A Phenomenological Investigation of Anomalous Performance in Flex Coaxial Cables

In this study, we investigate anomalous flex cable performance in a phased-array antenna on advanced EHF. A number of low-gain IF paths were found in one local-oscillator region during thermal cycle testing, subsequently causing an out-of-specification condition. Utilizing advanced diagnostic tools, the contractor determined that the root cause of the gain drop was a failure in a flex coaxial cable. We outline the development of an analytical phenomenology model employed in understanding the failure pathology and verifying the root cause. We have developed 3-D finite-element models using Ansoft High Frequency Structure Simulator (HFSS) that mimic the insertion loss behavior associated with cable failure modes. To synthesize a particular loss characteristic, we use an equivalent circuit model consisting of parallel- LC sections. A closed-form analytical expression for resonance frequency was derived, linking the circuit and physical parameters; linear regression is used to fit measured data to the HFSS cable model. Using the phenomenology model, we determined that the root cause of the failures is a delamination of the layered outer sheath of the cable. Two mechanisms were discovered: 1) spurs in the outer layer and 2) complete layer separation over short cable lengths. The first mechanism induces a small frequency-independent increase in the loss and is inversely proportional to the spur delamination angle. This loss is attributable to conventional aging and considered benign. With the second mechanism, the delamination gap induces a sharp resonance in the loss at a discrete frequency and is reminiscent of a low-order filter. This phenomenon is a malignant loss responsible for our anomalous out-of-specification condition. We completed a parametric study using the phenomenology model, and determined that: 1) resonance frequency is inversely proportional to the gap width (Wg) ; 2) resonance frequency is proportional to the square root of the gap size (?{- - dg}) ; 3) insertion loss amplitude is proportional to the air gap size and width; and 4) the Q of a given delamination region is proportional to the square root of the gap size.