Search

The measurement of the electromagnetic reflection characteristics of a device under test (DUT) is of interest in a large variety of applications. For this task, the established low-cost approach of using pulse-shaped excitation signals is more and more replaced by employing continuous binary coded excitation signals. The reason behind this is that reflection measurement systems with continuous signals offer a larger signal-to-noise ratio (SNR), which allows for a larger dynamic range compared to pulsed systems for the same measurement time. However, a correlator is needed to access the impulse response function (IRF) of the DUT. In this contribution, a reflection measurement system is presented using a hybrid correlation approach. In the latter, the correlation is performed using a combination of a hardware and a software correlator. The hardware of the developed measurement system has been optimized to achieve a high dynamic range and a high measurement accuracy. A demonstrator system has been realized using the so-called almost perfect auto-correlation sequences (APASs) with a chip clock of 5Gchips/s. With the optimized hardware, a dynamic range of 82 dB has been achieved. The measurement accuracy has been validated by means of distance measurements versus a high accuracy laser reference system achieving a maximum error of 25 ${\mu }\text{m}$ . Furthermore, the measured reflection coefficient of a low-pass filter has been compared with the results obtained with a vector network analyzer (VNA).

In the area of electromagnetic metrology, binary coded excitation signals become more and more important and various binary coded sequences are available. The measurement approach is to assess the impulse response function of a device under test by correlating the response signal with the excitation signal. In order to achieve a high measurement reproducibility as well as a high dynamic range, the generated binary coded signals have to provide low-noise. In this contribution, a low-noise signal generator realized with a field programmable gate array is presented. The performance investigation of different kinds of binary coded excitation signals and different correlation concepts have been practically investigated. With a chip rate of 5 Gchip/s, the generator can be utilized for ultra-wideband applications. In order to allow for a low-noise and long-term stable signal generation, a new clock generator concept is presented and results of phase noise measurements are shown. Furthermore, an algorithm to fast and precisely shifting the time lag between two binary coded signals for correlating excitation and response signals with a hardware correlator is presented. Finally, the realized demonstrator system is tested using two commonly used types of binary coded sequences.

Pseudo-random binary coded excitation signals are being increasingly used for electromagnetic network analysis. Various types of sequences with different correlation properties are available. Unfortunately, nonidealities of real measurement systems, like noise and nonlinearities, degrade the correlation functions. In this contribution a simulator is presented that takes several nonidealities into account. The impact of these nonidealities on two commonly used sequence types, the maximum length sequence (MLS) and the almost perfect autocorrelation sequence (APAS), on the achievable dynamic range (DR) are investigated. In the framework of the presented study, it was taken care that its results can be transferred into hardware requirements.

For the realization of an m-sequence-based reflec-tometer, a correlator performing a pulse compression is necessary in order to determine the impulse response function (IRF) of a device under test (DUT). Using a hardware correlator, a time delayed version of the m-sequence transmit signal is required. In this contribution, a concept for the generation of two m-sequences with a fast and precisely adjustable time lag between them is presented. Programmable linear feedback shift registers (LFSRs) are used for generating the m-sequences. The desired time lag is realized by using different initial states of the LFSRs. To guarantee that the time lag is set within one clock period, a select pulse generator (SPG) has been designed and implemented into the system. The concept has been validated by means of measurements using a realized demonstrator setup. According results show the fast time lag adjustment. Furthermore, reflection measurements have been performed with different coupler architectures and a coaxial cable with a length of 20m. Results obtained with two different couplers are compared regarding the achievable dynamic range.

In this contribution, an m-sequence-based system for network analysis is presented. This system is based on a new correlator concept, which combines the advantages of a hardware and a software correlator. In hardware, the multiplication of the received signal with a time lagged version of the excitation signal is performed. This enables a correlation for desired time lags resulting in a short measurement time, especially for long m-sequences. In software, integrate and dump (I&D) is performed, which enables a precise integration over one sequence length or integer multiples thereof, resulting in a high side maxima suppression. Furthermore, compared to a pure software correlator, the requirements with regard to the sample and hold stage are significantly lower. First, the fundamental properties of m-sequences and of the realized correlator are discussed. Afterwords, results with optimized measurement time and dynamic range of the system, derived from numerical simulations, are presented. Furthermore, to experimentally validate the system concept, a demonstrator has been realized. Measurement results showing the achievable dynamic range are presented.