Data processing: measuring – calibrating – or testing – Calibration or correction system – Timing
Reexamination Certificate
2001-03-05
2003-02-18
Hilten, John S. (Department: 2853)
Data processing: measuring, calibrating, or testing
Calibration or correction system
Timing
C356S005080
Reexamination Certificate
active
06522983
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates generally to time calibration of a measurement instruments and more particularly to a timebase calibration method for a sampling instrument, such as a equivalent time sampling oscilloscope.
An equivalent time sampling oscilloscope acquires a waveform record of a repetitive input signal in response to strobe pulses generated by the oscilloscope timebase. The timebase includes trigger circuitry, a voltage controlled oscillator and a strobe generator. The oscilloscope also includes front panel controls for setting acquisition and display parameters. The acquisition parameters include the time per division setting, acquisition window, sample spacing and the like. The acquisition parameters are interpreted by a controller, operating under program control, to produce the timebase parameters for generating the strobe pulses. The acquired waveform record is digitized and stored in oscilloscope memory for additional processing and display.
The voltage controlled gated oscillator is phase locked to a fixed reference clock provided in the oscilloscope. The oscillator initiates a clock signal upon receiving a trigger signal from the trigger circuitry. The strobe generator includes clock counter circuitry that provides a coarse delay and an interpolator for providing a fine delay. The controller loads a counter in the clock counter circuitry with a coarse delay value. The clock signal from the gated clock generator increments the counter. The clock counter circuitry generates an output pulse when the counter reaches the loaded terminal count. The output pulse is applied to a ramp generator that initiates a ramp signal that is applied to the intepolator. The controller provides digital-to-analog converter (DAC) codes (fine time delay values) that are converted to an analog values that act as threshold values for the ramp signal. The interpolator generates an output strobe pulse when the ramp generator signal crosses the fine time delay threshold value.
In order to achieve accurate timing and signal sample spacing for the acquired waveform, the frequency and hence the period of the clock signal needs to be known. The fixed clock reference in conjunction with direct counting type frequency measurement support circuitry in the oscilloscope determines the frequency of the gated oscillator. The fixed clock reference is also used to determine the dynamic range of the interpolator and generate a horizontal look-up table of nonlinearity corrected digital-to-analog converter values for the interpolator.
What is needed is a timebase calibration method for a sampling system that does not require direct counting type frequency measurement support circuitry. In addition, the timebase calibration method needs to be capable of determining the clock frequency of multiple clock generators. Further, there is a need for a timebase calibration method that determines the clock frequency of a gated oscillator that is not locked to a fixed reference clock.
SUMMARY OF THE INVENTION
Accordingly, the present invention is a timebase calibration method for an equivalent time sampling digitizing instrument having timebase circuitry that includes a strobe generator having coarse time delay circuitry and fine time delay circuitry. The coarse time delay circuitry receives a first portion of a strobe delay input for loading a counter with a coarse time delay value and the fine time delay circuitry receives a second portion of the strobe delay input for generating an analog input signal to an interpolator in the fine time delay circuitry. The analog signal is derived from a fine time delay look-up table of digital values. The coarse time delay circuitry is responsive to a received clock signal from a clock generator and the interpolator is responsive to the analog signal for generating a variable time delay strobe output pulse. The calibration method estimates the frequency of the clock signal from the clock generator using the power spectrum of the difference frequencies between the clock generator signal and a reference oscillator signal. A dynamic range for the interpolator is defined in digital-to-analog converter code values as a function of the clock signal period based on the variance of the difference of a plurality of acquired waveform record pairs of the calibration oscillator signal. Each record pair is acquired at a selected digital-to-analog converter code value with one record being acquired at a first coarse time delay and the other record being acquired with a second coarse time delay. A linear horizontal look-up table of digital-to-analog converter code value points is generated based on the defined dynamic range of the interpolator wherein the code value points are nominally separated by the same number of digital-to-analog converter code values. Residual nonlinearities of the interpolator are characterized over the defined dynamic range of the interpolator based on a plurality of waveform records acquired at an estimated zero crossing point of the calibration oscillator signal. The characterized residual anomalies are scaled to digital-to-analog converter code values, and combined with the digital-to-analog converter code values of the linear horizontal look-up table.
The estimation of the clock signal frequency further includes the steps of generating a coarse estimate of the clock signal frequency and using the coarse estimation as a starting point for generating a fine estimation of the clock signal frequency. Both the coarse and fine estimations of the clock signal frequency acquire waveform records at selected frequency settings of the calibration oscillator. An FFT is applied to the waveform records to obtain the frequency spectra of the difference frequencies of the clock signal and the calibration oscillator signal. The frequency where the non-DC component with the maximum power occurred is determined and indexed to frequency bins for the coarse estimation. The frequency data is squared and a second order polynomial curve fit is applied to squared frequency data. The minimum of the polynomial curve fit is selected as the coarse estimate of the clock signal frequency. For the fine estimation, the sideband power of the frequency where the non-DC component with the maximum power occurred is determined and indexed to frequency bins. The first frequency bin having a complete record of sideband powers is selected and a second order polynomial curve fit is applied to the sideband powers associated with the frequency bin. The frequency of the clock signal is calculated by combining the calibration oscillator frequency defined by the minimum of the polynomial curve fit with the difference frequency defined by the selected frequency bin.
The dynamic range of the intepolator is determined using coarse and fine characterization stages with each characterization stage having an iteration terminal count, a fixed end digital-to-analog converter code value, a start digital-to-analog converter code value, and a digital-to analog converter step value for varying the start digital-to-analog converter code value. For each stage an error array is initialized. The start digital-to-analog converter code values for the coarse characterization are the respective minimum and maximum digital-to-analog converter code values. First and second waveform records are acquired of the calibration oscillator signal using the strobe generator output pulses derived from the clock signal with the start digital-to-analog converter code values and the coarse time delay for the second waveform record being incremented by one from the coarse time delay for the first waveform record. A difference record is generated by performing a point-wise subtraction between the two waveform records. A variance value of the difference record is determined and appended to the error array. For each iteration, the start digital-to-analog converter code value is incremented by the digital-to analog converter step value to generate a new start digital-to-analog converter code value. A new pair of wav
Dobos Laszlo
Lester Kenneth J.
Bucher William K.
Hilten John S.
Sun Xiuqin
Tektronix Inc.
LandOfFree
Timebase calibration method for an equivalent time sampling... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Timebase calibration method for an equivalent time sampling..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Timebase calibration method for an equivalent time sampling... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3146174