Loading... Please wait...

Connector Intermateability Article

Posted

INTERMATEABILITY OF SMA, 3.5 MM AND 2.92 MM CONNECTORS

SMA, 3.5 mm and 2.92 mm connectors make reliable RF components and instrumentation possible. These seemingly mundane devices are the result of evolutionary and revolutionary design covering a 40 year period and involving several corporations. In their present state they represent highly evolved well defined interface systems.

It is not widely understood that these connector types are mechanically intermateable; meaning SMA, 3.5 mm, and 2.92 mm connectors of good quality and in good condition will intermate across connector families without damage. But what are the electrical ramifications of intermating of these connector types? Can they be used interchangeably without an appreciable difference in performance that is, will a significant impedance discontinuity occur when mixing types? Were they intended to be intermated in the first place? To properly answer these questions calls for a brief examination of the history and motivations behind these connectors.

THE SMA CONNECTOR

The SMA connector first appeared in the late ’50s as the “BRM,” manufactured by the Bendix Scintilla Division. In the ’60s it was popularized as the “OSM,” manufactured by Omni Spectra. In 1968 it received the “SMA” (Sub-miniature A) designation that we know today. The SMA connector uses a solid dielectric interface as opposed to an air dielectric. By definition, an air interface connector cannot have the designation “SMA.” Performance is rated to 18 GHz, but higher frequency variants are available.

The SMA was designed as a miniaturized economical connector for system application. It was never intended to be a precision connector for the laboratory. As it is only rated for 500 mate/de-mate operations it was designed for use on semi-rigid cable assemblies and their board or chassis mates. 

THE 3.5 MM CONNECTOR

The 3.5 mm connector is rated to 26.5 GHz, with a theoretical upper operating frequency of 34 GHz. It is specified to possess a minimum of 3000 mate/de-mate cycles per the IEEE P287/D3 standard.

The 3.5 mm connector was designed to be a precision interface for calibrated measurements of SMA- equipped devices It was created as a test connector for the SMA. As a result, an SMA-to-3.5 mm interface produces better results electrically than an SMA-to-SMA interface. The reason is due to an air gap that is formed between the solid dielectric interfaces of a mated pair of SMA connectors creating an impedance discontinuity.

With this information, it can be seen that an SMA-to-3.5 mm mated interface, using good quality connectors, is an acceptable (and intended) practice where adverse electrical performance through 18 GHz should not be expected.

THE 2.92 MM CONNECTOR

The 2.92 mm geometry with an SMA mateable interface was developed to provide coaxial connector performance to 40 GHz. In the mid- ’70s, Maury Microwave introduced the MPC3 connector using the aforementioned 2.92 mm geometry. Without an abundance of available instrumentation operating at 40 GHz, it found little usage. In the early 1980s, Weinschel Engineering utilized this geometry in an engineering design under a Department of Defense (DoD) contract.

Simultaneously, Wiltron Co. began a program to produce instrumentation operating to 40 GHz and would therefore use the 2.92 mm connector. The connector and instrumentation were introduced in 1983 by Wiltron (now Anritsu Corp.); the term “K-connector” was trademarked by Wiltron making reference to the connector’s frequency band of operation, the K-band. Intermateability with the SMA was not an original design objective for the 2.92 mm connector. The ability of these two connector types to intermate grew out of convenience; the 2.92 mm was based upon proven SMA geometry.

BASIC ASSUMPTIONS ESTABLISHED

Now that the basic relationship between these three connector types has been explained, characterizing the electrical performance of these connector types when inter-mated can be addressed. It has been established that a 3.5 mm-to-SMA mated interface is an acceptable and intended practice per the 3.5 mm connector design. All that remains is to demonstrate the electrical performance of a 3.5 mm-to-2.92 mm mated interface and a 2.92 mm-to-SMA interface.

A FEW WORDS REGARDING WEAR

The premise of intermateability is predicated upon connectors that are in serviceable condition. Misaligned center contacts or contact heights out-of-spec, worn outer conductors— especially in SMA designs—invite damage and a reduction in performance. Mating interfaces must be inspected and cleaned regularly. Pin height gauges are not reserved for metrology-grade applications; they are a good idea for anyone working with these connector types in frequent mate/de-mate scenarios.

