No open area test site (OATS) is free of ambient noise. If you can receive an FM radio signal or make a cellular phone call, you are in an environment with RF ambient signals.
Such is the case in the San Francisco Bay area. Nearly every television station in the area broadcasts with digital signals. As a result, digital TV will add to the area’s already healthy supply of RF ambients.
Before addressing the challenge presented by digital TV, it’s important to understand how Elliott Laboratories generally performs OATS testing in the high-ambient city of Sunnyvale, CA. We have provided compliance testing services to hundreds of Silicon Valley companies for nearly 20 years so we are very used to the question, “How can you perform OATS testing in this area?” We quickly quell ambient concerns by demonstrating reliable and reproducible measurements of radiated emissions.
Generating reliable OATS test results in this area requires a combination of technical acuity and test facilities designed for a high ambient location. While many OATS facilities use spectrum analyzers for performing measurements, we use Rohde & Schwarz receivers which have a tuned front end and excellent selectivity. Our solid ground planes, on which the OATS sites are constructed, cost significantly more than the typical mesh alternative but provide greater stability at higher frequencies.
In addition to special site construction, we use protocols such as signal substitution for resolving and eliminating ambient noise from test measurements. The recent introduction of digital TV signals provided us with the opportunity to put these protocols to the test.
KGO, a local network television station, agreed to help us confirm the accuracy of our techniques to measure emissions among digital TV signals. Upon our request, KGO turned on and off a test broadcast so we could take measurements with and without the digital TV ambients.
We chose a product to test that we knew generated RF at the same frequency (531 MHz) as one of the strong digital TV ambients. At this frequency, we measured less than a 1.0-dB difference when the test broadcast was turned off and then turned on, which is well within the range of repeatability for OATS testing.
Measurements also were performed using a second EUT so that there were two narrowband signals close to each other but not at exactly the same frequency. Again, the difference between the two sets of data was less than 1.0 dB.
How We Did It
There are many ways to make signal measurements at the same frequency as an ambient signal such as FM and TV broadcast or cellular phone signals. The method used depends on the nature of the EUT signal and that of the broadcast signal.
RF signals generally are categorized as either broadband or narrowband. With a narrowband signal, the energy is focused at a single frequency. A broadband signal has its energy distributed across a frequency range.
Signals that are repetitive in the time domain, such as clock signals, have a spectrum made up of discrete, narrowband signals. Narrowband signal sources in an EUT usually are from digital or analog oscillators and clocks. Typically, it is these narrowband emissions that dominate the emissions profile of an EUT.
Narrowband ambient signals usually emanate from:
- Analog TV.
- Radio stations such as the FM band.
- Local transmitters such as cellular telephones.
Many narrowband ambient signals, such as FM and analog TV broadcast signals, also are modulated with voice or data and have an easily identifiable shape when viewed on a receiver or spectrum analyzer.
Signals that are not repetitive are typically broadband signals. Broadband sources in an EUT tend to come from power supplies since the fundamental switching frequency of a power supply varies slightly with time, data modulated signals, and spread-spectrum oscillators.
Broadband ambient signals usually emanate from:
- Combustion engines, which usually cause problems below 300 MHz.
- Spread-spectrum radios at known frequencies and of a relatively low level.
- Digital TV broadcasts.
Because digital TV signals are broadband, they are more difficult to resolve from EUT signals than narrowband analog TV signals. Digital TV signals span nearly 6 MHz as compared to the signals from analog TV video and audio signals (Figures 1 and 2).
Measurement Techniques
EUT and Ambient Signals, Both Narrowband
Before you can differentiate between narrowband EUT and ambient signals, you first have to resolve them. Our operators accomplish this by reducing the bandwidth of the receiver or analyzer.
Then, it is relatively straightforward to maximize the signal from the EUT. Once the signal is maximized, there are three options:
- If the two signals are far enough apart in frequency, the receiver can be adjusted to tune out the interference source.
- Use a signal substitution measurement where a signal generator is used to produce a narrowband signal of the same amplitude as that from the EUT as observed on the receiver or spectrum analyzer. The signal-substitution method always is performed at the correct bandwidth. This technique is useful on signals without modulation.
Sometimes the influence of signal modulation can be estimated by measuring a harmonic of the signal where the ambient is not present and observing the difference in measurements between the signal substitution and the true EUT signal at this harmonic frequency. This method is widely accepted and has been verified to produce accurate results.
If necessary, move the antenna closer to the EUT to increase the amplitude of the EUT relative to the ambient signal. The resulting measurement then is extrapolated back to the correct test distance as detailed in the emissions standards.
EUT Signal Narrowband, Ambient Signal Broadband
A broadband ambient signal above the specification limit, when measured with the 120-kHz bandwidth required for measurements between 30 MHz and 1 GHz, typically will be well below the limit when measured with a narrower bandwidth. If the noise is uniformly spread across its frequency range, then reducing the analyzer or receiver bandwidth from 120 kHz to 10 kHz could decrease the measured ambient level by more than 10 dB.
Once the signal is maximized, the signal-substitution method can be used. Additionally, the narrowband EUT signal is very obvious to a trained eye when viewed amidst broadband noise (Figure 3).
For our digital TV broadcast test, the signal-substitution method was used for each of the two EUT signals and resulted in a signal-substitution measurement of 35.7 dB. The EUT then was measured directly with the ambient digital TV signal turned off and measured 36.5 dB.
Conclusion
It’s undoubtedly easier to perform OATS testing in low ambient environments. However, we continue to research and test new methodologies to remove ambient concerns from test results.
About the Authors
Eddie Pavlu is COO at Elliott Laboratories. He received a B.S. in electrical engineering from Fairleigh Dickinson University and an M.S. in electrical engineering from Stevens Institute of Technology. e-mail: [email protected].
Mark Briggs is the manager of EMC consulting services at Elliott Laboratories. He earned a BEng in electronic engineering and an MSc in electromagnetic compatibility from the University of York in England. e-mail: [email protected].
Elliott Laboratories, 684 W. Maude Ave., Sunnyvale, CA 94086, (408) 245-7800.
Copyright 2000 Nelson Publishing Inc.
May 2000