ITZWIRE wireline (hard-wire) vibrator test analysis.
====================================================
ITZwire 2.2018.01.10 January 2018
Adjust the way the noise levels are removed from the distortion values. Also adjust the windowing method used on the data to give smoother results. The main benefit is for low frequency broadband data, where the low frequencies are at very low drive levels so the signal to noise is poor.
ITZwire 2.2017.03.28 March 2017
Adjust the relative phase plots for mass and maseplate accelerometer versus reference to give better definition.
ITZwire 2.2016.06.09 June 2016
Adjust the frequency versus time plot to eliminate step effects.
ITZwire 2.2016.01.01 January 2016
Adjust the windowing for the distortion, phase and frequency calculations, to give better definition.
ITZwire 2.2014.04.13 April 2014
General clean to avoid divide by zero errors.
ITZwire 2.2013.10.13 October 2013
Add correlation coefficient to the start time and relative timing analyses to show the relationship between the reference and the data (force) trace samples. These samples are corrected in time using the raw cross-correlation results for time offet before the correlation coefficient is calculated.
ITZwire 2.2013.09.27 September 2013
Change the Relative Timing analysis so it does not use an analogue analysis designed for a sweep. This is to make it more robust for non-swept signals. Analogue analysis requires a series of sinusoidal data.
Adjust the distortion analysis to cope with very low frequency sweeps. Very low frequency sweeps will only have a few samples difference between the drive level and the harmonics, so drive energy can be included in the distortion analysis. Adjust so the background nose plus drive energy is discounted from the distortion calculation.
Adjust the distortion values for all distortion including noise to cope with very low frequency sweeps.
Adjust the phase response analysis to give greater definition.
ITZwire 2.2013.03.18 March 2013
Minor adjustment to the samples extracted as the sweep, by default. The threshold is reduced to 3% (from 4%) giving a few more samples in the extracted sweep time.
ITZwire 2.2013.01.14 January 2013
Minor adjustment to the Cross-correlation routine. The ouput array to decibel array calculation is now completed in Corcalc.
ITZwire 2.2012.04.25 April 2012
Adjust MODAL method to eliminate more outliers.
ITZwire 2.2012.03.16 March 2012
Adjust MODAL method. Replace most MEDIAN calls with new MODAL call for improved accuracy.
ITZwire 1.20120306 March 2012
Adjust the method to derive analogue measurements for start time errors again to enable pulsed sequences to be cross-checked.
ITZwire 1.20120213 February 2012
Adjust the method to derive analogue measurements for start time errors, to improve accuracy for extreme offsets between data and reference.
ITZwire 1.20110929 September 2011
Adjust the method to derive frequency from FFT calculations, to improve accuracy.
ITZwire 1.20110901 September 2011
Minor bug fix to ensure all files have the same sample interval
ITZwire 1.20110610 June 2011
Minor bug fix for long traces.
Reset the dB/octave response estimate for correlated data.
ITZwire 1.20110308 March 2011
Add capability for CSV data file input rather than SEGY.
Rewrite the analysis methods for stacking from two input traces, and for correlating with geophone (velocity) data.
ITZwire 1.20110105 January 2011
All ITZwire releases now numbered according to the date of compilation, i.e. 2011 Jan (01) 5th (05)
Adjust the phase analysis unwrap to cope with limited samples.
ITZwire 1.132 - November 2010
Adjust the relative phase calculations to make more robust.
Adjust the clock drift analysis, getting value from the least-squares calculated values not the real values so avoiding noise.
ITZwire 1.131 - October 2010
Adjust the file names shown during analysis to the DOS window.
Adjust the phase plot to show simple relative timing.
ITZwire 1.130 - September 2010
Add RELTIM analysis for simple analysis of relative timing between two traces by correlation. One page with minimal plots produced, no relative timing by analogue analysis just from the correlation and relative phase.
ITZwire 1.129.1 - May 2010
Adjust the windows for the results, moving the raw cross-correlation plot to the second page, and the fundamental force to the first page.
ITZwire 1.129 - March 2010
Summary values within plots changed to integer to ease definition.
Start time analysis improved by estimation before non-linear least squares fit.
ITZwire 1.128 - February 2010
Total traces per SEGY file set to 1,048,576.
Total samples per trace set to 131072.
ITZwire version 1.127 - 02/10/2009
Maximum Samples increased to 500000.
Add non-linear fit to the analogue assessment for relative clock timing errors. Use the non-linear fit to estimate the PPM error and start-time error.
ITZwire version 1.126 - 14/08/2009
Add total distortion to the total, odd and even harmonic distortion plot.
Add linear fit to the analogue assessment for relative clock timing errors.
Correct error in the graphing for phase and analogue timing which included the least-squares linear fit.
ITZwire version 1.125 - 1/07/2009
Adjust the start time graphing to add an analogue assessment for relative clock timing errors.
ITZwire version 1.124 - 15/02/2009
Adjust the graphing to fix array allocation bugs.
Add accelerometer phase plots to the independent acelerometer analyses.
ITZwire version 1.123 - 05/02/2009
Add capability for time offsets for the reference, force or accelerometer traces in the vibrator analyses.
