If you are tasked with measuring video signal crosstalk in distribution circuits

If you are tasked with measuring video signal crosstalk in distribution circuits, and want to know if the VM700T might be useful for this type of measurement, the short answer is "yes" - crosstalk testing can be accomplished with the VM700T.

The longer answer is a bit more involved because there is no "one button" test for crosstalk in the VM700T menus.  And, there are no written procedures on how to do this with the VM.  You will have to use available measurement applications, a little common sense, a little engineering ingenuity, and perhaps a little experimentation. 

Some important notes...

Perhaps the first item of note is that the VM700T channel to channel isolation (crosstalk) specification (for the instrument itself) is greater than 70 dB at Fsc.  In a situation where your test spec goal is, for example, more than 65 dB of isolation, this spec gives you about 5 dB of test margin - which should be adequate. 

The second thing to note is that if you have multiplechannels going through the device or system under test, you will need to make multiple tests.  Since the VM700T only has three input channels, you will need to perform some of the test by connecting and disconnecting the channels of interest in various combinations.  For example, if your system has 4 channels, there will be about 12 combinations of signal pairs to test if you test crosstalk in both directions ---- 1-2, 2-1, 1-3, 3-1, 1-4, 4-1, 2-3, 3-2, 2-4, 4-2, 3-4 & 4-3.  You may get some efficiency, or some interesting results, if you test three channels at once - which is the input limitation for the VM700T.  For example, channel 1 into 2 and 3.  And, one should certainly try these extra combinations.   One should also try various test scenarios with some or all of the device channels loaded.  I would recommend trying tests with signals applied to any 2, then any 3, then all 4 of your paths.  Fully activated circuits will often perform different from partially activated circuits. 

The third thing to note is that you need to obtain/use a good generator of high signal quality – and with specific test signals available.  You will probably achieve your best
(or easiest) results with a signal generator that can provide a 100% Red Field analog composite video signal.  If the generator can also provide this in a Y-C output format, it may be useful - but the use of Y-C may not be something that your system will not handle.   (The C or chroma component does not have sync.)  The Tektronix TG700/AVG7 or TG2000/AVG1 are able to provide 100% Red Field signals, in both composite and Y-C formats.  The 100% Red Field is recommended because it contains the highest amount of energy at Fsc, and would therefore be easiest to observe.  Other signals can be used, and one should always consider looking at possible crosstalk at frequencies created by sync signals. 

Please note that the following is offered as a set of general guidelines and not as specific
step by step procedures.  Your situation may be similar, or it might be quite different.  But, you should be able to obtain good repeatable results if you follow these general guidelines - and you test / experiment.

The following is offered as one possible test scenario.    

The first task is to characterize your signal generator.
-          Connect your test signal generator composite (or Y-C) outputs directly to the VM700T and run the Noise Spectrum application. 
-          Set the test signal generator to a 0% white field.
-          With a 0% Flat Field signal applied, you will still see a small peak of signal at Fsc.  This is residual subcarrier - the 3.58 (or 4.43) MHz stray signals that are "floating around in even the best generators.   You need to know where this is (and how much) so that when you run the signal through your system, you will be able to account for the "original anomalies".     For example - If, after running the signals through     your system, this small frequency component is now greater in amplitude, the additional contribution is from your device / system.

By the way, there is another point to keep in mind here.  The residual noise floor of the VM700T input channels is spec'd down to -80 dB.  This is also the bottom limit of the Noise Spectrum Measurement range.  The measurement will show information at lower levels, and the signals and numbers will be pretty close - but below -80, they will be outside specified operating parameters.     

