This is an author-created, un-copyedited version of the article G. N. Phung and U. Arz, "Parasitic Probe Effects in Measurements of Coplanar Waveguides with Narrow Ground Width,"2020 IEEE 24th Workshop on Signal and Power Integrity (SPI), Cologne, Germany, 2020, pp. 1-4. Copyright © 2020 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE. The definitive publisher-authenticated version is available online at: https://doi.org/10.1109/SPI48784.2020.9218166 Parasitic Probe Effects in Measurements of Coplanar Waveguides with Narrow Ground Width Gia Ngoc Phung Physikalisch-Technische Bundesanstalt (PTB) Bundesallee 100, 38116 Braunschweig gia.phung@ptb.de Uwe Arz Physikalisch-Technische Bundesanstalt (PTB) Bundesallee 100, 38116 Braunschweig uwe.arz@ptb.de Abstract—On-wafer measurements contain a large variety of parasitic effects degrading the accuracy of multiline Thru-Reflect Line (mTRL) calibration. These effects are caused by internal and external disturbances such as probe effects, multimode propagation, crosstalk between adjacent structures and radiation effects. While a lot of investigations have been performed for the most common coplanar waveguides (CPW) with nominal ground width, CPW with too narrow ground width have not been investigated thoroughly. This paper demonstrates how the probe effects deteriorate the mTRL-calibrated S-parameters for CPW structures with narrow ground width. Keywords— calibration, coplanar waveguides, multiline Thru- Reflect Line (mTRL), probes. I. INTRODUCTION On-wafer measurements are of fundamental importance for the characterization of components and devices in high-speed and microwave applications ranging from wireless communications, automotive radar and medical sensing. However, for on-wafer measurements, a lot of challenges need to be addressed. On-wafer probing with ground-signal- ground (GSG) probes itself contains a lot of parasitic effects which are not to be underestimated. On the one hand, these effects can be initiated by the impact of neighborhood, measurement environment, multimode propagation and the measurement instrumentation itself. On the other hand, they are caused by the actual characteristics of the device under test (DUT), mainly radiation and dispersion effects. In recent years, a major effort has been undertaken to investigate and to clarify the sources of these parasitic effects. In [1] the influence of microwave probes on calibrated on-wafer measurements is demonstrated for the coplanar waveguides (CPW) and thinfilm microstriplines (TFMSL) up to W-Band. Similar investigations have been performed in [2,3] for the extended frequency range up to 330 GHz. In [4,5] the occurrence of parasitic substrate modes was discussed. The latter investigation suggested measures to mitigate the propagation of substrate modes. In [6-8] the impact of radiation losses has been thoroughly explained. It has been detected that the influence of the CPW ground width is one of the main causes for radiation losses and dispersion effects [7]. The impact of wide CPW ground width has been clarified thoroughly in [8]. A recommendation has been proposed to keep the total CPW wtot which represents two times the ground-to-ground spacing plus double the ground width smaller than the formula given in [7,8]. A priori the ultimate maximum applicable CPW ground width which allows decent CPW characteristics is therefore defined. So, one would assume that reducing the CPW ground width would be the best choice to avoid any radiation and dispersion effects. However, this might not be true and gives rise to the question how the CPW characteristics would change when the CPW ground width is reduced to a minimum. Starting from a measurement example, this paper demonstrates how the S-parameters of the coplanar structures with reduced CPW ground width change in interaction with probe effects. For a better understanding, 3 D full-wave electromagnetic simulations in CST [9] are performed*. (a) (b) Fig. 1. a.) Electromagnetic model of a complete wafer with probe excitation in CST [9]; b.) Investigated wafer with three different calibration sets. Fig. 1 shows the investigated wafer which has been used in [10,11] and designed for the investigation of parasitic effects. The wafer consists of three different calibration sets with common CPW parameters of CPW signal width of w = 50 µm, a gap of s = 25 µm and varied CPW ground widths of wg = 50, 270 and 650 µm. The calibration set consists of a short as reflect standard, a 400 μm long CPW line as thru and eight additional lines with lengths between 500 and 20400 µm. All *) We use brand names only to better specify the experimental conditions. PTB does not endorse commercial products. Other products may work as well or better.