TESTING TELEVISION TRANSMISSION SYSTEMS

FOR MULTICHANNEL SOUND COMPATIBILITY

BY:

GEOFFREY N. MENDENHALL, P.E.

VICE PRESIDENT OF ENGINEERING

B ROADCAST ELECTRONICS INC.

Q UINCY, ILLINOIS

4100 N. 24th ST. 9 P.O. BOX 3606 QUINCY, IL. 62301 o TELEPHONE (217) 224-9600

BACKGROUND INFORMATION

Although stereo programming is not yet widely available to most television stations, there is great

interest in testing, evaluating, and preparing each station’s transmissions system for multichannel

transmission. One approach to testing, is to measure the various audio parameters including fre-

quency response separation, crosstalk, and distortion using a B-TSC encoder with the aural exciter

and a consumer type BTSC receiver-decoder for the “off-air” measurements. This approach may

lead to an incorrect evaluation of the transmission system due to errors in the encoding/decoding

process since precision encoders and precision modulation monitor/decoders are not yet available

with guaranteed minimum performance specifications that are applicable to the entire system.

This engineering application note describes a procedure for evaluating television transmission

systems using readily available standard test instruments and without relying on BTSC encoding and

decoding equipment.

The primary objective is to transmit the BTSC composite waveform to the stereo decoder in

the receiver at the correct level (deviation of the aural exciter) and without altering the amplitude and

phase relationships of the various components within this waveform. The composite signal path
from the output of the BTSC encoder to the input of the decoder in the receiver is subject to many

interacting and cumulative errors so it is necessary to devise a test procedure that can identify the

magnitude and type of error within each functional block in the system.

PROBLEM AREAS FOR THE COMPOSITE SIGNAL

There are three areas for signal degradation to occur:

1. The composite link to the TV aural FM modulator.
(BTSC stereo generator, SAP generator, PRO generator, composite processor, and STL equip-
ment)
2. The aural FM modulator.

3. The RF path to the demodulator.

(aural exciter, IPA, PA, diplexer, and antenna system.)

Each of these three areas has its own special effect on the baseband signal and each subsystem must

be individually optimized before the complete transmission system can give the best possible perfor-

mance.

THE COMPOSITE LINK

The composite path from the stereo, SAP, and PRO) generators to the aural FM! modulator

should be linear in both amplitude –vs- frequency and in phase -vs- frequency response. Simply

stated, this means that no frequency component within the baseband should be attenuated more than

any other frequency component. Furthermore, all frequency components should propagate thru the

system at the same speed (constant group delay) and thus arrive at the modulator at the same time.

Fquation-1A- and Eauation-1B Mathematically relate stereo separation to amplitude response.

Equation-2A- and Equation-2B mathematically relate stereo separation to phase response.

Figure-1 graphically shows the combined effect of amplitude and phase response on stereo
separation between the right and left channels.

Correct phasing and equal group delay of the (Fh) pilot tone is also essential to achieving

stereo separation.

The final stereo performance of the complete system will be determined by the algebraic

summation of the individual composite amplitude response and composite phase response of each

device within the composite signal path.

The aural exciter, STL link, and any other composite device should specify these composite

performance parameters so that total system performance can be easily predicted. In order to main-

tain a system separation capability of 4OdB as suggested by Zenith, the composite amplitude re-

sponse must be within +/- 0.17dB (5OHz to 47KHz) and the composite phase response must be less

than +/-1.15 degrees from linear phase (5OHz to 47KHz).

 COMPOSITE PPOCESSING

In an effort to achieve maximium modulation density (loudness), some FM broadcasters use

composite processing to remove the low energy overshoots in the amplitude of the composite

wavefom caused by complex audio input filtering. Overshoots will also occur in the peak to peak

amplitude of the BTSC composite waveform, but are not considered significant to the lower modu-

lation density (wider dynamics) desired in television broadcasting. Since overshoots have no effect

on compandor tracking or any other audio performance parameter other than achieving the last dB in

loudness, composite processing is not recomended for use with the BTSC system. The use of any

non-linear devices, such as clippers or limiters in the composite line will alter not only the peak

amplitude of the baseband, but also the frequency spectrum of the baseband. This generates several

types of distortion at the receiver.

Figure 2A and Figure 2B show the waveform and spectrum of unprocessed baseband while

Figure 2C and Figure 2D show the waveform and spectrum after l.0dB of composite clipping.

