There are many misunderstandings regarding how the type of manufacturing technology used in the assembly of printed circuit boards (PCBs) relates to automated test. To dispel some of the confusion, it’s important to answer a couple of questions. Is surface-mount technology (SMT) more difficult to test than through-hole technology (THT)? What roles do in-circuit testers (ICTs) and manufacturing defects analyzers (MDAs) play in testing these diverse technologies? Let’s examine the facts.
SMT vs. THT
Many circuit assemblies still are being manufactured using THT. This is a result of several factors:
- The circuit was designed prior to the advent of SMT.
- Circuit components are not available in SMT.
- The manufacturer doesn’t have SMT equipment to manufacture the assemblies.
- Repair is easier for THT assemblies.
- Simple circuits can be assembled less expensively with single-sided THT.
- Parts are available in smaller quantities for THT (no need to purchase reels of parts).
- Assembly of small lots is less expensive with THT than SMT.
With the recent industry-wide parts shortages, there have been reports of manufacturers redesigning SMT assemblies to be THT assemblies so parts can be purchased more readily. Another strategy is to design circuit boards to accept both SMT and through-hole parts so that the most readily available parts can be used. Because of these reasons, it doesn’t appear that THT will be going away in the immediate future.
Implications of Assembly Technology on Electrical Test
Electrically, a component is a component, and automated test equipment (ATE) can test it equivalently whether SMT or THT attaches it to the board. So, what’s the big deal about the implication of SMT vs. THT and test? There are at least three, all having an impact on the testing process:
- The fault spectrum changes.
- The method of test fixturing can be impacted.
- Repair costs are higher for SMT.
Fault Spectrum Changes
Opens and shorts are the most predominant failure mechanisms for PCBs (Figure 1, see the August 2001 issue of Evaluation Engineering). When assembly technology changes from THT to SMT, the fault spectrum tends to change from a predominance of shorts to a predominance of opens.
It’s tough to create an open connection in THT unless the flow process is not reaching part of the assembly; there are contaminants on the assembly, or a lead on the part is bent during the insertion process. Otherwise, the connection to the part is good. The bigger problem is that solder will bridge from one component lead to the next to create a short.
When soldering using SMT technology, the most significant problem is likely to be an open connection. This typically results from insufficient reflow. Since it is more difficult for manufacturers to catch opens with ATE, some tune their process toward using additional solder to make the manufacturing process lean more toward shorts than it would if left in the neutral state.
Test Fixturing
Testing both SMT and THT is done with bed-of-nails test fixtures. These fixtures use spring probes to contact the unit-under-test (UUT) on one or both sides to make an electrical connection to each node or net on the circuit. A node is one circuit on the assembly such as GROUND or ADDR0. With electrical contact to each net, the tester can access every component on the UUT for test and find all shorts.
The industry-standard spring probe for many years has been the 100-mil size. This is a result of typical THT probes for the vast majority of devices that have leads on 0.100² centers. As SMT necessitated smaller spacing, 75-mil probing was developed, then 50-mil, then 38-mil, and now even smaller spacing probes are available.
Features for THT boards generally are very inexpensive and reliable while fixturing for SMT boards can be nightmares if the boards are not designed with test in mind. There is no question that this leads to high fixture costs, but it also makes it extremely difficult to make highly reliable, long-lasting fixtures. Also, at times, not all of the nodes are accessible, compromising test coverage.
To remedy these problems, most circuit designers include some testability throughout the design and layout of the UUT. The best case has a through-hole lead, a test pad of 30- to 60-mil dia, or a via on one side of the circuit for each net. When test pads are included in the design, X-Y locations will be specified, making fixture layout and fabrication less expensive.
Dual-sided probing is a proven technology and has been used for a number of years. If access to both sides is necessary, it can be reliably provided; however, it does add appreciably to the cost of the customized test fixture.
As a result, with some forethought, the future for an SMT board can be just as inexpensive and reliable as a fixture for a THT board. Most manufacturers evolve to this methodology as they mature in the SMT design process. The minor additional expense up front makes life a lot better for all involved in the manufacturing and support of the board.
Repair Costs
Even though repairing THT PCBs isn’t necessarily a picnic, the time and cost required to test and repair SMT assemblies are higher. Troubleshooting is more difficult since pins can’t be isolated easily. Once the fault is found on the SMT assembly, specialized tools and expertise are needed to do the job, and repairs take longer.
As a result of this extra cost, it generally pays to perform as much test as you can on the assembly to find the faults early, especially before power-up when one failure can destroy other parts in the circuit. For boards that are inexpensive, it can be cost-effective to use multiple test passes during the process so boards can be discarded with the minimum number of parts installed.
