Since the shut-off of analog signals, most broadcasters have received calls from viewers regarding their inability to receive the station's signal. The quick answer is to check the theoretical footprint and assure viewers that they should be getting a good clean signal, and ask them to adjust their antenna or check the setup on their receiver. But after the call, a broadcaster is left wondering how the RF signal quality really is in a certain area.
When investigating reception complaints, there is a more productive approach to take since most broadcasters are monitoring and testing all video and MPEG signals in their facilities. They can simply extend this monitoring function to analyze the transmitter's RF signal and specific viewer locations. The RF leg in the signal chain is vital to ensuring quality and reliable reception.
Tools for taking 8-VSB measurements are available in different forms, including handheld SLMs and USB PC probes. They show and quantify problems, and also relate reception difficulties to their causes. These simple measurements can also point to transmitter or transmit antenna issues that may require attention.
Five simple measurements for proving signal quality
An 8-VSB measurement receiver provides analysis and insight into transmission or reception impairments with several key analyzing tests. These include:
- Visual — constellation and eye diagram analysis;
- Quantitative MER, EVM analysis;
- BER analysis;
- Level dBmV measurement; and
- Digital channel spectrum analyzer analysis.
Constellation diagram analysis
The receiver's demodulator is locked to the pilot carrier recovering I and Q amplitude and phase samples at the center of the symbol sample time. Plotted on a visual chart called a constellation diagram, these samples indicate how the detected I and Q signals vary from the ideal. (See Figure 1.) The I signal amplitude is used to recover the symbol values (-7 to +7).
The vertical lines on the constellation diagram represent the eight amplitude levels in 8-VSB. If I/Q samples fall on the vertical lines, this indicates the I (amplitude) component is ideal to the 8-VSB symbol.
Note: The position of the sample measured vertically along the symbol line indicates the value of the Q component. The Q component is not used for 8-VSB symbol recovery but is transmitted and used in signal quality analysis.
RF propagation and/or receiver influences can cause the sample points to fall to the left or right of the lines. When plotted I/Q samples are close to the lines, the receiver easily distinguishes the eight discrete symbol levels and recovers the proper symbol values. If the samples spread away from the lines, this indicates detected I/Q signal variations. If the variations grow too large and samples for one symbol level cross over to the adjacent symbol level, a symbol error results. While subjective, the constellation diagram provides a comprehensive visual indication of the 8-VSB channel signal quality.
Note: An eye diagram also shows RF reception signal quality. The seven center eye openings should be open and defined. Worsening signals collapse the eyes. (See Figure 2.)
MER, EVM
When time demands a single quantitative value to assess signal quality and relate it to the ability of a receiver to decode properly, the modulation error ratio (MER) measurement is referenced. MER is the ratio of the power of the signal to the power of the error vectors expressed in dB. For simplicity, MER is often described as the digital channel equivalent of an analog signal-to-noise ratio measurement.
Considering the constellation diagram, MER compares the actual location of a received symbol (I and Q voltage vectors) with its ideal location. (See Figure 3.) As the signal degrades, the received symbols are located further from their ideal locations and the measured MER values decrease. As location errors further worsen, the MER values decrease eventually to the point that symbols are incorrectly interpreted as adjacent symbols.
The larger the MER value, the better the signal quality. Typical performance of an 8-VSB transmitter should approach, if not exceed, a MER value of 30dB. A typical 8-VSB receiver can decode down to a MER of approximately 15dB. However, increasing signal power to a MER of 18dB or greater helps avoid teetering on the digital cliff. (See Figure 4.)
Note: A margin measurement associated with MER represents how far the MER value is from a typical receiver's threshold of visibility (TOV) or “digital cliff.” A typical TOV for 8-VSB is 15.2dB.
MER measurements may be expressed as an error vector magnitude (EVM) value. EVM is a percentage RMS value that represents the amplitude ratio of the RMS error vector amplitude to the largest symbol amplitude. EVM is the opposite of MER, since the lower the EVM percent value, the better the signal quality. EVM values for receivable signals range from 2 to 11 percent. EVM values above 10 percent quickly approach the TOV cliff. (Note: MER of 18 = EVM of 11.7 percent.)
BER analysis
To this point the measurements have been all demodulator-based, reflecting RF impairments. Further measurements gauge a receiver's ability to recover the MPEG packets. BER measurements indicate in real time the receiver's ability to decode and correct errors in the MPEG packets.
BER is a ratio of the number of bit errors to the total number of bits sent in a 1-second interval. The BER ratio or result should be a very small number, and it is expressed as a number times 10 with a negative exponent as the multiplier. A BER measurement is made possible by error correction encoding, providing the receiver with the ability to detect and count errors of the actual digital RF channel in real time.
The BER measurements before (pre) and after (post) a Reed-Solomon decoder indicate the uncorrected and corrected errors. A before-Reed-Solomon decoder or pre-FEC BER value is the raw bit errors in the digital payload. Values lower than 1×10-5 are considered acceptable, with 1×10-9 representing a near perfect signal. The after-Reed-Solomon correction or post-FEC BER value is the BER after all corrections (improvements) have been made. This value should be numerically lower than the pre-FEC BER, with values lower than 1×10-6 considered acceptable. (Note: Lower values have larger negative exponents 1×10-7.)
Some analyzers include a segment error rate (SER) measurement. SER is the number of MPEG segment errors per second. A small SER value is desirable, with 3 (typically averaged over a 20-second window) occurring at TOV. Once segment errors start to occur, the SER value rapidly increases as the signal worsens.
Level dBmV
While level measurements of an 8-VSB RF signal are less important than the other digital measurements to ensuring reception, they are needed to ensure a sufficient level is being input to the receiver to be above its noise floor. It is also crucial in identifying excessive signal levels, which can overdrive the receiver's input circuitry. This results in mixing and noise products that can cause signal degradation and poor signal quality.
Measuring a digital channel's signal level is different from sampling the NTSC video carrier at peak power sync times since the 8-VSB signal has no high-power carrier, and the spectral energy is random. To measure RF channel level and determine a single value, the average level existing through the channel band must be determined. This requires the energy be sampled at multiple points throughout the band and averaged into a single dBmV level measurement.
Digital receivers can easily be overdriven, and input levels above +15dBmV should be avoided. Factors such as attenuators, improved antenna location, antenna gain, directivity, download cable size or preamplifiers can improve reception signal power. While reception is possible to levels approaching -30dBmV, all other aspects of the transmission path and receiver's circuitry must be ideal. Improving reception levels to -20dBmV or greater reduces the chance of picture breakup when signal deterioration approaches the digital cliff.
Analyzing the digital channel spectrum
No 8-VSB RF analyzer would be complete without a spectrum analyzer to visually show the signal power as it is dispersed across the channel band. (See Figure 5.) Variations from the ideal flatness or even dispersion of signal energy through the center of the 5.4MHz channel band indicate influences of frequency and phase distortions. These variations may be transmitter- or transmission-path-related.
To conclude, a professional approach to answering 8-VSB broadcast-quality or reception coverage area questions is provided by the key 8-VSB analyzing tests summarized in this article. Routine field use and location monitoring provides RF performance information crucial to ensuring the health of the RF signal chain and ensuring reliable reception for viewers.
Glen Kropuenske is product manager in the Signal Quality Division at Sencore.
