Changes in insertion loss with temperature

Insertion-loss tracking is the ability of different assemblies of the same type and length to closely reproduce their inherent loss characteristics with changing temperature. As with phase tracking, the closer the assemblies track, the lower the residual systemic error and the more consistent the beamwidth, sidelobe suppression, and beam steering. System range, jamming and clutter resistance, and overall accuracy are ultimately superior. Poor insertion-loss tracking is generally attributable to poor control over materials and processes during cable manufacturing or to using cable assemblies with different characteristics.

Attenuation change (% db) vs. temperature, bulk property

Attenuation change (% db) vs. temperature, bulk property - graph

This curve is a property of bulk cable (0.190 in. diameter, solid center conductor, PFA jacket cable); very low loss cable assemblies may deviate slightly due to connector effects over temperature. Insertion-loss change is in percent dB relative to 0 dB change at 25°C. This measurement was taken at 18 GHz; it is a weak function of frequency and cable type. Insertion-loss change is primarily due to the change in conductivity of the silver with temperature. To a lesser extent, the loss tangent of the ePTFE is also a function of temperature.

The Attenuation Change vs. Temperature, Bulk Property graph is a plot of data that illustrates typical insertion-loss tracking of Gore microwave coaxial assemblies over the temperature range of -100 to +150°C. Typical insertion loss tracks within two percent for a given cable type, depending on the assembly length. You can achieve tighter insertion loss tracking by selecting cables from the same material batch. Please consult Gore for details.

Changes in phase length with temperature

Gore microwave coaxial assemblies provide improved performance and reliability in a phase-sensitive system while also reducing its complexity and cost. The low loss and VSWR characteristics of our products are well known. Perhaps not so well known is our exceptional electrical length stability under various environmental conditions. The purpose of this information is to familiarize you with that stability.

The key to this enhanced stability is our ePTFE cable dielectric. The ePTFE used is manufactured exclusively by Gore and is used in no other manufacturer's microwave assemblies. It is, in essence, the key component of our microwave cable assembly.

The information presented herein was extracted from a variety of actual qualification tests and applications. However, due to extensive variations in system requirements, it may or may not be applicable to your requirements and should not be directly specified without our consultation. Gore routinely provides performance guarantees to specific requirements. With our unique vertical integration, we have an unmatched capability to define the phase performance of our assemblies.

Temperature change (phase vs. temperature)

The extent of electrical length change over temperature range is referred to as phase-temperature response. The flatter the response curve with reference to a specified temperature (usually 25°C), the better. In addition to the flatness of the phase-temperature response curve, the extent of hysteresis is important. Hysteresis is the change in electrical length measured at a particular temperature when cycling toward a temperature extreme and upon return to that particular temperature from the extreme. It is very difficult to apply error-correction techniques to a system using transmission lines that exhibit hysteresis characteristics because their electrical lengths are different when returning to a temperature from cold versus returning to it from hot.

Assembly fingerprints (phase vs. temperature)

When considered together, the phase/temperature response and the hysteresis characteristics of a particular microwave transmission-line-assembly type constitute that assembly type's fingerprint. The fingerprint for each assembly type is unique, within limits. Some assembly types have wide limits, usually due to a lack of control over materials and manufacturing processes.

Gore microwave coaxial assemblies have very tight limits of variation on their fingerprint because Gore has total control over variables. We can and do guarantee limits if desired.

The fingerprints for full-density PTFE semi-rigid assemblies and for Gore ePTFE assemblies are very different and are used to illustrate the differences between assembly types. The Change in Electrical Length with Temperature graph below shows the difference over temperature between solid PTFE and Gore ePTFE. Electrical length is normalized to zero degrees electrical length at 25°C. All data are relative to this value.

Change in electrical length with temperature: Gore expanded PTFE cable vs. full density PTFE cable

Change in electrical length with temperature: Gore expanded PTFE cable vs. full density PTFE cable

The graph below, Phase Change and Temperature G5 Cable, Bulk Property, is the fingerprint for G5, one of our most common cable types. The data show the mean phase performance value and expected lower and upper phase performance windows. The vertical scale is shown in units of phase change in degrees/GHz/ft. These measurements do not include measurement error.

Phase change vs. temperature G5 cable, bulk property

Phase change vs. temperature G5 cable, bulk property

This curve represents the nominal phase change in degrees per GHz per foot for bulk G5 cable (0.190 in. diameter, solid center conductor, PFA jacket cable). Phase change is relative to zero ppm at 25°C. Phase change is a complex function driven by elongation of the cable with increasing temperature and a state change within the ePTFE that occurs at approximately 15°C. Phase change is repeatable with temperature cycling. The exact phase characteristics vary slightly among cable types. Contact Gore before specifying any phase-versus-temperature characteristics.

Phase Change vs. Temperature G5 Cable, Bulk Property

For example, assume that a 10 ft. Gore microwave coaxial assembly operates at 4.2 GHz over a temperature change of +25 to -55°C. From the Phase Change vs Temperature G5 Cable, Bulk Property graph, you see a .24 deg/GHz/ft. of phase length change. 