Normal Test Processes
After the bare board is tested, some basic tests are performed to ensure that the board is assembled correctly. These tests are the same whether using SMT or THT.
First, the board is tested for opens and shorts. Generally, this is done using a relatively low limit, such as 10 W, so the threshold is below the value of most components on the board but high enough to eliminate false failures from probe and wiring resistance that may be present. Then the UUT is tested for resistance, capacitance, inductance, diode junctions, and active part operation.
These tests are standard for both ICTs and MDAs and generally will find the vast majority of faults, complete with good diagnostic capability. For example, the tester might report that R101 is 10.050 kW but should have been between 980 and 1.020 kW—the fault of having a wrong component installed.
The MDA tester, often called an analog ICT, assumes that the components are good and endeavors to find how they are improperly installed on the PCB. ICTs can take the testing a step further by checking ICs in the circuit for operation. Taking the step from MDA to full ICT improves the test coverage by a few percent, but at multiple times tester/test fixture development cost.
What’s the SMT vs. THT Problem?
So, what’s the difference in testing SMT vs. THT? Let’s take a look at the difference in the fault spectrum of mostly shorts to mostly opens.
For most components, it makes no difference at all. Testing either SMT or THT will find all of the shorts unless SMT doesn’t allow full nodal coverage. For detecting opens, either technology allows testing to find most opens. For example, when testing a resistor or capacitor, if there is an open, it will be detected when the part is measured.
The difference is finding open connections in ICs. For the typical test, a mapping is made of the diode junctions present between IC pins and the power/ground rails of the UUT. This works fine most of the time, but in many cases, an IC is connected to both ends of the trace. This looks like two parallel diodes to the test system. If one is missing due to an open connection in the IC, the tester will miss it since it will measure the other diode that is in parallel, not knowing the difference.
What’s the SMT vs. THT Solution?
Various methods have been devised to detect this class of faults. Some of the techniques include modeling the diodes in the device as a transistor or sensing current out of other pins as current is sourced into the device. While these techniques work well in some static cases, they can break down when the part changes from one vendor to another.
Another technique is to use full ICT for the ICs. This is effective for finding faults but expensive as a result of initial purchase price and programming costs for each UUT.
If the board being tested is designed with boundary scan technology (BST), a special serial bus is used within ICs or around the UUT’s circuits that allows a general-purpose BST test controller to stimulate circuits and read back the resultant outputs (Figure 2, see the August 2001 issue of Evaluation Engineering). This is a UUT power-up technique and needs to be designed into the board for effective testing—but can allow effective test with a minimal number of access points where high-circuit densities preclude full nodal access.
Another technique called vectorless test, such as the Agilent TestJet technology, the Teradyne FrameScan/Wavescan, and the GenRad Opens Xpress, uses a special top probe over the IC in conjunction with the other standard probes already present (Figure 3, see the August 2001 issue of Evaluation Engineering). By providing a stimulus and measuring the coupling through the IC, open connections can be accurately detected. Each vendor’s method is different and has its plusses and minuses, but all are similar in their ultimate purpose and use.
Vectorless test can be used effectively even with IC vendor changes. It doesn’t require the expensive programming of full ICT techniques. In some cases, such as with Agilent TestJet technology, the hardware/software algorithms can even be extended to detect reversed capacitors—a tough testing problem otherwise.
Another solution to the SMT open-connection quandary is to ignore it at the circuit-board test level. In some cases, the manufacturing process may be controlled well enough that the number of undetected faults may be below the tolerable fault level. In other cases, the circuit may be functionally tested in final assembly to determine if any undetected open faults are present. If there is a fault, an open connection is less likely to cause damage than most other types of circuit faults.
Summary
Testing methodologies often are considered as being drastically different when testing SMT as opposed to THT. We have seen how these differences can be very minimal, in particular when some thought has been given to the testing process during design of the circuit board. The difference in fault spectrum caused by manufacturing technology can be insignificant enough to be ignored in some cases or accommodated via techniques such as vectorless test.
About the Author
Ken Hallmen, P.E., is the marketing manager at CheckSum. Previously, he held engineering, managerial, and marketing positions at John Fluke Manufacturing and Tektronix. Mr. Hallmen has a B.S. and an M.S.E.E. from the University of Washington and an M.B.A. from Seattle University. CheckSum, 6120 195th St. NE, Arlington, WA 98223, 360-435-5510.
Return to EE Home Page
Published by EE-Evaluation Engineering
All contents © 2001 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher
August 2001