Tektronix, Inc.

Pavel R Zivny

Maria Agoston

Mark L Guenther

Michael Nelson

Kan Tan

Laszlo J Dobos

Edward E Averill

Ronald M Henricksen

Wenzheng Sun

A test and measurement system has one or more inputs connectable to a device under test (DUT), and one or more processors configured to execute code that causes the one or more processors to: gather a set of training waveforms by acquiring one or more waveforms from one or more DUTs or from simulated waveforms, remove noise from the set of training waveforms to produce a set of noiseless training waveforms, and use the set of noiseless training waveforms as a training set to train a neural network to predict a measurement value for a DUT, producing a trained neural network. A method of training a neural network having receiving one or more waveforms from one or more DUTs, or generating one or more waveforms from a waveform simulator, removing noise from a set of training waveforms gathered from the one or more waveforms to produce a set of noiseless training waveforms, and use the set of noiseless training waveforms as a training set to train a neural network to predict a measurement value for a DUT, producing a trained neural network.

Machine learning model training using de-noised data and model prediction with noise correction

A method of dynamically determining an oscilloscope noise characteristic includes retrieving a power spectral density (PSD) model of noise from a storage based upon a current configuration of the oscilloscope, generating a representation of any filtering being applied to a waveform generated by a device under test, using the PSD and the representation to produce a modified power spectral density, and using the modified power spectral density to determine a dynamic oscilloscope noise characteristic. A test and measurement instrument has one or more inputs to acquire waveforms from a device under test (DUT), one or more processors to retrieve a power spectral density (PSD) model of noise from a database, generate a representation of any filtering being applied to a waveform generated by the DUT, use the PSD and the representation to produce a modified PSD, and use the modified PSD to determine a dynamic instrument noise characteristic.

Semi-automated oscilloscope noise compensation based on power spectral density characterization

A test and measurement device includes an input port to receive an input signal, a sampling circuit structured to generate a sample from the input signal, in which generating the sample from the input signal produces an amount of kickout energy, and an energy reducing circuit coupled between the sampling circuit and one or more other components of the test and measurement device, the energy reducing circuit structured to decrease the amount of kickout energy from the sampling circuit. The energy reducing circuit may include or be combined with a pick-off circuit. Methods are also described.

Termination trigger pick-off for remote head sampler

Methods and systems are described for analyzing signal impairments using a test and measurement instrument. A method may include decomposing aggregate signal impairments into signal impairments that are correlated and uncorrelated to an acquired data pattern. The uncorrelated signal impairments may be further decomposed into periodic signal impairments (e.g., PJ) and non-periodic uncorrelated signal impairments. A PDF of the non-periodic uncorrelated signal impairments may be mathematically integrated, thereby producing an estimated cumulative distribution function (CDF) curve. Random signal impairments may be estimated as an unbound Gaussian distribution. The CDF curve of the non-periodic uncorrelated signal impairments and the unbound Gaussian distribution may be plotted in Q-space on a display device. Non-periodic bounded uncorrelated signal impairments (e.g., NP-BUJ) PDF may then be isolated. Bounded uncorrelated signal impairments PDF may then be synthesized. Complete uncorrelated signal impairments PDF may be synthesized. A synthesis of the decomposed components can be performed at a user-defined bit error rate to generate the total estimated jitter (e.g., TJ@BER or TN@BER).

Bounded uncorrelated signal impairment deconvolution for total signal impairment extrapolation

Method and systems are described for estimating signal impairments, in particular jitter that includes uncorrelated, non-periodic signal impairments. One system may take the form of an oscilloscope. The estimates may take the form of a probability density function (PDF) for uncorrelated signal impairments that has been modified to replace low probability regions with a known approximation and an extrapolation of the known approximation.

Methods and systems for analyzing decomposed uncorrelated signal impairments

A method for analyzing jitter using a test and measurement instrument includes obtaining a collection of time interval error (TIE) values corresponding to composite jitter of a waveform, optionally decomposing the composite jitter into jitter components that are correlated to the data pattern and components that are uncorrelated to the data pattern, and using a spectral approach to decompose the jitter components into jitter components that are recognizable as deterministic and jitter components that are unrecognizable as deterministic. Thereafter, the jitter components analyzed in the frequency domain are converted back to the time domain, and subtracted from the composite jitter, thereby isolating uncorrelated residual jitter. The uncorrelated residual jitter is decomposed into bounded uncorrelated jitter and random jitter, for example, by integrating a probability density (PDF) function of the residual jitter and analyzing the resulting cumulative distribution function (CDF) curve in Q-space.

Method for decomposing and analyzing jitter using spectral analysis and time-domain probability density

A digitizing instrument is used for modifying pattern data and jitter and noise components of a communication signal. In a typical implementation, the midpoints of a rising edge slope and horizontal portion of the communication signal are determined and multiple digital data records are acquired at the midpoints. The data sample records are transformed to frequency components and the random jitter and noise, and periodic jitter and noise components are determined. A correlated pattern data and the jitter and noise components are matrix elements in a simulated signal channel having communication system elements. Each correlated pattern data and jitter and noise component may be modified for each of the communication system element. The selectively modified correlated pattern data and jitter and noise components are combined to produce a modified communication signal that is displayed as a numeric table, eye diagram or bit error rate presentation.

System for independently modifying jitter and noise components in a signal digitizing instrument

Parameters of a spread spectrum clock signal in a communication signal are characterized by acquiring voltage samples of the communication signal at a nominal time location of an edge of the communication signal. The voltage samples are converted to time samples and the difference between the maximum and minimum time values is determined at the nominal time location. A spread spectrum clock magnitude number is generated by dividing the difference between the maximum and minimum time values by the nominal time location of the acquired voltage samples of the spread spectrum clock signal. A spread spectrum modulation profile of a spread spectrum clock signal is estimated by over sampling the time samples using an aliased index value to generate over sampled triangular waveforms representing the spread spectrum clock modulation profile. One of the over sampled triangular waveforms is use to generate the spread spectrum clock modulation profile.

Method of characterizing parameters and removing spectral components of a spread spectrum clock in a communications signal

A sequential equivalent-time sampling oscilloscope simultaneously acquires samples of a signal-under-test and a reference clock that is asynchronous with the signal-under-test. The oscilloscope converts the samples of the reference clock into timestamps and uses the timestamps to compensate for the jitter of the oscilloscope and to indicate the absolute jitter of the signal-under-test.

Sequential equivalent&#x2014;time sampling with an asynchronous reference clock

A trigger system for a digital oscilloscope operating in external clock mode. Every n pulses, the trigger system generates a trigger signal. The trigger system therefore provides a trigger every n samples of the input signal waveform.

External clock count based auto trigger for an oscilloscope

A method of detection of metastability of a trigger in a digital storage oscilloscope compares a predetermined address used to produce a trigger ready signal with a current address determined by the occurrence of a trigger event after the trigger ready signal. If the difference between the two addresses is within a predetermined limit, a metastable condition is indicated, and data acquired is treated as corrupted data.

Growing community of inventors

Portland, OR, United States of America

Pavel R Zivny