Now that you know the performance characteristics of your generator (and VM), it is time to apply the signals to your system.
-          Apply a 0% White Field (Black Burst) to one of the system channels.
-          Apply a 100% Red Field to another one of the system channels. 
-          Apply the system output channel with the 0% White Field to VM700T channel A.
-          Apply the other output signal (100% Red Field) to channel B. 
-          Set up the VM700T so that the Sync reference is only derived from one channel –
         Channel A will do for this discussion.
-    In the Waveform Mode…
          Press the Menu button
          Touch the Sync softkey
          Choose A as the sync source and disable the "Lock to Source" feature. 
              (Touch the softkey to un-illuminate it.) 
-          Go back to waveform mode and select any line in the active video field
         (line 100, for example)
-          Go to the Measure Mode and select Noise Spectrum.
-          In the Noise spectrum mode, enable the cursor.
           Press the Menu button
           Touch the Cursor softkey.
           Touch the Cursor 1 Active softkey.  The cursor will appear and can be moved to                       the component of interest with the front panel knob.  If you set it on the Fsc            component, at the bottom of the screen, you will get the signal level value (in dB), and the frequency.

Now, choose input B as your source.  You should see the 100% Red signal as a very tall (near to 0 dB) frequency component at Fsc. 

If you now select the A channel, the signal component at Fsc will consist of the residual subcarrier of the original generator, plus the added crosstalk signals from the system.  If the levels are still below -65, you can be fairly sure that the added crosstalk components are well under your -65 target.

At this point, you should also add 100% Red signals to the other input ports of your system in order to see if the total additive effects "crosstalk" back into the channel of interest.   Be sure to try all reasonable combinations of channels through your system.

Another note.  It is important to terminate any unused outputs of the signal generator, any unused inputs and outputs of the system under test, and any unused inputs on the VM700T.  You want to minimize the possibility of other stray signal effects getting into your measurement problem. 

Make your observations, record your results.


Within the VM700T, there is another tool that can be used to better visualize or separate signal errors caused by the source generator.  This is known as the Relative to Reference mode.  In the Noise Spectrum menu items, this is one of the choices that are available at the bottom of the screen.  Relative to Ref has the capability of subtracting the original generator characteristics from the system measurement - giving the net differences that are the actual system crosstalk components that you are looking for. 

To use this feature…
Go back to the previous paragraphs where you are first characterizing your signal generator.  With the 0% Flat Field signal (black burst) applied to input channel A on
the VM700T, and with the Noise Spectrum measurement running,...
-          Press the Menu button
-          Touch the Reference softkey. 
-          Then, touch the Store Reference 1 (or 2) softkey to save the reference data.

  Now, put your system back into the test signal loop, and…
  Again, select channel A - the blackburst signal in the Noise Spectrum mode.
-          Press the Menu button.
-          Touch the Relative to Ref softkey. 
-          Touch the Use Reference 1  (or 2) softkey - where the first generator data sample was stored.

At this time, if you now see a frequency component at Fsc, it's amplitude will be veryclose (within a dB or so) to the device/system’s contribution only.  Most all   contributions from the generator and the VM700T have been effectively subtracted out.

It is important to make note of the VM700T's noise floor and the limits of the Noise Spectrum measurement at this point.  In the Relative to Ref mode, the digitizer will still try to subtract and display, but once you go below about -90 dB, the display will start
to look "odd".  This is normal.  The digitizer system has hit its lowest limits.  However, if you are only interested in the components at Fsc, a spike above -80 dB A spike at 4.43 MHz (or 3.58) will be a useable indication of your crosstalk situation.

Some additional cross-checks can be done by using other applications in the VM700T such as the Signal to Noise measurements (in the Auto mode), or by expanding the Waveform mode vertically to look for presence of a known signal components in the "unknown" channel.  These are more difficult to explain in text, and require a bit more engineer skill levels.  And, neither of these modes will specifically give you information about crosstalk at your specific frequency of interest (Fsc of 4.43 MHz).   The Noise Spectrum techniques will be the easiest to implement, the easiest to visualize, and probably the easiest to explain/document.      
Prime Members Start Your Free Trial of Britbox with Prime Video Channels

No comments