SUMMARY OF TYPES OF DISTORTION

CAUSED BY COMPOSITE PROCESSING

1. Intermodulation of all baseband frequency components causing extraneous spectral

components
2. Harmonic distortion of baseband causing degradation of crosstalk and separation.
3. Modulation of pilot injection level causing loss of lock at the synchronous detector.
 The received audio is high in intermodulation distortion and non-correlated information due to

aliasing of the extraneous spectral components added by composite processing. If minimum, system

distortion is the goal, composite processing should not be used. Audio processing should he per-

formed before the audio is multiplexed into baseband.

Distortion of the composite baseband signal, can also be caused by transient intermodulation

distortion (TIM) within the. amplifier stages. Transient intermodulation distortion of the baseband

signal is caused by the same mechanisms that produce TIM in audio signals. The composite amplifi-

ers must have sufficient feedback bandwidth to accept baseband frequencies to l00kHz and should

slew symmetrically to minimize slew-induced distortion. The TIM performance becomes largely a

matter of operational amplifier selection and circuit configuration.

AURAL MODULATOR LINEARITY

The composite baseband signal is translated to a frequency modulated carrier frequency by the

modulated oscillator. Frequency modulation is produced by applying the composite baseband signal

to a voltage tunable RF oscillator. The modulated oscillator usually operates at the carrier frequency

and is voltage tuned by varactor diodes, operating in a parallel LC circuit.

To have perfect modulation linearity, the P.F output frequency must change in direct propor-

tion to the composite modulating voltage applied to the varactor diodes. This requirement implies

that the capacitance of the varactor diodes must change as nearly the square of the modulating

voltage.
 Unfortunately, the voltage versus capacitance characteristic of practical varactor diodes is not

the desired square law relationship. All varactor-tuned oscillators have an inherently non-linear

modulating characteristic. This non-linearity is very predictable and repeatable for a given circuit

configuration, making correction by complementary predistortion of the modulating signal feasible.

Suitable predistortion can be applied by using a piece-wise linear approximation to the desired

complementary transfer function.

Any distortion of the baseband signal caused by the modulated oscillator will have secondary

effects on stereo, SAP, and PPO crosstalk, which are quite noticeable at the receiver in spite of the

rather small amounts of distortion to the baseband. For example, if the harmonic distortion to the

haseband is increased from .05% to 1.0%, as much as 26dB additional crosstalk into the SAP can be

expected.

THE RF PATH

THE AURAL TRANSMITTER SIDEBAND STRUCTURE

The frequency modulated RF output spectrum contains

many sideband frequency components, theoretically an infinite number. They consist of pairs of

sideband components spaced from the carrier frequency by multiples of the modulating frequency.

The total transmitter RF output power remains constant with modulation, but the distribution of that

power into the sidebands varies with the Modulation index so that power at the carrier frequency is

reduced by the amount of power added to the sidebands.

OCCUPIED BANDWIDTH

After examining the resulting spectra, it becomes clear that the occupied bandwidth of an FM

signal is far greater than the amount of deviation from the carrier that one might incorrectly assume

as the bandwidth. In fact, the occupied bandwidth is infinite if all the sidebands are taken into

account, so that a frequency modulation system requires the transmission of all of these sidebands

for perfect demodulation of information. In practice, a signal of acceptable quality can be transmit-
ted in the limited bandwidth assigned to the TV aural channel.

EFFECTS OF BANDWIDFH LIMITATION

Practical considerations in the transmitter RF circuitry make it necessary to restrict the RF

bandwidth. As a result, the higher order sidebands will be altered in amplitude and phase. Band-

width limitation will cause distortion in any FM system. The amount of distortion in any practical

FM system will depend on the amount of bandwidth available versus the modulation index being

transmitted.

LIMITING FACTORS WITHIN TIE AURAL TRANSMITTER

Relating the specific quantitat

iveeffect of the bandwidth limitations imposed by a particular

transmitter to the actual distortion of the demodulated composite baseband is a complicated problem

indeed. Some of the factorsinvolved are:

1. Total number of tuned circuits involved.

2. Amplitude and phase response of the total combination of tuned circuits in the P.F path.

3. Amount of drive (saturation effects) to each amplifier stage.

4. Non-linear transfer function within each amplifier stage.

IMPROVEMENT OF THE AURAL RF PATH

The following design techniques can help improve the transmitter’s bandwidth:
1. Maximize bandwidth by using a broadband exciter and a broadband IPA stage.

