 Via Too Deep to Image Successful SEM examination would also not be possible due to the depth of the hole. Most PCB produces also don t have SEM Grind Down PCB Sample Closer to Pad Close up of the PCB surface copper after drilling and prior to metallisation to make the interconnection. The sample has been ground down to reduce the step height between the capture pad and the top of the lam inate, it is possible to grind down closer to the copper pad surface

24 Ultrasonic Clean Sample in IPA Surface of the Capture Pads ay High Magnification

25 Testing Blind Via Hole Quality in Printed Board Manufacture Minimum tools required Customers defective boards Grinding wheels Very fine scalpel blade Very good quality tweezers Magnification low and high Good deal of patience Printed Board Assembly Featuring Blind Via Holes If open circuits are suspected on blind vias this is a way to look at the interface of the capture pad and copper plating. Often there is great debate on what is good and bad, failure or not. Often it is difficult to see the interface or conduct analysis of the interfaces. Grind the PCB to just below the capture pad

26 Printed Circuit Board Featuring Blind Via Holes Printed circuit boards could be tested prior to assembly using standard thermal shock tests defined in international standards. The photograph shows the track and capture pads through the laminate as we start to grind to the back of the pad. Grind the PCB to just below the capture pad Printed Board Assembly Featuring Blind Via Holes Capture pad & track can then be peeled from the base of the via hole The image shows the track and pad with thin layer of resin still below the copper foil prior to peeling the copper from the laminate and via. With care you can still continue to grind with grit paper to just touch the copper surface which reduces the strain on the copper during peeling, you must not expose the base of the via barrel

27 Printed Board Assembly Featuring Blind Via Holes Capture pad & track If a test board was used for this can then be peeled assessment it would be possible to Away from the base of the via hole use a bond tester to measure peel force of the track and its separation force Blind Via Connection to Inner layer Bottom of via hole Via hole capture pad 0.020

28 Examples of Via & Capture Interfaces Testing Blind Via Hole Connection The following is a simple concept for a test coupon for quality control of blind via holes. Daisy chain pattern of blind vias tested for continuity

29 Testing Blind Via Hole Connection Daisy chain pattern of blind vias tested for continuity, if they fail after some form of stress like temperature cycling, bending etc. they can be tested visually using back microsection techniques outlined previously. The length of the capture pad or pad to pad separation should be as long as possible to allow cutting and peeling of the track. The width of the track should be equal to or larger than the via capture pad top aid peeling and avoid breaking at the pad to track interface Optical Examination of Via Connection With a true plated interface between the via and pad. M arks on both surfaces should correspond after pad separation. Grind sample board up to the base of the capture pad, peel off the capture pad between blind via holes. Examine the via and capture pad interfaces. Depending on the design of the test coupon it would also be possible to conduct peel tests on via holes

30 Blind Via Test Board Option It may be possible to make a test board from etched double sided laminate. Laser drill the blind vias through etched apertures in the copper foil on side one of the board down to the back of the copper. Then peel the copper off for inspection. In this way no grinding of the board is required. What is Pad Cratering Pad cratering is the partial or complete separation of the copper termination pad, typically seen on area array designs after mechanical strain has been applied. This failure is different to pad lifting when heat is applied during rework. However pad cratering has been reported on board assemblies with large BGA packages directly after reflow soldering and assumed to have occurred during cooling Pad lifting has also been seen on boards which have been reworked but this may be due to pop corning of devices. In this case the copper pads were retained on the solder sphere and separated from the PCB. It is easier to see pad cratering on area array devices with optical inspection than x-ray although pad separation and track breaks have been investigated by the author and Dave Bernard with x-ray in one of their technical papers

31 PCB Pad Cratering Different examples of pad separation from the PCB and component package surface PCB Pad Cratering Copper needles provide greater surface adhesion to the laminate surface but this in turn may impact electrical performance of the laminate on multilayer designs. It s a balance between mechanical and electrical performance. The hardness of the epoxy and the solder alloy increases the possibility of separation

32 Possible Cratering Causes Pad cratering seen with lead-free, less ductile alloy Can be related to the PCB laminate, harder epoxy Commonly seen during flex or drop testing Difference in package expansion rates, possible impact of reflow and cooling steps Does not always results in open circuit failure!! PTH/Surface Mount Pad Surface Silver Tin

33 PTH/Surface Mount Pad Surface Nickel gold pad Copper OSP Organic Surface Protector PTH/Surface Mount Pad Surface Solder Levelled

34 Surface Finish Thickness Measurement PTH/Surface Mount Pad Surface Nickel/gold foot is an unusual formation that can lead to shorts or reduction in surface insulation. The photograph shows a gold hallow around the surface mount pads. Nickel has plated onto the surface of the laminate and the edge of the solder mask. This has also allowed gold to coat the surface of the nickel reducing the design gap Microsectioning of the nickel/gold foot clearly shows the plating defect. The nickel can be seen to extend from the bottom of the pad across the laminate and up the solder mask wall

