|Ask the Experts|
February 22, 2012
How did the Class 3 pass/fail criteria in IPC-A-610 come about? Was there field testing and or life cycle testing to support these requirements? If our products need to meet Class 3, but the heel fillets on one gull wing component do not comply with Class 3, but do meet Class 2, and if we do not rework the components, can we meet Class 3 if we conduct additional testing and the product passes the tests? What tests would we need to conduct?
|Expert Panel Responses|
The criteriafor all of the IPC standards are gathered from all over the industry. The datacomes from extensive testing and years of practical experience. To yourquestion of, "can we meet class three criteria through testing?" the shortanswer is, "No". In order for an assembly to "meet the criteria of theIPC-A-610" it must do just that; have observable conditions that falls withinthe scope of the stated criteria. Now that's notto say your customer can't take exception to the criteria. In the opening pagesof the IPC-A-610 you'll find the order of precedence. At the top of the orderof precedence is the customer requirements (contract). If the customer contractsays that the assembly must be built in accordance with the IPC-A-610 criteria,then that is what you need to follow, no exceptions. If the customer statessomething along the lines of, "… built in accordance with IPC-A-610 unlesstesting can prove the reliability of the assembly", then you have room fortesting. If you customerallows testing for exceptions to the criteria, the customer will have tospecify what tests are allowed. If you'd like to beinvolved in the standards development process, Please join us at one of thecommittee meetings held periodically throughout the year. Anyone who uses thecriteria can benefit from knowing where the criteria originates and how thestandards are developed. Check out http://www.ipc.org/Committeepage.aspxfor a list of all the IPC committees and the documents they develop.
Manager of Assembly Technology
The heel fillet requirement has its history in MIL-STD-2000 andprevious versions of military and NASA standards and has been verified by test.Unfortunately, those tests were done so long ago finding the data is probablyimpossible. The heel fillet is accepted as the important part of theconnection’s strength, so there is some concern depending on the expectedoperating environment. Thermal excursions and mechanical stimuli(vibration/shock) will stress the connection more than if the hardware residesin a benign environment. Additional testingcan be a way to reach a "comfort factor", but you must keep in mind that theadditional testing can also shorten the life of the connection.
NASA/Marshall Space Flight Center
This response to thisquestion could take forever; however, let me try to answer it this way. The plated through hole andwire wrap sections are based upon historical and best manufacturing practiceswhich have been in place for the last 50 years. Although the base laminatematerials have changed over the years along with the introduction of lead-freematerial some changes had to be made to the manufacturing processes. Thoselead-free changes were incorporated after many tests were designed and roundrobin tests were conducted by member companies of IPC. Those results weresubmitted to the various task group membership committees, reviewed, expandedand finally accepted and incorporated into the documents. As for the surface mountsection which was introduced in the 80’s in revision A, all those examples weresubmitted with the appropriate test results defining their goodness andreliability. The Class 3 customers and users were part of this acceptanceprocess and if the back up information was not sufficient to satisfy theirrequirements, additional testing was conducted to prove the technology. All ofthis was conducted along with the reliability people and reliability taskgroups who supported the various specifications. It must be kept in mind thetotal quantity of solder joints created during the period of time where thisdocument was created. The major companies contributing to the original documentwere companies like Bell Labs, IBM, Motorola, Martin Marietta, NorthrupGrumman, Lockheed Martin, Hughes, JPL, NASA, Digital Equipment, Raytheon,Harris, Collins Radio, Rockwell, Texas Instrument, Tektronix, Boeing, AT&T,Northern Telecom, General Electric, Westinghouse, Magnavox, Honeywell, TraceLabs, Litton Guidance, E.I Dupont, Sanders Associates, Singer, Naval Avionics,Canadian Marconi, Nelco, Enthone, Unisys, Sandia Lags, SCI, Pace, Hexacon,Lawrence Livermore Labs, Computing Devices, plus many more. The number of solder joints created and the technologyprovided by these companies, has been the baseline for the 610 specifications. So the bottom line is, thework was done to prove the point of reliability and goodness. Nowto answer the question as to what happens when one lead does not have thefillet height to meet the requirement of class 3 which is a heel fillet whichis equal to thickness of the lead. Is this lead reworked or is it dispositionedto use as is? This can only be answered by the customer who has the knowledgeof where the product is going to be used and the environment where the productwill be used. Would I generate a test to check it out would depend upon howmuch money do you have. We can tests these solder joints to failure and stillnot know how they will survive in their operational life. My recommendationswould be to check the manufacturing process, measure the amount of solder pasteapplied to those locations, define the solderability of the components, and ifthese things are acceptable then I would use the component as is with someongoing monitoring.
Vice President, Technical Director
|Submit A Comment|
Free Newsletter Subscription
Circuitnet is built for professionals who bear the responsibility of looking ahead, imagining the future, and preparing for it.
Insert Your Email Address