A Technique for Optimal On-Wafer Device Spacing at Millimeter-Wave Frequencies

This dataset contains the simulated data of a ground-signal-ground on-wafer probe landed on a microstrip variable-impedance device (DUT) with a 775 micron (um) microstrip line neighboring the device and without any line neighboring the device. Two variables were investigated in the data: the impedance of the DUT and the X,Y location of the neighboring line. The reflection coefficient of the probe was recorded from 1 Gigahertz (GHz) to 150 GHz. We use a metric discussed in the complementary paper that we deem 'maximum error' which is the maximum value, across the frequency band, of the absolute difference between the probe reflection coefficient, at a specific DUT impedance and neighboring line location, and the probe reflection coefficient with no line nearby. Figure 2,3, and 4 are all different conditions of DUT impedances and neighboring line locations. In the paper, red curves and red X markers correspond to when the maximum error metric has exceeded 0.03. Green curves and green checkmark markers correspond to when the maximum error metric is below 0.015 and the yellow curves and yellow diamond markers correspond to when the maximum error metric is between 0.015 and 0.03. This dataset also contains the measured and simulated data for the probe reflection coefficient when landed on the output of a high-electron-mobility transistor (HEMT) with and without a nearby 775um line. The bias point of the HEMT device was Vds: 10V and Ids: 10mA. The HEMT measurement with no line nearby was used as the impedance of the DUT for the probe simulation. The probe simulation was calibrated using an Open-Short-Load (OSL) calibration technique so that the measurement and simulation reference planes were the same.

[
  "006:55"
]

[
  "en"
]

[
  "006:045"
]

[
  "Electronics:Electromagnetics",
  "Electronics:Semiconductors"
]