A transfer matrix modification for accurate terahertz FMCW thickness measurements

Schreiner, Nina S.; Baccouche, B.; Molter, Daniel; Sauer-Greff, W.; Urbansky, R.; Friederich, Fabian

doi:10.1109/ucmmt.2017.8068507

Cited by 1 publication

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To obtain the resulting measurement signal, the product is downsampled according to our sampling rate. The block diagram in Figure 5b illustrates our modified TMM approach which is derived in [31] in detail. In contrast to [29,30], an even more efficient calculation of the signal model is achieved by only calculating the matrix of the samples of the measurement signal.…”

Section: Model-based Signal Processing Approachmentioning

confidence: 99%

“…Using Equations (6) and (7), the combined matrix of the l-th layer results in Ml+1[n] = Dl,l+1[n]∗Pl+1[n]. To calculate the complete transfer matrix bold-italicM[n], the matrices of the single layers are multiplied, and hence the equivalent time-domain function hequi[n] = M10[n]/M00[n] is multiplied with the simulated measurement signal [31].…”

Section: Model-based Signal Processing Approachmentioning

confidence: 99%

See 1 more Smart Citation

Multilayer Thickness Measurements below the Rayleigh Limit Using FMCW Millimeter and Terahertz Waves

Schreiner

Sauer-Greff

Urbansky

et al. 2019

Sensors

Self Cite

View full text Add to dashboard Cite

We present thickness measurements with millimeter and terahertz waves using frequency-modulated continuous-wave (FMCW) sensors. In contrast to terahertz time-domain spectroscopy (TDS), our FMCW systems provide a higher penetration depth and measurement rates of several kilohertz at frequency modulation bandwidths of up to 175 GHz. In order to resolve thicknesses below the Rayleigh resolution limit given by the modulation bandwidth, we employed a model-based signal processing technique. Within this contribution, we analyzed the influence of multiple reflections adapting a modified transfer matrix method. Based on a brute force optimization, we processed the models and compared them with the measured signal in parallel on a graphics processing unit, which allows fast calculations in less than 1 s. TDS measurements were used for the validation of our results on industrial samples. Finally, we present results obtained with reduced frequency modulation bandwidths, opening the window to future miniaturization based on monolithic microwave integrated circuit (MMIC) radar units.

show abstract

Section: Model-based Signal Processing Approachmentioning

confidence: 99%