Some recent posts requesting information about Spectrum Analyzers for use in WxSat reception prompted me to put together my experiences over the past 10 years struggling with inexpensive spectrum analysis instrumentation for 137 Mhz and 1691 Mhz as well as some popular i.f. frequencies.
Why use a Spectrum Analyzer? The analyzer is useful at 1691 and 1687 for various tasks. One of these is dish aiming, especially for VAS at 1687 Mhz. Every last drop of signal is crucial for low bit error rates in VAS reception. Because the signal bandwidth is 4 Mhz and the data rate is 2.11 Mbits, a good signal to noise ratio is essential to get anything at all. Aiming the dish by trial and error, while useful, is not ideal. Bit synchronization circuitry in the VAS front end requires a string of 1000 or so bits at the beginning of a 50 Millisecond data block to get locked in synch. Add to that the capture time for the PLL to acquire clock synch before frame sync can be established and we start getting a relatively long time period before we know whether we are "on target" with dish to satellite aiming. Moving the dish a little, waiting and then repeating the process a few seconds later gets old fast! Finding the precise aiming point is simpler with a Spectrum Analyzer - aim for peak reading. The analyzer is also useful for evaluating the system Noise Figure of feedhorn/LNA/Downconverter combinations. Degradation of any of these components can be evaluated quickly with the visual analysis afforded by the analyzer.
In the VHF (137 MHZ) world, the analyzer is useful for analyzing interfering signals from whatever source - a computer, American Airlines, or another satellite. Antenna patterns are easily evaluated by watching a pass on the analyzer. Remedial action is easy to evaluate - does Geoff Chesters' Helical Filter really tame the Wild Lindenblad? (It does).
For us old fashioned (cheap) folks manually tracking an HRPT dish, an analyzer is a must.
My first attempt at spectrum analysis was a so called "pan adapter" or fixed frequency analyzer centered at 10.7 Mhz. The trick is to use a communications receiver to tune the frequency of interest and watch the results in the receivers' i.f. bandpass. I used the Heath Panadapter for many years to watch i.f. signals. This has since been replaced by an invention called "The Poor Mans Spectrum Analyzer".
Ham Radio magazine in September, 1986 published plans by Murray Barlowe where a TV tuner was pressed into service as an analyzer. The way this thing works is a sawtooth waveform is repetitively fed into the tuning varactor, causing the tuner to sweep all frequencies from DC to over 1 Ghz. A simple detector on the tuner output produces a DC signal corresponding to the amplitude of the tuned frequency. An oscilloscope on the output triggered by the sawtooth makes an excellent spectrum analyzer. Murray markets this thing via his company, Science Workshop, PO Box 310, Bethpage, NY 11714, tel 516-731-7628. You can't go wrong for $65, but watch out! Bells and whistles quickly run up the bottom line. I have been enhancing my Poor Mans Analyzer over the years, mostly every time I go to the Dayton Hamfest and stop by the Science Workshop booth. This years addition was a little different - a $24.95 book titled "Build your own Spectrum Analyzer" by Murray Barlowe. This book, besides a comprehensive source of documentation and schematics for the Poor Mans' analyzer and all its options, is a great collection of Poor Man's owners notes, photos and enhancements.
In the Summer 1991 I came across the front cover of Radio Electronics magazine (presently Electronics Now). Featured was a spectrum analyzer plug in card for a computer. It was the TV tuner trick all over again, however this time the sawtooth tuning voltage is generated by a computer driven D/A converter. The output is measured by an A/D converter and displayed on the computer screen. This all sounds better than it actually is. The computer is not fast enough to sweep the entire spectrum and measure the output of each point, displaying it in real time. It literally takes seconds to sweep and display the spectrum. Computer generated noise is prevalent throughout the spectrum. Is it real or it just that new SVGA adapter? Who knows? The unit I have is the one featured in the article, by DKD Instruments, 1406 Parkhurst, Simi Valley CA 93065; tel 805-581-5771. $400 for the kit and software. These folks are not very friendly to deal with. The components in their parts kit are of poor quality and if you happen to be missing any parts, as I was, they will tell you to pick them up locally. I don't know what fantasy world they live in, but my local Radio Shack typically carries a full line of batteries and always has some little kid playing on the floor with a toy firetruck. Speaker wire and audio cables they can also handle, but a TOKO .078 Uhenry slug tuned inductor? I can't recommend this Analyzer to anybody.
The other variant in recent years has been fully assembled versions of the TV tuner analyzer. A&A Electronics has one for around $600.
Industrial quality analyzers are also coming down in price. Tektroniks and HP have models available for under $5,000, B&K has the Model 2610 for $2895 and ITC offers models beginning at $1295.
What Should I do? Do you need to spend a lot for Spectrum Analysis? Yes, if you want to plug and play. No, if you don't mind some assembly time and are good at electronic projects with a minimum (or no) instructions. The Poor Man's Analyzer does almost everything the more expensive analyzers do. The important exceptions are calibrated displays, built in attenuators and compact single instrument package. I'm very happy with how useful my Poor Man's Analyzer is.