Synchronous Vs Envelope Detection Updated 10/05/01

Some demodulators provide a choice between envelope and synchronous detection modes. When running the FCC Proof of Performance tests, its important to use the proper detection mode. Especially when measuring differential phase. Let's take a look at the two modes and see what they have to offer.

Envelope Detectors

Envelope detectors are usually considered to be the simplest form of detector. In fact, they can be quite complicated or they can be as simple as a diode and a low-pass filter. The performance of an envelope detector, when used to demodulate a television signal, is limited by the relative amplitudes of the signal being demodulated (bigger is better), and an effect called quadrature distortion. Quadrature distortion is a result of the asymmetry of our television signal. It's vestigial sideband coming from the modulator or transmitter and nearly single sideband within the television's IF circuit. One nice thing about envelope detectors (in addition to their low cost) is that they are not sensitive to the phase of the visual carrier. This will become important when we look at differential phase tests.

Synchronous Detectors

Synchronous detectors are considerably more complex than simple envelope detectors. They consist of phase locked loop and multiplier circuits. Demodulation is performed by multiplying the modulated carrier by a sine wave that is phase locked to the incoming carrier. Synchronous detectors are a subset of "product" detectors. If you are an amateur radio operator, you may have listened to single sideband suppressed carrier signals using receivers with a BFO for re-inserting the carrier and envelope detection; and other receivers with real product detectors. SSB signals sound much better using product detectors.

The advantage of synchronous detection is that it causes less distortion than envelope detection and works well with single sideband signals. It is the preferred detection method for most tests. Another characteristic of synchronous detectors is that they are phase sensitive. The amplitude of the demodulated signal is a function of the relative phases of the incoming carrier and the carrier generated inside the receiver. In the extreme case, if the phase of the modulated carrier and the regenerated carrier in the demod is 90 degrees (they are in quadrature) the detector output would be zero!

Insidious ICPM

The phase sensitive nature of synchronous detectors allows us to measure Incidental Carrier Phase Modulation (ICPM) of the signal being evaluated. ICPM sounds complicated but its not as bad as you might think. Remember that the video signal is amplitude modulated. If the phase of the visual carrier changes in response to the modulating (video) signal, we have ICPM, incidental carrier phase modulation. If ICPM is present, the most likely indication from a television set is sound problems. That may be a little surprising. However, the sound circuit inside most television receivers uses the frequency difference between the visual and aural carriers to produce the 4.5 MHz sound IF. If the phase of the visual carrier is changing in response to the video signal, then the phase difference between the visual and aural carriers is changing. This is essentially the same as the aural frequency being modulated by the video signal. The result can be video related sound buzz and/or poor stereo separation. If the demodulator has two synchronous detectors, one in phase with the incoming signal and one at 90 degrees (in quadrature), ICPM can be measured. If there is no ICPM, the quadrature output should be zero. If ICPM is present the amount of carrier phase shift will be reflected in the quadrature output. By applying the quadrature output to the x axis and the video output to the y axis of a display device, ICPM will be displayed as a tilt in the resulting display. Mathematically, the amount of ICPM is equal to ARCTAN (quadrature signal/in-phase signal). The test signal used for ICPM measurements is the linearity staircase -- the same test signal we use to measure differential phase and gain.

The Problem

If ICPM is present, the phase of the visual carrier, relative to the regenerated carrier in the demodulator, is different at different levels on the linearity staircase.

There's also a 3.58 MHz subcarrier riding on the linearity staircase — it’s the one we use for diff gain and phase tests. If the phase of the visual carrier changes as we go up the linearity staircase, then the phase of the 3.58 MHz subcarrier also changes as we go up the linearity staircase. Differential phase is measured as the maximum phase deviation of the color signal relative to the reference segment at zero IRE on the linearity staircase. In this case (using a synchronous detector) ICPM looks exactly like differential phase!