35 Bob Willis - Texas Instruments Workshop Solder wicking Three examples of solder wicking; On the left due to incorrect process parameters in the 80s with vapour phase reflow In the centre due to slow wetting on copper OSP. The copper surface had been exposed to a wash off operation which impacted the solderability of the PCB. The example on the right is on a nickel gold board due to poor wash off practices on a night shift in high volume production Solderability Testing with Wetting Balance

36 Wetting Balance Solderability Testing Wetting Balance Solderability Testing

37 Solder Spot Wetting Test Pattern In the design we use six 0.020" parallel tracks on a 0.040" pitch, track and gap are and are approximately 0.900" long. The gang solder mask image has a clearance of around the pattern. The pattern should be placed on both sides of the board or both sides of a multi panel on the break-out scrap area Solder Spot Wetting Test Pattern

38 Solder Spot Wetting Test Pattern CSP CSP CSP CSP + + BGA BGA BGA BGA CSP CSP CSP CSP + + BGA BGA BGA BGA Impact on PCB Solder Finishes Images from NPL Lead-Free Workshops

39 Solder Paste Stencil Pattern Solder paste pattern consists mm (0.020 ) square apertures, can have slight rounds on corners. The gap between the apertures increased from 0.16 to 1.16mm in 0.05mm increments

40 Tin Whiskers The images above show tin whiskers on the surface of a plated through hole printed circuit board coated with tin. The boards were produced and shipped to a manufacturing site in Europe and, when examined prior to assembly, found to have whisker growth. Tin has become popular on printed boards as one of the alternative coatings, tin has also become the finish of choice in the component manufacturing industry. However, many people have shown concerns over the formation of whiskers. There has been a considerable amount of work and technical articles produced on whisker formation and potential for failure in electronics. There is still a lot of work being undertaken around the world on this subject. The reason is we still do not have guaranteed whisker free products or the process where the chemistry is being used is not being maintained correctly. If you follow guidelines on designing a whisker free process, use materials that should not form tin whiskers, somehow they still appear. This problem is not specifically a lead-free issue as it has been around for years as a possible problem. The increasing use of tin as a component finish and printed board alternative coating has highlighted the potential for failure. Organisations like NEMI and JEDEC have provided guidelines on what causes whisker formation, ways of accelerating testing for whisker formation and preventative strategies Current understanding is tin whiskers form due to stress formed in the plating. This causes tin filaments or single crystals of tin, commonly 2-5 microns long to be forced out from the surface of the plating during the life of the product. To reduce the possibility of formation engineers have looked at baking to reduce stress, changing or increasing the thickness of barrier layers. This is the third example of whiskers examined on different products, each theoretically should not have occurred based on current recommendations and specifications. The other two are related to RF shielding Test Method for Measurement of the Propensity for Conformal Coatings to Inhibit Tin Whiskering Martin Wickham and Christopher Hunt, MAT28, December 2008 Testing PCBs & Components with XRF Engineer measuring surface finish thickness on PCB pads

41 Potential Process Failures Copper dendrites can cause intermittent product failures during product operation. Copper ferns grow from one electrode to another in the presence of a moisture layer on the surface of the board White residues caused by the use of an incorrect cleaning process or chemistry. The flux is not soluble in the chosen cleaning process or some other reaction has caused the flux to become insoluble Printed board failure due to a conductive short formed on inner layers of a multilayer boards generally referred to as a CAF failure. CAF stands for Conductive Anodic Filamentation and related to the materials and conditions in side the board Potential Process Failures Printed board failure due to a conductive short formed on inner layers of a multilayer boards generally referred to as a CAF failure. CAF - stands for Conductive Anodic Filamentation and is related to the materials and conditions inside the board

42 Microsection Inspection Check position of plated through hole centre line during sectioning of the samples Microsection Inspection

43 Microsection Inspection Microsection preparation requires experience and time there are no shortcuts if you want all the information to allow correct interpretation of your results. Two examples of a plated through via 0.2mm after lead-free process trials and then temperature cycling of the board over 1000 cycles must show up some changes Blind Via Laser or Control Depth Drill

44 Microsection Inspection Inner Layer Separation

45 Ionic Contamination Measurement Measurement of contamination on a board is also a matter of considering the total result, the total time and the rate at which the contamination rises over time. Remember the ionic contamination can come from soldering materials, the printed board surfaces, fabrication materials and other sources. Many people are happy that the result shows it to be less than the specification limit Printed Board Quality Control Thermal shock testing with fluidised sand bath and water, left. Five cycles between 25degC & 260degC looking for change of resistance between the cycles. The sample holder and measuring system is shown on the right

46 Printed Board Quality Control Plated through hole Multilayer board Close up of test samples prepared for thermal shock testing, the examples are a PTH and a multilayer circuit. The fingers are used to make the connections in the test fixture for immersion into the water and band bath Printed Board Quality Control Typical example of a crack forming at the knee of the hole on a plated through hole board during testing. This would cause a change in resistance during the test

47 Printed Board Quality Control Two coupons after testing showing different degrees of cracking at the knee of the hole. This would result in a change in resistance SEM Examination Industry example images of pads after stripping courtesy of Shipley

