Solder paste analysis: Maintaining harmony between OEM and EMS
AT A GLANCE
- Analysis of manufacturing defects shows that solder problems are largest single cause of failure. Many of these soldering failures can be traced back to poor housekeeping of solder paste
- Whilst there is not yet an IEC or IPC standard for open time testing, IEC 61189-5 and IPC-TM-650 do include comprehensive tests for slump, tack, solder balling and wetting (and in the case of the latter, a test for spreading)
- Gen3 Systems has introduced a Solder Paste Analyser that performs all the tests detailed by the IEC and IPC in a single machine. Now manufacturers can check the quality of solder paste quickly and efficiently on the shop floor prior to use
ARTICLE
Virtually overnight, the reputation of one of the world’s largest and most famous toy manufacturers was on the line.
If the recent woes of US toy manufacturer Mattel’s demonstrate one thing it’s how even well managed companies can rapidly find themselves in hot water when the manufacturing process goes wrong.
As widely reported in the press, one of Mattel’s Chinese subcontractors ignored the company’s precise specification and used a lead-based paint to coat die cast toy cars. Mattel’s quality control procedures soon flagged up the problem, but too late to prevent nearly 500,000 cars hitting the department store shelves1.
While the short term cost of rectifying the problem wasn’t insignificant, the cost of the longer term damage to Mattel’s previously impeccable brand image and reputation will be much more difficult to quantify. Mattel may well emerge unscathed, but only time will tell and how well the company manages its way through this crisis.
While perhaps the highest profile name to be hit by such problems, Mattel is far from unique. And if something as simple as the paint on a toy car can go wrong, imagine what could happen to a complex, high reliability electronics assembly. Moreover, when things turn sour neither the OEM nor EMS is likely to hold their hands up willingly leading to an expensive and morale-sapping investigation.
But this is exactly the risk run by Western-based OEMs when seeking to reduce manufacturing costs by offshoring to lower cost regions such as Eastern Europe, China, India, and more recently Russia, Brazil and Vietnam.
The logic is sound: the OEM can concentrate on what it’s good at – designing, marketing and improving innovative products – while the assembler – either a sister company or an EMS – invests in the capital equipment, processes and engineering expertise demanded by six-sigma manufacturing.
A match made in heaven then… until something goes wrong. Offshoring to a manufacturing facility halfway around the world makes the manufacturing process doubly difficult to control. And if a company like Mattel can find itself in trouble, then any company can.
In the electronics industry, analysis of manufacturing defects shows that solder problems are largest single cause of failure. And with surface mount technology now the dominant assembly method, many of these soldering failures can be traced back to poor housekeeping of solder paste creating problems during printing, placement and reflow.
The International Electrotechnical Commission (IEC) and Association Connecting Electronics Industries (IPC) standards bodies have quantitative test standards for solder paste (although the current standards only cover tin/lead pastes). Publication of equivalent standards for lead-free alloys is planned for 2009). IEC 61189-5 and IPC-TM-650 include four and five solder paste characteristic tests respectively.
These tests provide a framework for ensuring the solder paste is up to the job but are carried out by the solder paste manufacturer prior to despatch. However, until now, the tests have been complex and time consuming, requiring a separate machine for each test. But a new generation of all-in-one machines allows a semi-skilled operative to completes the entire suite of tests in 15 minutes. Now, by checking the characteristics of the paste before use manufacturers can minimise defects that may otherwise manifest themselves in the field – when rectification is very expensive and damage to reputation is painful.
Paste problems
Modern solder pastes and stencil printing equipment have advanced dramatically to meet the demands of high volume or high mix manufacture of fine pitch, densely populated electronics assemblies. Paste manufacturers go to great lengths to ensure their products meet the specifications laid down in the IEC and IPC standards.
However, solder paste is a complex combination of solder powder, flux vehicle and other additives that are there to enhance the paste’s performance during the printing process (see sidebar “Anatomy of a solder paste”) and the in the time between manufacture and use, paste properties can change.