ELECTRICAL PERFORMANCE AT THE CONNECTOR INTERFACE

Across the SMA, 3.5 mm and 2.92 mm connector types, a governing factor in electrical and mechanical performance is center contact height—a measure of protrusion or recession of the center contact with respect to the connector reference plane. Depending upon whether the connector is a LPC (Laboratory Precision Connector)-based or GPC (General Precision Connector)- based design, the center contact height has a prescribed allowable range; temperature and wear can impact variation within this range. Center contact height is a compromise between good mechanical performance, that is, non-destructive contact with its mating connector, and good electrical performance. But how does center contact height influence electrical performance?

Although there are other areas of potential variation, connector center contact height variation plays a major role in defining interface electrical and mechanical performance. Good quality 3.5 mm, 2.92 mm and SMA connector those whose interfaces are produced in accordance to the IEEE 287 (3.5 mm and 2.92 mm) and MIL-STD-348 (SMA) specification will have their critical dimensions constrained. These specifications/standards serve to support the practice of intermating these connector types. It is important to note that the IEEE 287 and MIL-STD-348 specifications address interface dimensions only; they do not touch upon design issues associated with a connector’s “back end” or cable interface side. This issue is to be resolved by the manufacturer.

PERFORMANCE QUESTIONS ANSWERED

An experiment was devised to investigate the impact of center contact height (mated contact gap) variability on interface electrical performance

within mixed mated pairs of connec- tor types (3.5 mm-to-2.92 mm and 2.92 mm-to-SMA, for example).

To establish a reference, electrical performance vs. center contact height was examined using a mated pair of 3.5 mm connectors. Single-port VNA measurements through 26.5 GHz were made with LPC grade 3.5 mm connector adapters coupled to a 3.5 mm sliding load from an HP85052B 3.5 mm calibration kit. The adapter/ sliding load arrangement is shown in Figure 4.

A sliding load was selected for two reasons: (1) it contains a sufficiently long 50 Ω air line section that will fa- cilitate gate placement for time do- main gating operations used in the data analysis portion of the experiment; (2) the center contact height of the sliding load 3.5 mm interface is variable via an adjustment screw lo- cated at the rear of the sliding load. This allows for precise control of the gap between the 3.5 mm socket and 3.5 mm pin contacts when mated.

The experiment consists of the following steps: • Using the mated pair of 3.5 mm connectors as an example, one end of the 3.5 mm LPC grade socket adapter was threaded into port 1 of the VNA • Using a center contact height gauge, commonly called a pin height gauge, the sliding load interface-pin type was gauged and center contact height was adjusted to 0.0000 inches. •The sliding load was then threaded onto the LPC 3.5 mm socket interface and tightened to the appropriate torque specification. The result is a 3.5 mm socket-to-pin mated interface •An S-parameter measurement was made of the 3.5 mm mated interface; the resulting S11 data was recorded for subsequent examination of the mated interface’s VSWR and impedance. The sliding load was removed from the 3.5 mm socket interface and contact height was lowered by 0.001 inches, as measured with the contact height gauge •Repeat Steps 3 and 4 up to a contact height of –0.005 inches; the negative sign indicates a recessed condition, i.e., a position below the connector reference plane

PERFORMANCE QUESTIONS ANSWERED: SUMMARY OF FINDINGS

A rise in impedance occurs when terminating the mated interface pairs with a sliding load. This rise is due to a limitation inherent to the sliding load—the load’s “cut-off frequency.” The sliding load depicted in Figure 4 is recommended for use between 2.5 and 26.5 GHz. Below 1.8 GHz, the load displays an impedance of less than 50 Ω. At very low frequency, near DC, the sliding load is a short. This irregularity causes reflected im- pedance measurements to rise with time, and the VSWR to tend towards a value slightly above 1.00:1 as the frequency tends to zero (see Appen- dix A). With this in mind, the impedance and VSWR data are accurate in that they faithfully portray the behavior of a mated pair interface under the noted load conditions.