2. Use a single-tube design or a broadband, completely solid state, design where feasible.

3. Optimize both the input circuit and output circuit of the tuned stage for the best possible
bandwidth.

4. Minimize the number of interactive tuned networks.

5. Use a delay equalized multiple cavity diplexer.

6. Use a broadband antenna system with a low standing wave ratio at the aural carrier frequency.
 SYSTEM TEST PROCEDURE

The composite amplitude and phase characteristics must be measured to a high degree of

accuracy. (tenths of a decibel and tenths of a degree from phase linear).

A high accuracy audio network analyzer could be used to directly measure the composite

characteristics, but most stations do not have access to this equipment.

Another simple yet effective way to evaluate the system performance is to send a multi-tone

test signal consisting of a low (L+R) audio frequency and ultrasonic (L-R) frequency components of

equal values through the system and display the resulting waveform on an oscilloscope whose sweep

is synchronized to the low frequency audio component. The resulting waveform is shown in Figure-

3.

The amplitude of the (L+R) ancl (L-P.) components should be exactly equal at each point

throughout the composite system to the demodulator. The propagation time through the system

should also be equal for (L+P.) and (L-R) components. The key property of this test signal is that

the (L+R) and (L-R) components are equal (1:1 ratio) so that any change in this ratio due to system

problems can easily be observed on an oscilloscope. The composite signal output from the BTSC

stereo generator does not have a fixed and equal ratio between (L+R) and (L-R) so it cannot be used
for this test. Figure-4 shows what the BTSC composite baseband looks like if viewed on an oscillo-

scope with the peak-to-peak amplitude shown as a function of time. It is difficult to accurately

measure the amplitude ratio and phase relationship of (L+R) to (L-R) since the ratios vary with the

level of companding and are never equal.

The required (1:1) test signal can be obtained from a standard FM broadcast stereo generator

by turning the pilot off and modulating only one channel since the (L+R) and (L-R) information is

output in equal amounts under these conditions.

The TZ-30 TV stereo generator has a special test mode to provide the required 1:1 ratio test

signal with or without the (Fh) pilot tone.

INTERPRETING THE COMPOSITE WAVEFORM

During all of the tests the external trigger input to the oscilloscope is connected to the audio

generator which feeds only one input of the stereo generator. The other audio input is shorted and

the pilot is turned off. The composite output from a wideband RP demodulator such as the Boonton

model 82AD or the Hewlett-Packard model 8901A modulation analyzer is fed to the wideband

vertical input of the oscilloscope. The composite waveform can also be checked at other points

within the system to determine the error contribution from each subsystem. If both the amplitude and

phase response are correct, the base line of the waveform will be perfectly flat even when closely

examined by expanding the vertical scale on the oscilloscope as shown in Figure-5.

 EXPANDED SCALE TO CHECK BASELINE FLATNESS

An amplitude and delay equalizer for the composite baseband is available as part of the TZ-30

BTSC stereo generator. Equalization for amplitude and phase deficiencies in the STL or Aural

exciter will improve the overall system performance. The adjustments of the equalizer are made

while observing the demodulated composite baseband to minimize deviation from a flat baseline.
 If the baseline deviates from flat in a (curved or bowed) symmetrical manner as shown in

Figure-6A- and Figure-6B there is an amplitude error only.

 If the baseline deviates from flat in a (straight line) tilted rnanner as sbown in Fi(jure-7A and

Figtire-7B, there is a phase (time delay) error only.

MEASURING STEREO SEPARATION DIRECTLY FROM THE

COMPOSITE WAVEFORM

Figure-8 illustrates a composite waveform with a mixture of amplitude and phase errors as

indicated by the asymmetrical deviation of the base line from flat. The separation can be calculated

by taking twenty times the log to the base ten of the ratio of the total peak to peak value of the

waveform to the peak to peak deviation from flat base line. The sample calculation in Figure-8

shows a separation of approximately 28dB.

HOW TO ADJUST THE AURAL TRANSMITTER FOR BEST

BTSC-MCS PERFORMANCE

All optimization should be done with the transmitter connected to the normal diplexer and

antenna system. The transmitter is first tuned for normal output power and proper efficiency accord-

ing to the manufacturer’s instruction manual. The meter readings should closely agree with those

listed on the manufacturer’s final test data sheet.

A simple method for centering the transmitter passband on the carrier frequency involves

adjustment for minimum synchronous AM. Synchronous AM is a measure of the amount of inciden-

tal amplitude modulation introduced onto the carrier by the presence of FM modulation. This mea-

surement is very useful for determining the proper tuning of the aural transmitter. Since all transmit-

ters have limited bandwidth, there will be a slight drop-off in power output as the carrier frequency is

swept to either side of the center frequency.

