Allen F. Horn

Conductor profile effects on the propagation constant of microstrip transmission lines

We experimentally show that the increase in conductor loss due to roughness is larger than the factor of two predicted by the most widely used roughness factor models. This is consistent with a recent numerical study of the effect of random roughness on conductor loss. The data also show that, for thin substrates, increasing the conductor profile substantially slows the effective velocity of propagation and also increases dispersion, independent of the composition of the dielectric material. Measurements are compared with results from a new conductor model as used in a 3-D planar EM analysis that includes an excess inductance related to the conductor profile. It is shown that this accounts quantitatively for both the insertion loss and phase constant effects.

Conductor surface-roughness effect in the loss tangent measurement of low-loss organic substrates from 30 GHz to 70 GHz

This work shows the effect of the conductor surface roughness (CSR) in the measurement of the loss tangent using the microstrip ring resonator method (RRM) in three low-loss, organic dielectric substrates (RT/duroid 6002, 6202 and 5880) and each, with two types of copper metallization. In addition, relative permittivity and loss tangent measurements are presented from 30 GHz to 70 GHz for the three substrates for the first time using the RRM. The copper metallization types used for the study are the rolled metallization with a root mean square (RMS) CSR value of 0.27 um and the Electro-deposited (ED), with an RMS CSR of 1.89 um. Results show that the extracted loss tangent values from the ED metallization are in excellent agreement with the values extracted with a Split Cylinder Cavity at about 34 GHz for all materials studied. On the other hand, the loss tangent values extracted with ring resonators fabricated on a rolled metallization, show errors of about 50%, when compared to those extracted with the same Split-Cylinder cavity. No significant errors were detected in the extracted relative permittivity with the two CSR values, and errors below 6% were achieved when compared to the values extracted with the Split Cylinder method.

The influence of test method, conductor profile and substrate anisotropy on the permittivity values required for accurate modeling of high frequency planar circuits

Purpose – The purpose of this paper is to help high frequency circuit designers understand how to choose the best permittivity value for a laminate material for accurate modeling.Design/methodology/approach – In this paper, experimental measurements of the performance of simple circuits are compared to various mathematical and software models.Findings – Higher permittivity values were obtained using samples with bonded copper foil compared to samples etched free of foil. These higher values yielded better agreement between measured and modelled performance using current automated design software. High profile foil on thin laminates was found to increase the surface impedance of the conductor and change the propagation constant and apparent permittivity of the laminate by 15 percent or more. It was also demonstrated that, under some circumstances, the anisotropy of the substrate could result in differences in measured and modelled performance.Research limitations/implications – Only a limited number of cir...

Thermal stability of the dielectric properties of the low-loss, organic material RT/duroid 6002 from 30 GHz to 70 GHz

For the first time, the thermal stability of the dielectric properties, i.e. the relative permittivity and the loss tangent, are presented for RT/duroid® 6002 from 30 GHz to 70 GHz over the temperature range between 20 °C and 200 °C, using the microstrip ring resonator method at two different microstrip impedances. High-frequency-resolution, Multiline TRL calibrations were performed at each temperature point to increase the accuracy of the measurements. Measurements show a remarkably-stable normalized temperature coefficient of the relative permittivity of −17.6 ppm/°C across the entire bandwidth. Likewise, the normalized loss tangent temperature coefficient had a value of about 0.00118 °C−1, with little variations throughout the measurement bandwidth.

Measurement of simulated active device and RF trace heating in high frequency circuit laminates

Purpose – The purpose of this paper is to quantify the effects of thermal conductivity (TC), dielectric constant and dissipation factor (DF) of circuit laminates on the temperature rise with active components and RF trace heating.Design/methodology/approach – Temperature rise measurements were made on surface mounted chip resistors (to simulate active components) at various dissipated power levels, with and without “via farms”. The RF heating temperature rise of 50 ohm microstrip transmission lines on 0.5 mm laminates was also measured by the same method.Findings – The chip resistor temperature rise correlated with the independently measured TC of the laminate materials. The use of a “via farm” substantially reduced the temperature rise in all materials, but the higher TC laminates still conferred a measurable advantage. The trace temperature rise due to RF heating correlated with both TC and DF.Research limitations/implications – It was shown that the one‐dimensional heat transfer model does not accurate...