Worse still, once the paste is opened and “conditioned” (allowed to reach room temperature and stirred to ensure a homogenous mix of the paste’s constituents) it begins to deteriorate rapidly. This deterioration is due to evaporation of solvents and increased oxidation of the solder spheres due to the air that has been introduced during mixing. The period during which the paste is acceptable for production is known as the “open time”.
Paste that has exceeded its open time, or is in poor condition due to inadequate storage or storage beyond its shelf life will cause faults during printing, placement and reflow. Defining open time is critical for ensuring that identical products made in separate factories enjoy the same quality levels. That’s not to say that open time will be the equal in every location. A specific brand and type of solder paste may have a shorter open time if, for example, it is used in a warmer or more humid environment.
Apart from the open time, the paste has other key properties that determine its performance at various stages of the manufacturing process. The key property during printing is slump. Slump describes a physical property of solder paste which determines how the cross sectional profile of paste printed on a pad “relaxes” after printing. A degree of slump is inevitable, but excessive slump can allow the paste to bridge the gap between adjacent pads, especially on fine pitch PCBs.
Tack is the property of the paste that is most relevant to placement. Tack is a measure of the stickiness of the paste and is essential to ensure that placed components stay put during the placement cycle and subsequent handling prior to reflow (see figure 1). If the paste has insufficient tack, components fall off the PCB or become misaligned, or can tombstone during reflow. Each of these faults results in expensive rework. Maintenance of tack after the paste is applied to the board (tack decreases as the solvents in the solder paste evaporate) is essential to allow a reasonable time between placement and printing.
Figure 1: Solder paste samples during tack testing
Other assembly faults due to substandard paste don’t manifest themselves until during or after reflow. Worst still, some of these faults may not appear until the product is in the field when the cost of rectification has increased a thousandfold compared to repair at the automated test station. (Think back to the example of Mattel.) These faults are excessive solder balling and spreading, and poor wetting.
Solder balling is a phenomenon caused by lack of heat and oxidation and can result from insufficient cleaning of the surface to be soldered by the flux (i.e. poor solderability) and insufficient heating of the solder paste. As the solder starts to melt small balls form around the joint but as the rest of the solder melts the solder balls fail to coagulate with the main mass. Later, the solder balls can break away from the joint and cause short circuits.
Wetting determines the integrity of the solder joint. Molten solder introduced to a contaminated or oxidised copper pad fails to wet properly and forms a weak or “dry” joint. A weak or dry joint can fail immediately or may survive for years in the field until failing catastrophically. One of the flux’s tasks is to remove contaminants and oxides to promote good wetting.
Excessive spreading occurs when the flux is too active and can cause two faults: solder reflowing over the edge of the pad, and excessive residues on the boards. The flux’s activity is a balance between ensuring the solder wets properly and leaving excessive contaminants on the PCB. Insufficient activity limits cleaning and deoxidisation and residues can cause problems for automatic test equipment (ATE) by clogging test probes and causing phantom faults. Moreover, excessive residues can cause electrochemical reactions in the presence of moisture and a bias voltage resulting in short circuits in the field.
An all-in-one test system
Whilst there is not yet an IEC or IPC standard for open time, IEC 61189-5 and IPC-TM-650 do include comprehensive test for slump, tack, solder balling and wetting (and in the case of the latter, a test for spreading) (see sidebar “Defining the tests”). Committees made up of independent technical experts from OEMs, EMSs and equipment manufacturers have devised the tests. They represent a qualitative benchmark for determining solder paste suitability with good Gauge Repeatability & Reproducibility (Gauge R&R).
Fortunately, equipment manufacturers have eased the process by introducing test equipment that automates the individual tests, allowing semi-skilled operatives to perform them and saving time. Nonetheless, five pieces of test equipment represent a lot of capital equipment outlay and require plenty of expensive real estate on the shop floor.
But now, one manufacturer, Gen3 Systems, has released a single integrated machine that performs all the tests detailed by the IEC and IPC, plus an open time test. Dubbed the SPA1000 Solder Paste Analyser (see figure 2), the machine promises to improve quality, increase process yield and reduce inspection and rework. The SPA 1000 performs slump, solder ball (see figure 3), tack (see figure 4) and wetting in accordance with IEC 61189-5 and IPC-TM-650 and adds an IPC-TM-650 spread test (see figure 5) and the manufacturer’s own open time test.
