RF wafer probing with improved contact repeatability using nanometer positioning

Abstract

This paper presents an improved technique for monitoring and controlling the contact condition of on-wafer RF probes with nanometer accuracy to enhance the measurement repeatability. The setup consists of a vector network analyzer, a modified probe station with a planar calibration substrate aligned under microwave GSG probe through a closed-loop nanopositioner and a camera system. A fully one-port SOL calibration is performed in the frequency range 0.05-50 GHz. A repeatability study based on standard deviations of the measured data considering both conventional and proposed approaches is described. From these experimental results, the improvement of the technique proposed is achieved by accurately controlling the probe contacts.

On-Wafer Series-Through Broadband Measurement of Sub-fF55-nm MOS RF Voltage-Tunable Capacitors

“…Indeed, Fig. 5(b) shows the effective capacitances determined using (7) and (8) together with the approximate capacitance discussed above for V = 2.5V. C11 clearly shows a non-constant frequency behavior.…”

“…the effective capacitance C11 is mainly related to the phase-shift [equation 7]. S21 data together with relation (8) show that measured transmission coefficient │S21│vary with the voltage whereas the phase-shift Φ21 remains nearly constant. At the first order, by neglecting the denominator term of (8), the effective capacitance can be determined directly by 9' ≈ ( 9' ) 2 + ⁄ .…”

“…In contrast, C21 indicates a flat response up to 25 GHz. Above 25 GHz, the model using a single capacitance C is no more available (slope of the curve) and the VNA becomes more sensitive to stochastic measurement errors (fluctuations of the signal [8]). The measurement uncertainty imputable to the probe contact repeatability ( , ) was highlighted by performing six capacitance-voltage measurements of the same device.…”

On-Wafer Broadband Microwave Measurement of High Impedance Devices-CPW Test Structures with Integrated Metallic Nano-Resistances

“…Moreover, stochastic errors such as drift become predominant above 20GHz. Therefore, quantitative analysis should be ideally performed below 20 GHz including in the ultra-low frequency regime that offers a broad range for electrical mechanisms studies [6]. Finally, low-case measurement uncertainty can be achieved if the full measurement campaign (including calibration, de-embedding) are performed in a short period to lower stochastic errors such as drift inherent to environmental variations.…”

Nano-probing station incorporating MEMS probes for 1D device RF on-wafer characterization

“…Conventional RF test structures require probing pads whose dimensions are around 50 50 µm 2 to accommodate the probe tip geometry (example: pitch of 100 µm, contact area of 30 30 µm 2 ). Consequently, the main barriers are: (i) the extrinsic parasitic capacitance associated with the pad that is in the order of 2 fF mask the impedance of the nanodevice, rendering deembedding methods inaccurate [3], and (ii) the manual positioning of the probe onto the CPW test structure generates misalignment measurement errors that impact notably on the measured impedance of the nano-device itself.…”