The linearity staircase signal is used for both ICPM and Differential Gain & Phase measurements

How to get around the problem

The best way to avoid confusing diff phase and ICPM is to use an envelope detector for the differential phase tests. Envelope detectors are not phase sensitive and, therefore, do not respond to Incidental Carrier Phase Modulation. If your precision demod does not have an envelope detector (unfortunately several of the newer ones don't -- a major oversight on the part of the manufacturers) you may still be able to perform accurate differential phase tests if it has a quadrature output port. If it has only synchronous detection and no quadrature output port, it should not be used for differential phase tests! With no quadrature output, you have no way to differentiate between ICPM and differential phase. I have seen as much as 30 degrees of ICPM in some modulators. This is unusual and most modulators exhibit very little ICPM, but unless you can determine how much ICPM is present, you have no idea how much differential phase is present. Again, this is only a problem when using synchronous detection.

Accurate Differential Phase Measurements Using Synchronous Detection

If your demodulator has synchronous detection and a quadrature output, accurate differential phase tests results can be obtained by subtracting the amount of ICPM from the differential phase measurement on a step by step basis. Here's the process: Measure differential phase, relative to the phase of the packet at the bottom of the linearity staircase (0 IRE level) for each riser, and log the results (be sure to include the sign). Measure ICPM, relative to the lowest level of 100 IRE, for each riser in the linearity staircase, and log the results (be sure to include the sign). For each riser, subtract the ICPM number from the differential phase number. The largest magnitude of the calculated results is the correct number. Here are the results from of an actual test. The test was performed using a Tek 1450 demod that has both synchronous and envelope detectors (so I could compare the results) and a Tek VM700A video test set. I used a full field FCC composite test signal for the test.

Measured using synchronous detection   D.P. Measured using envelope detection (compare to column 4) 1 Riser 2 Measured D.P. 3 Measured ICPM 4 Actual D.P. col 2- col 3 1 0.3 -0.4 0.7 0.8 2 0.0 -0.8 0.8 0.8 3 -0.8 -1.4 0.6 0.7 4 -2.0 -2.2 0.2 0.2 5 -2.9 -3.1 0.2 0.3

Notice that differential phase measured using the synchronous detector without correcting for ICPM was -2.9 degrees. The correct result was +0.8 degrees. An error of 3.7 degrees! This type of result is not unusual.

If you're familiar with the way the Tek VM700 displays differential phase, you might want to watch an animation demonstrating what happened during the above tests. Click here to see the animation

Let's get real!

OK, you need to be a techno-masochist to go through what I just described (but it works!). Obviously, this process would take a very long time for the typical cable system. So what to do? If you don't mind taking advantage of a loophole in the rules, you don't need to run the test on every channel. This is one of the tests that only needs to be performed on a limited number of channels (9 for a system that has a highest NTSC visual carrier frequency between 500 and 600 MHz, for example). Every channel is required to meet the specification, but you don't need to prove that it does (unless asked). So, you could run the tests on the minimum number of channels and let it go at that. A better approach is to take a quick look at ICPM for each channel before running the diff phase test. In most cases ICPM will be very small. Make a note of those that exceed a certain amount -- 2 degrees for example -- and run the thorough version of the tests only on those channels (or avoid them altogether). For the channels with ICPM of less than 2 degrees, as long as they are within +/- 8 degrees, they pass. The details of running the ICPM test are a little much for this paper. Check the instructions for your test gear for the details, or contact me for more information. Here are some demods that are capable of performing accurate differential phase tests. I realize that the list is not complete. If you know of other demods with both synchronous and envelope detection modes, or demods with only synchronous detection that also have a quadrature output, please forward the information to me at garya@tvms.net or use the "Contact Us" form. I'll add it to the table. Company Model Envelope Quad Out Comments           Tektronix 1450 Y Y No longer in production Tektronix 1350 Y Y  www.tek.com Tektronix DS1200 N Y No longer available Videotek DM-192 Y Y  www.videotek.com Videotek DM-154 N Y   Videotek DM-200 N Y    Modulation Sciences  MSI-320  N Y  www.modsci.com Rohde & Schwarz EFA 93 Y Y www.rohde-schwarz.com For accurate diff phase tests, the demod must have an envelope detector and/or quadrature output. If neither is available, you won't be able to tell whether you are measuring diff phase or ICPM! Back to Technical Papers Index

Home Products Tech Papers Contact Us

Gary Andrews Television Measurement Services garya@tvms.net www.tvms.net