Figure 2: SPA 1000 solder paste tester
Figure 3: Samples ready for microscopic inspection for solder balling after reflow
Figure 4: Tack testing graphical outputs. The SPA 1000 provides traceability for ISO 9001 QA procedures
Figure 5: The SPA 1000’s testing suite includes the IPC’s recommended spreading test
The open time test is performed by printing samples using a just-opened container of solder paste at regular intervals (for example, every 30 minutes) and conducting a tack test. When the tack force drops below an acceptable limit the paste is deemed unsuitable for use. The time between the first test and the test where the paste no longer meets the tack force requirement is the open time. Once the open time is exceeded during production, the solder paste is discarded.
The SPA 1000 runs from a Windows™-based PC running either Vista™ or XP™ and presents results via a familiar graphical user interface (GUI) with 10 and 30x magnification and image capture (see figure 6). The machine can be used by a semi-skilled operative in around 15 minutes and takes up a fraction of the space of the individual test equipment required for all six tests. It provides good Gauge R&R and is simple to calibrate.
Figure 6: Results are presented via a familiar graphical user interface (GUI) with 10 and 30x magnification and image capture
Quality from start to finish
Warranty returns are an expensive business, yet are just the start of the heartache when products fail in the hands of the consumer. The ensuing investigation into who is responsible for the problems is time consuming, undermines morale and deflects effort from making a profit. And it is doubly difficult when, as in often the case in the electronics sector, an external company has assembled the product.
With the majority of these faults caused by solder defects, it is the responsibility of the OEM to specify the soldering process and materials, and of the EMS to adhere to those specifications and be able to prove it in later audits in order to maintain a harmonious relationship.
The IEC and IPC standards provide the ideal quantitative tests with good Gauge R&R so that they can be used as a benchmark in manufacturing sites whether they are in California, Hungary or Shenzhen. Now that Gen3 Systems has introduced a Solder Paste Analyser that performs all the tests detailed by the IEC and IPC in a single machine, manufacturers can check the quality of solder paste quickly and efficiently on the shop floor. These checks will improve first time yield and minimise crippling warranty returns.
1. Reference www.iht.com/articles/2007/08/15/business/15imports.php
SIDEBAR 1
Anatomy of a solder paste(1)
A solder paste (also known as a solder cream) is a suspension of solder alloy (in spherical or ellipsoidal powder form) in a flux vehicle. The diameter of the solder powder spheres and flux are classified depending on size and type. Other agents are added to prevent separation of the powder from the flux and to assist during paste application. In general, a solder paste comprises 90 percent by weight solder powder, five percent resin, four percent solvent and one percent additives.
The most commonly used solder paste for traditional tin/lead soldering is Sn62Pb36Ag2 with the lead-free alternatives adhering to the IPC’s recommended SAC305 alloy comprising Sn96.5 Ag3.0Cu0.5 (SAC305) as defined by J-STD-006. Other tin/lead or lead-free alloys are in use.
In addition to the normal requirements for a flux, solder paste fluxes have extra components which affect the paste rheology and hence how easy the paste is to print or dispense. These additions help to prevent the evaporation of the solvent element of the flux during storage, or while the paste is in use and help to prevent excessive slumping.
During printing air is mixed in with the paste. This air increases oxidation on the surface of the solder spheres. Excessive oxidation rapidly consumes the low flux content of modern solder paste, causing it to fail to reflow properly later in the process. This is the main reason why the “open time” (the time that the flux is exposed to the atmosphere) should be limited, and why paste should not be saved and reused after it has been loaded onto a printer stencil.
During reflow of solder paste heating must be carefully controlled. The pre-heat time is critical and should last long enough to drive off moisture so that the solder doesn’t “explode” at the higher temperatures of reflow (and spread paste over the board) yet not so long that oxides form on the surfaces to be soldered.
1. Reference: “A Comprehensive Guide to the Design and Manufacture of Printed Board Assemblies”, William Macleod Ross, Electrochemical Publications, Vol. 1, 1996, pp 648.