OFFREF offset in positive time for reference trace (milliseconds)
OFFFOR offset in positive time for ground force trace (milliseconds)
OFFACC offset in positive time for accelerometer traces (milliseconds)
ITZwire version 1.122 - 15/11/2008
Adjust the phase analysis to avoid noise at the end of the sweep
Add start-time error values to the top of the start-time analysis.
Adjust the force plots to cope with low signal values, maximum force less than 10 units.
ITZwire version 1.121 - 23/08/2008
Adjust the sample selection for the window.
Adjust the graph windows for independent accelerometer analyses.
ITZwire version 1.120 - 02/08/2008
Adjust plots for better definition
ITZwire version 1.119 - 09/05/2008
Add multiplier MFORCE for the ground force trace to allow inversion of the ground force for negative polarity accelerometers.
ITZwire version 1.118 - 15/02/2008
Add minimum plot capability and start-time analysis to independent accelerometer analyses.
ITZwire version 1.117 - 19/01/2008
Correct error in raw correlation wavelet display.
Minor bug-fixes incorporating bug fixes for sub-routines from ITZ, and the graphing subroutine for the raw correlation plot.
ITZwire version 1.116 - 1/08/2007
Minor bug-fixes. Correct bug in MEDIANX subroutine.
FFID numbers with current systems may exceed 999999, fix problems handling larger FFID numbers.
ITZwire version 1.115 - 16/07/2007
Adjust number of sample handling to cope with values over 65536
Adjust the number of windows used for phase, frequency and distortion plots.
Fine timing response is calculated from the slope of the phase
ITZwire version 1.114 - 14/05/2007
Add capability for independent accelerometer test data analysis.
Independent accelerometer data can be simply recorded by the recording system. This avoids the need for expensive test equipment, apart from the independent accelerometers and charge-coupling and amplifier interface equipment.
Additional results given for the accelerometer data.
ITZwire version 1.113 - 04/04/2007
Adjust the frequency versus time plot for better accuracy.
Change phase calculation by windowed phase calculations, minor loss in accuracy compensated for by a gain in reliability for special sweeps.
Change frequency end points to -24dB points.
Add estimate of basic sweep parameters to plots.
ITZwire version 1.112 - 27/03/2007
Add frequency range at -6, -12, -24 and -48 dB.
Add the basic QC values to each plot, start time, signal to noise, average and maximum values, etc.
Bug fix for polarity of some problem files. Polarity is now fixed by the raw correlation.
Add estimate of dB/octave for log sweeps.
ITZwire version 1.111 - 30/10/2006
Adjust the phase calculation removing artificial offsets from the FFT process.
Adjustment to the start-time-error calculation based on least squares linear fit of the phsae response.
ITZwire version 1.110 - 10/8/2006
Adjust the way the samples are automatically limited to those with valid sweep. Better avoiding of noise spikes.
Adjust the frequency versus time for greater accuracy.
Adjust the relative phase plot so the start-point is adjusted to remove the median offset.
Add an average force reading to the fundamental force output graph.
ITZwire version 1.109 - 25/5/2006
Adjust statistics to better show timing error calculation values.
ITZwire version 1.108 - 12/3/2006
No visible difference to user results, same processing algorithms. Changes to modular structure of the program to ease changes.
Add PLWIRE.F for results plots. Smaller executable size.
ITZwire version 1.107 - 9/3/2006
Change full plot force and reference trace data plots for colour showing overlap of response.
ITZwire version 1.106 - 19/11/2005
Eliminate checks for sample amplitude.
Adjust phase response form maximum definition. Change unwrap algorithm.
Adjust cross-correlation envelope smoothing response (moving average of 3 samples).
ITZwire version 1.105 - 17/10/2005
Version 1.105 of ITZwire calculates relative phase from force phase minus reference trace phase, then eliminates spikes over Pi/2.
Checks for the polarity from the raw cross-correlation and forces the phase plot to suit polarity.
Instances of data with spiking/noise in the data could show incorrect phase where normal polarity data could be shown with phase difference +/-180 degrees.
Smoothing used for final phase plot, to reduce noise.
ITZwire version 1.104 - 17/10/2005
This version of ITZwire uses a routine to reduce the data to the expected sweep, BUT will ignore the self-calculated values if the user has already specified the start and end time/samples.
The plots with time information will use real (decimal) information rather than round-down to whole second as in previous versions.
ITZwire version 1.103 - 13/10/2005
Correction to the distortion plots. Previously these were labelled on the X axis by the frequency bandwidth so distortion could be cross-referenced against frequency. However the distortion was actually plotted against time (just labelled against frequency).
This version of ITZwire uses a routine to reduce the data to the expected sweep.
The routine detects when the data is within 1% of the maximum amplitude for the reference trace (normally the cleanest data) and sets the start and end time to suit.
If you get a reference trace with spikes after the sweep finishes, you will still see distortion in the distortion plot, after the sweep finishes, so this is now labelled by time.
So, you should not need to specify the sweep start and end time unless you get noisy/spiking data. If you see a lot of distortion in the distortion plot, check the maximum time on the distortion plot agrees with the expected sweep time. If not try setting the sweep start and end time to avoid the noisy/spiky data.
Reword phase plot to state comparison of phase of ground force against true reference.