Measuring the Invasiveness of High-Impedance Probes

Transcription

1 Measuring the Invasiveness of High-Impedance Probes Uwe Arz 1 Pavel Kabos 2 Dylan F. Williams 2 1 Physikalisch-Technische Bundesanstalt, Braunschweig, Germany 2 National Institute of Standards and Technology, Boulder, CO, USA

2 Introduction Outline Introduction High-impedance probe (HIP) characterization Application to time-domain measurements HIP invasiveness measurement Validation of invasiveness model Outlook: 3-port measurements Summary and conclusions 2

3 Introduction 3

4 Introduction Why use high-impedance probes? minimize loading effects on circuit under test reduced invasiveness superior to hand-held oscilloscope probes suitable for fine-pitch applications high bandwidth Problem: Probe properties are usually not accounted for! 4

5 HIP characterization Standard procedure: FAILS! Method A: WORKS! Reference Planes of Coaxial 1 st -tier Cal. Coaxial 1 st -tier Cal. Coaxial 1 st -tier Cal. Probe 1 Probe 2 Reference Planes of On-Wafer 2 nd -tier Cal. ( probe-tip ref. plane) Initial Ref. Plane of TRL Calibration deembed GSG-probe and CPW line from preliminary measurement CPW line standard (GSG=Probe 1 and HIP=Probe 2) 5

6 Application to time-domain measurements Measurement setup Pulse source On-wafer waveform measurement Voltage at tip of high-impedance probe 2 mm CPW line 0.2 Voltage from oscilloscope measurement (HIP characterization averaged from 5 different line lengths) reference voltage (VTip calculated from source) Voltage from oscilloscope measurement (HIP characterization derived from 2mm line) Voltage from Oscilloscope Measurement (ideal HIP-probe 20x attenuation assumed) Voltage (V) calibration of time-domain measurements is possible with new probe characterization procedure! Time (ns) (MTT Trans. Feb. 03, pp ) 6

7 HIP invasiveness Experimental setup 50 Ω broadband coaxial load Port 1 VNA GSG HIP GSG Port 2 VNA (a) (a) reference plane of on-wafer cal. CPW line standards of different lengths CPW transmission properties disturbed! 7

8 HIP invasiveness Influence on measurement of two CPW line standards S 11 in db -40 S 21 in db THRU measurement without HIP THRU measurement with HIP L19 measurement without HIP L19 measurement with HIP Frequency in GHz -7.5 THRU measurement without HIP THRU measurement with HIP L19 measurement without HIP L19 measurement with HIP Frequency in GHz 8

9 HIP invasiveness Measurement procedure perform Multiline TRL calibration (ref. planes at position (a) give results shown on previous slide) measure one specific CPW line standard with reference planes at position (b) isolate invasiveness error box 50 Ω termination GSG HIP GSG (a) (b) (b) (a) CPW line standard invasiness error box 9

10 HIP invasiveness Invasiveness error box properties: S-to-Z transformation 2000 I 1 Z -Z I 2 Z 11 -Z 12 Z 22 -Z 12 Re[Z ij ], Im[Z ij ] in Ω Re[Z ij ] Im[Z ij ] Re[Z 11 ] Re[Z 21 ] Re[Z 22 ] Im[Z 11 ] Im[Z 21 ] Im[Z 22 ] Frequency in GHz V 1 V 2 Z 12 General equivalentcircuit model: only shunt element Z 12 required, equal to load impedance! 10

11 HIP invasiveness Calculation of load impedance from HIP characterization use 2-port HIP characterization (S-parameters S ij ) together with reflexion measurement of broadband coaxial load Γ term Γ load 2-port HIP characterization (S-parameters S ij ) Γ term conversion to load impedance: 11

12 HIP invasiveness Comparison of measured vs. calculated load impedance 2000 Z load calculation Z load = Z 21 (measured impedance parameter of invasiveness error box) 0 Z load from RC model (C=0.04pF, R=1kΩ) 1500 Z load calculation Z load = Z 21 (measured impedance parameter of invasiveness error box) Z load in Ω 1000 Z load from RC model (C=0.04 pf, R=1 kω) Note: manufacturer spec is 0.02 pf! arg(z load ) in degrees Frequency in GHz Frequency in GHz magnitude of Z load phase of Z load 12

13 Error box validation Model of CPW measurement disturbed by HIP: cascade of transmission matrices (S-to-T transformation) 50 Ω termination GSG HIP GSG CPW line standard (a) (b) (b) (a) (a) (b) (b) (a) T ab T inv T ba 13

14 Error box validation Results with T inv from THRU measurement -20 S 11 in db THRU measurement with HIP disturbance T H RU calculation from H IP 2-port-characterization [1] THRU calculation from HIP invasiveness measurement L19 measurement with HIP disturbance L 19 calculation from H IP 2-port-characterization [1] L19 calculation from HIP invasiveness measurement Frequency in G H z T inv derived from THRU allows reconstruction of L19 measurement! 14

15 Outlook: 3-port measurements Invasiveness in a more complex environment: compare measurement w/o HIP against measurements w/ 2 different HIP positions P 3 (1) 0 P 1 DUT S 11 (db) -5 P 2 P 3 (2) -10 S 11 DUT with HIP in position (1) S 11 DUT with HIP in position (2) S11 Two-port DUT Frequency (GHz) 15

16 Summary and conclusions 16