SIDEBAR 2
Defining the tests
The IEC’s solder paste tests (defined in IEC 61189) are slump (5X08), solder balling (5X09), tack (5X10) and wetting (5X11). IPC-TM-650 specifies five equivalent tests for solder paste: slump (2.4.35), solder balling (2.4.43), tack (2.4.44), wetting (2.4.45) and spreading (2.4.46).
The IPC test for slump specifies stencil printing on two coupon with defined pads of various sizes and spacings, then storing one for 10 to 20 minutes at 25 +/- 50C and 50 percent RH +/- 10 percent while heating the other to 150 +/- 100C for 10 to 15 minutes. The heated specimen is cooled to ambient and both specimens are examined for bridging.
The IPC test for solder balling calls for two specimens to be printed with stencils of 0.2 or 0.1 mm thickness and three holes of 6.5 mm or 1.5 mm diameter on a pitch of at least 10 mm. The specimens are then reflowed using a solder bath or hot plate and examined under magnification. The results are compared with visual samples.
The IPC test for tack measures the force required to separate a test probe from the printed paste sample at progressively increased intervals between printing and testing. The probe is applied to the paste at a rate of 2.5 mm/min +/- 0.5 mm/min with a force of 300 g +/- 30 g and then within 5 seconds withdrawn at the same rate. The data is presented in graphical format as the paste “ages”. Typical measurements are time to reach 80 percent of the peak value, peak force in grams with expected variation or time over which the peak force is maintained or for the tack force to decline to 80 percent of its peak value.
The IPC wetting test calls for a copper specimen to be cleaned with liquid copper cleaner, washed with water, rinsed with isopropyl alcohol, dried and then placed in boiling deionised water for 10 minutes before being air dried. The specimen is then printed with solder paste, reflowed and examined under magnification for dewetting or non-wetting.
Spreading is tested under the IPC method by using five circular solder preforms and five copper coupons. The preforms are placed on the coupons and reflowed solder paste flux is then dissolved in reagent grade 2-propanol and 0.05 ml of this solution is dropped into the middle of the preform. By applying to a hotplate or solder bath for 15 seconds, the solder is melted. The flux residues are cleaned from the specimens and the spread (in mm2) of solder is estimated by comparing with reference circles of known area.
(More details on the above tests.)
The open time test isn’t defined by the IPC or IEC. However, a typical test adheres to the following procedure:
- Print specimen and note time.
- Conduct a tack test.
- Note results.
- Repeat steps 1, 2 and 3 every 30 minutes
- When the tack force falls below an acceptable limit (say 30 percent of the peak tack force) note time.
- The open time is the period between the first and last (failure) tack test.
For example, for a particular paste the acceptable limit is 10g. The following results were revealed by the tack test:
- 30g for the first test.
- 32g after 30 mins.
- 34g after 1 hour.
- 30g after 1.5 hrs.
- 28g after 2 hrs.
- 25g after 2.5 hrs.
- 20g after 3 hrs.
- 15g after 3.5 hrs.
- 10g after 4 hrs.
Therefore, the open time is 4 hrs. (Note that different condition will results in different paste performance characteristics.)
FOR MORE INFORMATION
Gen3 Systems distributes the Solder Paste Analyser SPA 1000 in collaboration with alliance partner Sharemate Technology of Taiwan.
If you can’t find the information you need at Gen3 Systems’ website at www.gen3systems.com, then please contact Gen3 Systems direct on:
Telephone: +44 (0)1252 521500, Fax: +44 (0)1252 52 1112, or e-mail.
ABOUT THE AUTHOR
Graham Naisbitt (e-mail) is Managing Director of Gen3 Systems Limited and is a member of the IEC’s TC91 WG3, the working group that formulates test standards for the assembly industry. Naisbitt is also Leader of Solderability Testing Standard IEC 60068-2-69, Co-leader of Solderability Testing Standard IEC 60068-2-54, and Member of IPC-J-STD 002 and IPC-J-STD 003.
Download the PDF (2444 KB)29th Jul 2008