Fingerprints and your system

To compare the impact on your system of the fingerprints shown in the Phase Change vs. Temperature graph, certain requirements of the application must first be defined:

  • Operating frequency
  • Assembly length
  • Operating and nonoperating temperature ranges
  • System tolerance for phase change with changing temperature

The latter depends on system performance specifications and the acceptable complexity of error correction (and its associated cost). Values achieved by these calculations are nominal; actual assembly characteristics vary based on a variety of design characteristics and requirements. Contact W. L. Gore & Associates for specific data relating to your particular needs.

Phase tracking

Phase tracking is the ability of multiple assemblies of the same type and length to closely reproduce their inherent fingerprint with changing temperature. The more closely assemblies track to each other the better because error-corrected systems are generally designed to operate at the mean of the tracking window. The larger the window, the larger the residual systemic error. In military applications, such errors typically affect beamwidth, sidelobe suppression, and beam steering, which in turn affect system range, clutter and jamming resistance, and overall accuracy. In digital applications, such errors increase signal skew.

Assemblies most often fail to phase track adequately due to 1) poor materials and process control during cable and assembly manufacturing, and 2) using a mix of assemblies, constructed from different manufacturers' components. The former can be avoided by ensuring that the assembly manufacturer has total control over its processes and periodically proves it by testing. The latter most often occurs when assemblies are built by an assembly house (perhaps in your facility) from purchased bulk cable. The bulk cable and connectors are usually generic and are obtained from more than one supplier.

Phase stability with bending

The phase stability of coaxial assemblies with bending or flexure is very important in phase-sensitive systems because assemblies are almost always subjected to bending during installation, routine maintenance, and/or actual use. Phase changes with bending must be minimal and predictable for the impact of those changes on system performance to be assessed during system design. Predictability is particularly important for systems such as phased arrays, where many paths must ultimately be matched in electrical length within tight tolerances.

Also important are phase changes induced during system maintenance. If cables are temporarily moved to achieve access, it is unlikely that they will be returned to their exact original position. Even if they are, changes in electrical length will almost certainly occur. Minimizing these changes is very important; otherwise, system performance will suffer.

Unfortunately, phase changes with flexure cannot be avoided completely because coaxial cable is a cylindrical component. When a cylindrical component is bent, the circumference of the outside of the bend is larger than that of the inside. This modifies the geometry of the cylinder and the stresses within it due to extension and compression.

Coaxial cable consists of multiple concentric, cylindrical components. Since their diameters, materials, and actual construction differ, modification of their geometries and stresses with bending and torque differ accordingly. These differences ultimately translate into changes in electrical length. Gore microwave coaxial assemblies offer the best available phase stability with bending because of the extreme flexibility and resilience of our dielectric and our superior shielding technology.

This shielding technique was introduced by Gore in 1975 and, over the years, has demonstrated its superior flexibility, shielding effectiveness, and phase stability.

This proven design-when used in combination with Gore ePTFE dielectric-eliminates many of the bend-induced stresses that are found in other assemblies. The result is far less phase change when our assemblies are bent or flexed.

As a point of reference, the guidelines in Table 1 may be helpful (see below). Worst-case conditions are shown on the left, median conditions in the center, and best-case on the right.

It is important to establish the stability characteristics of the assembly in the actual system environment. Unfortunately, this is not always possible or practical, so a test must be used that provides an indication of relative phase stability.

One such test simulates the median conditions of occasional bending, medium bend radius, and single, reversed bend plane. To make the test more stringent, a large bend angle can be incorporated (a worst-case condition). In this test, use a mandrel with a radius 1.5 times the minimum recommended cable bend radius.

Make four measurements of electrical length: 1) with the assembly relaxed (in a straight position); 2) with the assembly wrapped 360 degrees clockwise on the mandrel to provide a large bend angle; 3) with the assembly wrapped 360 degrees counterclockwise on the mandrel to simulate a reversed bend plane; and 4) returned to the relaxed state.

The FB Test Cable graph illustrates the typical results when this test is performed on low-loss 0.190 inch diameter assemblies. Variations of this test or other tests can be performed to more closely simulate your conditions.

FB Test Cable: 2.25" Radius Coil Test

Phase Change with Flex
Loss Change with Flex

Electrical and mechanical stability in severe environment

In addition to temperature change, assemblies used in microwave systems are exposed to environmental conditions such as vibration, shock, humidity, flexure, and torque.

Gore microwave coaxial assemblies subjected to vibration and shock environments over long periods of time do not fail mechanically and electrically due to metal fatigue or cracking of solder joints, as is the case with semi-rigid assemblies. Further, they do not exhibit self-generated noise or microphonism during vibration, and they retain their electrical characteristics-electrical length, phase and loss stability, insertion loss, and VSWR.

Table 1

  Worst stability Median stability Best stability
Flexure Continuous Occasional One time
Bend radius Minimum Medium Large
Bend angle <=360 <=180 <=90
Bend plane Multiple Single, reversed Single
Torque level Severe Mild None