Measurement of Timebase Jitter for Component Video

The present invention relates to timing measurements, and more particularly to the measurement of timebase jitter for component video.

In a gen-locked television system it is desirable to eliminate any phase jitter from the system. A phase jitter of as little as three nanoseconds can produce a noticeable chrominance aberration on edges when observed on a picture monitor. For composite video signals vectorscopes have been traditionally used to observe timing of the television system. However when only component video signals are available, vectorscopes cannot be used to measure timing errors including jitter. With the advent of component television studios and with the introduction of high definition television (HDTV) systems that use only component video signals, there is no measurement instrument comparable to the vectorscope for measuring timing jitter.

Therefore what is desired is a method of measurement of timebase jitter for component video that provides accurate jitter determinations of one nanosecond or less.

Accordingly the present invention provides a method of measurement of timebase jitter for component video by comparing the component output of a reference signal generator with the component output of a local phase locked signal generator. A sync signal from the reference signal generator is input via a first delay line to the local phase locked signal generator. The component video signals from the reference signal generator are input to a second delay line. The delay lines may be lengths of coaxial cable. A reference component video from one of the signal generators is compared with another component of the other signal generator visually on an oscilloscope or other type of waveform monitor. The timing between the two channels is adjusted to be equal, and the resulting jitter is then displayed or automatically processed to determine the timebase jitter of the television system.

The objects, advantages and other novel features of the present invention are apparent from the following detailed description when read in conjunction with the attached claims and appended drawing.

FIG. 1 is a block diagram of a system for measuring timebase jitter for component video according to the present invention.

FIG. 2 is a graphic display of a signal used to measure timebase jitter according to the present invention.

FIG. 3 is a graphic representation of timing convergence ideally and in the presence of timing jitter.

FIG. 4 is a block diagram of an alternate embodiment of a system for measuring timebase jitter according to the present invention.

FIG. 5 is a graphic representation of timing jitter as determined by the system of FIG. 4.

Referring now to FIG. 1 a reference signal generator 10 provides component video signals (GBR or Y,Pb,Pr) and a sync signal. The sync signal is input via a first delay line 12 to a local phase locked signal generator 14 that outputs component video signals. The component video signals from the reference signal generator 10 are input to a second delay line 16. Normally the G or Y video component is considered to be a reference component, and the other two video components are compared against this reference component to provide time coincidence of the outputs of the component video components. The reference video component from one of the signal generators 10, 14 is input to one channel of a dual channel waveform display device 18, such as an oscilloscope or television waveform monitor, and one of the other components from the other signal generator is input to the second channel of the waveform display device. The display device 18 is capable of inverting and varying the gain of channel B and summing the inverted, gain equalized signal with the signal on channel A.

A test signal that may be used for timebase jitter measurement is disclosed in U.S. Pat. No. 4,829,366 issued on May 9, 1989 to Bruce J. Penney entitled "Method and Apparatus for Measuring Delay and/or Gain Difference Using Two Different Frequency Sources." This test signal outputs two burst signals at slightly different frequencies, such as 5.00 MHz and 5.002 MHz, one for the reference component and the other for the other components, that are timed to be in phase in the middle of a video horizontal line. The resulting display, as shown in FIG. 2, is in the form of a "bowtie" with the "null" point determining the relative timing between the two component video signals input to the display device. The "bowtie" test signal also includes timing markers on the reference component channel. The timing between the reference component of one signal generator 10, 14 and one of the other components of the other signal generator is adjusted by the respective delay lines 12, 16 as well as by a reference timing adjustment control in the local phase locked signal generator 14 so that the null occurs at the middle of the horizontal line. Ideally as shown in FIG. 3 the intersection of the "bowtie" display should converge to a unique point. However due to timing jitter there is a region about the center point of the horizontal line within which the intersection is indeterminate. The width of this indeterminate region is the measure of the timing jitter for the local phase locked signal generator. The timing markers of the "bowtie" test signal are used to measure this indeterminate region. Since the component video signals are derived from a single master clock within the signal generator, once timing is established between the reference component of one generator and one of the other components of the other signal generator system timing is complete. Otherwise characterization of the other component channels is necessary in the same manner.

An alternative signal for automatic timebase jitter measurement is to use a test signal that provides a fixed frequency sinusoidal signal for both signal generators 10, 14, such as a five megahertz sinusoidal signal as shown in FIG. 4. Only the reference component from each signal generator 10, 14 is used. The reference signal generator 10 is coupled to the local phase locked signal generator 14 using the first and second delay lines 12, 16 as described previously. The reference components are input to respective attenuators 20, 22 to equalize the amplitudes between the two channels, and the reference timing adjustment in the local signal generator 14 together with the delay lines 12, 16 are used to adjust the timing between the two channels. The outputs of the two attenuators 20, 22 are input to a modulator 24, and the output of the modulator is input to a lowpass filter 26. If the reference component is V1*cos(wt), then the resulting genlocked component is V2*cos(wt+.phi.(t)) where V1 and V2 are the respective amplitudes, w is the carrier frequency and .phi.(t) is the phase jitter over time. The output of the lowpass filter 26 is V*cos(.phi.(t)) which is the timing jitter component. The timing jitter component is input to a processor 28, such as a fast Fourier transform (FFT) processor, and the results are displayed on an appropriate display device 30. As shown in FIG. 5 the result is a frequency spectrum about the carrier frequency shifted to d.c. from which the timing jitter is measured.

When viewed on an oscilloscope the jitter component .phi.(t) appears as a time instantaneous jitter, cos.phi.. Starting with V*cos(.phi.(t)), the signal is normalized to obtain cos(.phi.(t)). Taking the arccosine of the normalized signal produces .phi.(t). A phase-locked, or gen-locked, loop is defined to be in lock if -90.degree..ltoreq..phi..ltoreq.+90.degree. at the lock frequency. For the system to remain genlocked, the value of cos(.phi.(t)) has a magnitude less than one. The phase jitter is input to the FFT processor 28 to produce a spectrum about d.c. and each harmonic of the horizontal line rate, i.e., the spectral density of the jitter component. The time instantaneous jitter, before FFT processing, also is useful to observe the span of peak-to-peak jitter that may be expected in the system 14 under test. The spectral density after FFT processing may identify causes of the jitter, such as hum, interference and the like.

Thus the present invention provides a measurement of timebase jitter for component video by using an appropriate test signal and a reference signal generator, adjusting the timing between the reference signal generator and a local phase locked signal generator, and displaying the result in the form of an indeterminate region, the width of which is a measure of the jitter, or in the form of a spectrum about a d.c. transformed carrier signal.