48 Merlin Circuits Stripping Procedure 20 g / Litre DEGUSSA Gold Stripper 645 plus10 g / Litre Potassium Cyanide Confirm the solution is appropriate for the suppliers nickel gold process chemistry Use solution at room temperature Stripping printed board gold samples to be conducted in a fume cupboard with suitable protective clothing e.g. gloves, eye protection lab coat 0.06 to 0.08 microns of Immersion Gold will normally strip in 10 to 12 seconds Take the sample board and gently move back and forth in the solution until the Nickel becomes visible, quickly remove and rinse thoroughly in demin water and the waste collected for appropriate disposal via authorised waste streams Record time taken to strip the Gold and view Nickel surface under suitable magnification. (It should be emphasized that the solution is highly toxic and only used by an experienced technician. Under no circumstances should the cleaner be allowed to come into contact with acids) Pads Before/After Stripping Examples of two surface mount pads before and after stripping, fast stripping time indicated thin gold and a poor surface when inspected

49 Example Pads After Stripping

50

51 Blow Holes & Pin Holes in Solder Joints Blow Holes & Pin Holes in Solder Joints

52 Testing PCBs for Outgassing Testing PCBs for Outgassing The video clips show testing of a PCB in production for outgassing. Originally this was suggested as a test method with some criteria based on the number of bubbles or rate of outgassing. One engineer even built a machine for counting the bubbles

53 Copper Pad Erosion with Lead-Free Copper Pad Erosion Although examples of copper erosion have been highlighted in the industry there is little evidence to date of this being an issue. In the case ofsingle sided boards the apparent erosion may have been due to preparation of the copper for OSP treatment. In this case copper is mildly etched, excessive prep may have removed more copper around the pads as other areas of the tracking would be protected by the solder mask. Where mechanical cleaning is used and incorrectly controlled the copper can be reduced around the hole leading to a apparent copper reduction. Further investigation of the problem and the examples circulated in the industry will be further reviewed and where appropriate further trials conducted. It is interesting to note that the defects highlighted have not been shown to cause failures. At first examination we would consider them all rejectable based on our existing knowledge of tin/lead joints. That knowledge is again not necessarily based on failures but the inspection criteria for solder joints in circulation today. Perhaps we do need to re-look at some of the visual criteria we use in industry for lead-free? Recent trials have been conducted on selective soldering systems with lead-free alloys. In this case where the boards did have a very thin copper plating and been solder levelled with lead-free the copper removal was significant. After exposing the boards to a high temperature during selective soldering for an extended time copper pads were full dissolved from the surface of the board. This is not typical and should not occur when a sound plated layer is present in a well controlled lead-free assembly facility. However it does demonstrate what can happen Outgassing from PCB Via Holes Oil being used on the surface of the vias and the sample board heated to simulate reflow. Even with epoxy and copper plating outgassing can still occur during reflow soldering leading to voiding under large packages

54 Potential Process Failures Copper dendrites can cause intermittent product failures during product operation. Copper ferns grow from one electrode to another in the presence of a moisture layer on the surface of the board White residues caused by the use of an incorrect cleaning process or chemistry. The flux is not soluble in the chosen cleaning process or some other reaction has caused the flux to become insoluble Printed board failure due to a conductive short formed on inner layers of a multilayer boards generally referred to as a CAF failure. CAF stands for Conductive Anodic Filamentation and related to the materials and conditions inside the board CAF Failure in PCB

55 PCB Delamination PCB Delamination Company found delamination of boards in production after reflow soldering. Moisture in the board was blamed for the failure but it was not the root cause. Moisture is in every board, baking does eliminate it for a few hours but also has it disadvantages. A simple destructive examination was conducted on a section of the board, the blister was cut with a sharp knife and the section removed from the surface of the board. Looking at the copper surface it is clear that no bond had ever been made between the resin and the copper. This was probably related to the copper preparation PCB Delamination PCB Delamination Company found delamination of boards in production when using Thermount a non woven glass laminate. Moisture in the board was blamed for the failure but it was not the root cause. A simple destructive examination was conducted on a section of the board, the blister was cut with a sharp knife and the section removed from the surface of the board. Looking at the copper surface it is clear that no bond had ever been made between the resin and the copper. This could have been caused by the copper preparation or the press cycle which is a higher temperature for this material.

56 PCB Delamination Through Via Failure

57 Open Intermittent Circuit Solder Mask Lifting

58 Solder Mask Lifting Printed Board Quality Control Dye pen testing of solder mask

59 Printed Board Quality Control Dye pen testing of solder mask Printed Board Quality Control IPC-TM , called by IPC SM-840 4B-3B-2B-HB-F-H-2H-3H-4H-5H-6H

60 Basic Printed Board Manufacture Test board for the smallest chip capacitors and resistors available in industry today next size down from NPL Process Defect Database

61 NPL Process Defect Database NPL Process Defect Database

62 NPL Process Defect Database Do you have any questions?

63