VSWR: In all three mated interface combinations (3.5 mm-to-3.5 mm, 2.92 mm-to-3.5 mm and SMA- to-2.92 mm), the VSWR never exceeded 1.054:1 through 26.5 GHz, over the range of experimental center contact gaps. The 3.5 mm and 2.92 mm-to-3.5 mm mated interfaces produced very similar VSWR results over the range of gaps, having a nearly identical spread in maximum VSWR values: 1.005:1 to 1.050:1 through 26.5 GHz. The SMA-to-2.92 mm mated interface produced a much tighter, but overall higher spread of maximum values: 1.038:1 to 1.054:1 through 26.5 GHz.

Impedance: Impedance was examined via a time domain step response. As expected, the mated interface impedance closely mirrored VSWR on all mated pair combinations. Out of the mated pair combinations tested, the 3.5 mm mated interface produced the most ideal response having a nearly flat transition occurring at a contact gap of 0.12 mils (0.00012"). A close second in terms of ideal response was the 2.92 mm-to-3.5 mm mated interface, producing a similarly flat transition at a contact gap of 2.65 mils (0.00265"). In third place was the SMA-to-2.92 mm mated pair, providing its best impedance transition performance at a 1.50 mil (0.0015”) contact gap.

CONCLUSION

The experiment’s purpose was to investigate the influence of center contact gap variations upon a mated pair’s electrical performance; specifically, high frequency electrical performance. Gap variability was accomplished by varying the center contact height of one connector within the mated pair. By systematically changing the gap, real-life mated connector pair performance across mixed interface types was modeled. Although differences in performance were noted, no significant advantages or disadvantages in electrical performance were observed across the mixed interface mated pairs. In short, the intermixing of 3.5 mm, 2.92 mm and SMA interface types was not found to produce adverse performance and significantly different results when using connectors made by a reputable manufacturer.

The 3.5 mm and 2.92 mm-to-3.5 mm mated interfaces produced somewhat better results compared to the SMA-to-2.92 mm mated interface. However, these results are the product of tightly controlled contact heights; center contact heights on the order of -0.00012" are not the norm and are associated with costly, special purpose LPC connector types. The majority of applications employing 3.5 and 2.92 mm connectors utilize “test grade” versions of these connector types, that is, a grade of connector where center contact height is held at lower levels to accommodate frequent handling and use over temperature, thus wider mated center contact gaps will be encountered.

Test grade connectors can be made to comply with many of the IEEE 287 standards, but in specific areas, designers may choose to depart from these criteria for the sake of ease of manufacturing and/or cost concerns. In the end, the performance differences between these interface combinations are indeed very small even under controlled conditions. Under non-controlled conditions, such as those that prevail in all but the most demanding metrology applications, these differences become insignificant. ■

ACKNOWLEDGMENTS

The author extends his thanks to the following individuals without whose help this technical note would not have been possible: Harmon Ban- ning, technologist, W.L. Gore & Associates Inc., Thomas Clupper, senior electrical engineer, W.L. Gore & Associates Inc., Jon Hvitfelt, RF connector designer, W.L. Gore & Associates Inc. and Marc Maury, president, Maury Microwave Corp.

References

1. Institute of Electrical and Electronic Engi- neers Inc., IEEE P287/D3 – Provisional, Standard for Precision Coaxial Connectors (DC to 110 GHz), July 2005.

2. Maury Microwave Corp., Application Note 5A-011, Improving SMA Tests with APC3.5 Hardware, November 18, 1999.

3. Maury Microwave Corp., Application Note 5A-021, Microwave Coaxial Connector Technology: A Continuing Evolution, De- cember 13, 2005.

4. Agilent Technologies Inc., Application Note, General RF Connector Information, December 2000.

5. Department of Defense, United States of America, Interface Standard for Radio Frequency Connector Interfaces, MIL- STD-348A, April 20, 1988.

6. Department of Defense, United States of America, Performance Specification: Gen- eral Specification for Connectors, Coaxial, Radio Frequency MIL-PRF-39012E, April 27, 2005.

Author: PAUL PINO

W.L. Gore & Associates Inc. Elkton, MD