 This slight change in RF output level follows the waveform of the signal being applied to the

FM modulator causing AM modulation in synchronization with the FM modulation. Minimizing

synchronous AM- will assure that the transmitter passband is centered on the aural channel. Care

must be taken when making these measurements that the test set-up does not introduce synchronous

AM- and give erroneous readings which would cause the operator to mistune the transmitter to

compensate for errors in the measuring equipment.

The input impedance of the envelope detector must provide a nearly perfect match so that there

is a very low VSWR on the sampling line. Any significant VSW- on the sampling, line will produce

synchronous AM at the detector because the position of the voltage peak caused by the standing

wave moves along this line with FM modulation. Unfortunately, the AM detectors supplied with

some modulation monitors do not provide a good enough match to be useful for this measurement.

Precision envelope detectors are available that present a good match (3OdB return loss) to the

sampling line.

A typical adjustment procedure is to FM modulate 100% at 400Hz and fine-adjust the

transmitter’s input tuning and output tuning controls for minimum 400Hz AM modulation as de-

tected by a wideband envelope detector (diode and line probe). It is helpful to display the demodu-

lated output from the AM detector on an oscilloscope while making this adjustment.

Note that as the minimum point of syncronous AM is reached, the demodulated output from

the AM detector will double in frequency to 80OHz, because the fall-off in output power is sym-

metrical about the center frequency causing the amplitude variations to go through two complete

cycles for every one FM sweep cycle. This effect is illustrated in Figure-9. It should be possible to

minimize synchronous AM while maintaining output power and efficiency in a properly designed

power amplifier.
 Another more sensitive test is to tune for minimum intermodulation distortion in left only or

right only stereo transmissions. Stereo separation will also vary with tuning.

For stations employing a SAP, transmitter tuning becomes very critical to minimizing crosstalk into

the SAP. Modulate one charmer only on the stereo generator to 100% with a 7867Hz tone. This

will place the upper second harmonic (L-R) stereo sideband on tot) of 78.67KHz SAP. Activate the

SAP at normal injection level without Mutation on the SAP. Tune the transmitter for minimum

output from the SAP demodulator. This adjustment can also be made by listening to the residual
SAP audio while normal stereo programming is being broadcast.
FIELD ADJUSTMENT TECHNIQUES
1. Tune for minimum synchronous AM noise.

2. Tune for minimum IMD in left or right only channel.

3. Tune for minimum crosstalk into unmodulated SAP subcarrier.

In any of these tests, the input tuning is frequently more critical than the plate tuning. This is

because the impedance match into the input capacitance becomes the bandwidth limiting factor.

Even though the amplitude response appears flattened when the input is heavily driven into satura-

tion, the phase response still bas a serious effect on the higher order FM sidebands.

TEST EQUIPMENT SET-UP

Figure-10 shows block diagram of the required test equipment set-up for making composite wave-

form measurements. Note that the composite baseband is checked at various points along the trans-

mission path in order to verify the performance of each subsystem.

 Observing the composite waveform while using a low modulating frequency of 5OHz will

usually indicate any low frequency problems due to coupling capacitors in the system that are of

insufficient size. Composite tests using a high modulating frequency of 15kHz will reveal rolloff in

the high frequency response of the system which attenuates the (L-R) components more than the

15kHz component.

A precision envelope detector is also included in the test set-up so that the synchronous AM

waveforms can be observed while tuning the aural transmitter.

 ACKNOWLEDGMENTS

The author wishes to thank Rick Carpenter and Bill Resch for their assistance in conducting

the tests and editing this paper.

Special thanks to Katliy Klingler for typing and word processing, Jeff Houghton for the illus-

trations, and Kim Dopheide for word processing.

THE AUTHOR

Geoffrey N. Mendenhall. earned his BEE degree front the Georgia Institute of Technology in

Atlanta, Georgia.

Mr. Mendenhall has designed communications and telemetrv equipment for E.F. Johnson and

Harris Corporation. He led tbe design efforts for both the Harris MS-15 product line and the Broad-

cast Electronics FX-30 FM exciter.

His practical field experience has involved engineering and operations work for several radio

and television stations.

Mr. Mendenhall is presently serving as Vice President of Engineering for Broadcast Electron-

ics Inc. in Quincy, Illinois. He is an active amateur radio operator and is experimenting with home

satellite reception.

The author holds two U.S. Patents for electronic designs and is a registered professional

engineer in the State of Illinois. He has authored numerous technical papers and is member of the

Institute of Electrical and Electronics Engineers.