Designing for impedance testing: maintaining signal integrity on impedance-controlled PCBs is easier when the design accounts for the test method. (Cover Story).

TDR impedance testing of high-speed boards is routine and reliable. For tighter specifications and high-speed performance on standard low-cost base materials, however, automated TDR impedance testing has its advantages. The economics apply whether the need is to test coupons or representative traces on the board. And while testing is one aspect of the equation, it is also vital to understand the production processes most likely to influence yields.

Where high-volume test is required there is a choice between fixture and flying probe systems. At first look, it would seem that a fixture system would offer the best throughput. However, in contrast to bare-board testing (BBT) of opens and shorts, a TDR test takes just under one second for single-ended measurement and around 1.5 seconds for differential. Therefore, move time on a flying probe system is not a great factor in test speed. More important, a flying probe impedance tester can self-verify at the probe tip. This is achieved through the use of built-in traceable air-lines on the test table (FIGURE 1). Fixture-based testers using a matrix of RF coaxial switches cannot self-verify in this way, although in applications where a compromise in repeatability and reproducibility (R&R) is acceptable they are a valuable alternative to flying probes.

[FIGURE 1 OMITTED]

A precision reference airline is the recommended standard for air-line impedance and is calculated as

[Z.sub.0] = 59.939 loge (D/d)

where D is the diameter of the inner conductor and d is the bore of the outer conductor. For more on traceable airline calibration, see Ide (1).

Coupons or Boards

Automated testing can take place on coupons or boards, and both methods have pros and cons. It should be emphasized that, from a fabricator's standpoint, impedance test is all about achieving the desired impedance results with good repeatability. It is inherently easier to take repeatable measurements on a test coupon because the test footprint and ground connection can be carefully defined and the coupon configured for easy measurement. Some companies prefer on-board measurement of either special test traces or the actual production traces (FIGURE 2a and 2b).

[FIGURE 2 OMITTED]

Ultimately, it is up to the designer and fabricator to choose the best method for a particular application. Some board layouts are suitable for on-board testing, others less so. Consider, too, that when a coupon is used as a test vehicle, the impedance test can be run in parallel to bare-board test. This can provide a significant reduction in throughput time, and it also means one less step in the PCB production process (and thus one less chance for damage).

Like all test methods, it comes down to compromise. High volumes and lower-cost boards favor coupons. Onboard test may look better for higher-cost boards in prototype quantities. Both coupon and on-board impedance test can benefit from increased test repeatability when the test is performed on an automated TDR system.

The best fabricators take statistical data from impedance test and feed it back into the production process. Often these data suggest that nominal line widths may need to be altered from the original design. This comes as something of a surprise to PCB designers accustomed to submitting Gerber files to manufacturers which in turn fabricate the boards without adjustments. High-end CAD systems sometimes assume homogenous and perfect materials when calculating line widths; often, these tools provide a "perfect world" calculation of line widths.

In actual production environments, PCBs will vary depending on the source of prepreg and core materials, the glass resin ratio after pressing and the etched trace geometry. Ideally, designers should identify controlled impedance traces separately from other traces of the same nominal width. Some new impedance design systems permit users to model one step closer to real-world processes by employing boundary element method field solvers capable of setting Er on a layer-by-layer basis, and compensating for resin-rich areas between traces (2) (FIGURE 3). If possible, permit front-end engineers to alter the nominal dimensions (within acceptable limits) in order to achieve maximum yields (FIGURE 4).

[FIGURES 3-4 OMITTED]

Orienting Signal Pairs

Whether testing coupons or boards, certain steps should be taken when designing for impedance test.

Testing is simplified if all test signal/ground pairs are oriented the same way. Setup and programming are minimized if footprints are as consistent as possible. Even when automated impedance test is planned, there may be times when the board will also be tested manually. A growing number of OEMs perform sample testing of impedance on new board batches or when they qualify new suppliers. These tests are most likely to be performed with manually operated systems. Board test will be far more successful if a compromise in test footprint is achieved, one which offers good signal integrity and easy location both for manual and automated systems. RF test cables are not very flexible and an operator will take longer if the probe needs to rotate 90[degrees] for each test.

Sample coupon Gerber files are available from Polar Instruments and come in IPC-D-317A (3) or IPC-2141 (4). Also, the updated IPC-TM-650 will include suggested pad and via sizes. In practice, aim for as much standardization as possible, while keeping in mind that many companies demand special probing requirements on a particular application. (Shuttle-mounted, quick-change heads can help minimize setup time when a variety of probing requirements are present.)

Available real-time SPC software is capable of outputting control charts, and running gage R&R permits monitoring of process yield and the capability of the measurement system. Achieving the highest levels of R&R requires an automated system with probe tip impedance verification built into the TDR. Running SPC on a fixture-based system will be equally valuable, especially as extra care is needed for verifying the results at the probe tip on such systems.

In any TDR system, suitable precautions must be taken against ESD. High-frequency measurement systems have proven to be susceptible to static damage, far more so than an open/shorts tester, for example. In an automated system, ESD from the operator is less likely to cause a problem than with a manual system, but static induced by board handling and separation remains to be dealt with. This can be done through static isolation hardware, ionized air or a combination. Ideally, static isolation hardware will pre- and post-discharge the board and keep the tester isolated from the probes when not in operation. Failure to keep control of ESD in an automated TDR may 1) ruin repeatability, as the damage may sometimes be progressive and gradually degrade readings, or 2) be catastrophic, resulting in an inability to test and expensive repair.

Employing precision on-board airline references for regular verification helps flag progressive deterioration and prevent problems. Probe tip measurement verification tracks changes in the measurement system. Probe tip air-lines may also be used for calibration, but require care as regular recalibration at the probe tip can mask progressive deterioration in the cabling and measurement system. The probes on automated testers should be regarded as consumable, and often the system will indicate when probe maintenance is required. The RF performance of the spring pin probe is finite: discard and replace worn probe card assemblies when the system prompts for maintenance. Probe assemblies look simple but are carefully designed for signal integrity; replacing pins with nonstandard parts or surplus pins from a BBT is not a good idea and can result in compromised or inaccurate measurement.

Occasionally there may be some misunderstanding between an OEM and a fabricator over board measurements. When trying to track the source of a variation between an automated TDR system and the design specification of the board, it may be necessary to microsection. This is one of the strongest points in favor of testing on coupons; should the discrepancy prove to be a measurement error, the board is still usable.

Points in favor of automating TDR test:

1 Improves throughput.

2. Reduces reliance on operator skill.

3. Is traceable, on system reference airlines, guaranteeing results.

4. Uses SPC software that maintains control of test equipment and process.

5. Affords fast setup, and reduces need for operator training.

REFERENCES

(1.) J. P. Ide, "Traceability for Radio Frequency Coaxial Line Standards" National Physical Laboratory, 1992.

(2.) J. Alan Staniforth and Martyn Gaudion, "The Effect of EtchTaper, Prepreg and Resin Flow on the Value of Differential Impedance" IPC Printed Circuits Expo Proceedings, March 2002.

(3.) IPC-D-317A, "Design Guide For Electronic Packaging Utilizing High Speed Techniques," January 1995.

(4.) IPC-2141, "Controlled Impedance Circuit Boards and High Speed Logic Design," April 1996.

MARTYN GAUDION is director of sales and marketing at Polar Instruments Ltd. (polarinstruments.com), a supplier of controlled impedance test and design systems, and a contributor to the IPC high-speed, high-frequency standards committees. Prior to Polar, he was responsible for test engineering on high bandwidth portable oscilloscopes at Tektronix. He can be reached at [e-mail].