Spectrum Analyzer Buyer’s Guide

Having a scope in a home lab used to be a real luxury, but these days, its fairly common for the home gamer to have a sophisticated storage scope (or two) hanging around. Dedicated spectrum analyzers are a bit less common, but they have also dropped in price while growing in capabilities. Want to buy your very own spectrum analyzer? [Kiss Analog] has a buyer’s guide for what to consider.

If you’ve already got a scope, it may have a Fast Fourier Transform (FFT) function, and he talks about how it could be used in place of a spectrum analyzer or vice versa. But it really depends on what you’re planning on using it for. If you’re doing compliance testing for emissions, an analyzer is invaluable. If you like building transmitters or even just oscillators for other purposes, viewing the output on a spectrum analyzer can show you how well or poorly your design is performing. Any application where you need to visualize large swaths of the RF spectrum is a candidate for a spectrum analyzer.

Towards the end of the video, you’ll get to see some actual uses on a Uni-T UTS3021B. While those are at the higher end of the hobby price spectrum (no pun intended), it has many features that would have required an instrument ten times that price in years gone by.

There are also some very inexpensive options out there. While it is true, to a degree, that you get what you pay for, it is also true that even these cheap options would be amazing to an engineer from the 1990s. Yes, of course. You could do it with a 555.

8 thoughts on “Spectrum Analyzer Buyer’s Guide

  1. A UTS3021B is £1700.

    I bought a secondhand R+S CMU-300 for £300.

    It comes with a spectrum analyser function that can even take up to 50W RF on the input. It goes up to 2GHz as I recall.
    Also comes with an RF sig gen. So you can then check attenuators at specific frequencies, etc.

    Plus (if you need it) you can also test 2G BTS’s!

    There must be all sorts of old telecoms equipment that has a spectrum analyser function.

  2. Unless we’re talking about high end – and I mean “high end” (ie. $100k+) Agilent equipment,
    I’d rather have a refurb’d “old school” Tek model.

    Got a Tek 494AP spec analyzer, refurb’d, calibrated, etc – the unit is absolutely
    superior in terms of measurment accuracy to any of the chinesium hobbyist toys.
    A 494AP went for more than a typical house “back-in-the-day”.
    May not have all the fancy bells and whistles, but all that is meaningless if it
    doesn’t produce accurate measurement results.

  3. The FFT of an oscilloscope CANNOT substitute for a spectrum analyzer! How many bits of dynamic range does an oscilloscope provide? Eight? Tek oscilloscopes could provide eleven for repeating waveforms (equivalent time sampling in Tek lingo). Each bit is 6dB and small change, right? So, even at twelve bits (some LeCroy lower-frequency oscilloscopes offer 12 bits), you are getting a 72dB dynamic range. And as the frequencies are pushed higher, the number of bits goes down. A spectrum analyzer can EASILY provide 90dB of dynamic range, if not more. So, this is the big difference between an oscilloscope running an FFT and spectrum analyzer: Dynamic range. Analog Devices makes some super-duper ADCs these days, but as the frequencies are pushed higher, even an older HP8565 40GHz spectrum analyzer will do better. (If the frequencies are so high that harmonic mixers are used, the noise floor rises and that eats into the dynamic range. But at such high frequencies, say beyond 100GHz, the effective number of bits of an oscilloscope will drop to a few bits only.) The problem with older analyzers like the HP8565 is that their IF bandwidth is not as wide as in a modern instrument. But the oscillators of the older analyzers like the HP8565 are YIG-based, while many modern instruments are all synthesizer-based. The jury on phase noise performance of YIG-based vs. all-synthesized oscillators is still out.

    1. On most 8-bit scopes FFT is a joke that mostly makes me think less of the scope, or at least should be just disregarded, but there are some exceptions. The FFT in Siglents is “usable”, (It get’s some 70 to 80dB out of an 8-bit ADC by using really deep (million points or so) FFT, and that is an impressive feature on itself. I’ve also seen a demo of an R&S scope receiving and FFT-ing commercial radio channels by just sticking a piece of wire in it’s input.

      And sure, a scope with FFT is no substitue for a “real” spectrum analyzer for the classical use of spectrum analysers, but when the feature is implemented solidly enough it can be a useful addition to an oscilloscope. It would also be complementary to a stand alone spectrum analyzer for low frequency stuff. But in the low frequency range, a good audio card, complemented with some front end and software is very likely the best options if you want to keep it affordable.

      1. On Rigol, downloading the 12+ Mpts and running FFT in Octave yields reasonable useful results for e.g. noise spectrum measurements, where the dynamic range is not that huge.

        But I agree that the built-in 1000 pts FFT on entry-level scopes is an useless joke.

  4. I did not realize that the site was run by pro-communists. Might as well be russian. They actually deleted my comment for encouraging people to buy something not made by communism (and child slave labor for that matter… We all know what’s going on in china). It would appear that the moderators of hackaday are likely russian sympathizers. Very sad.

  5. You really have to watch your back with used test equipment. There is a lot of abused and way out of calibration gear floating around. If the 50 Ohm input of an analyzer is blown, good luck getting an affordable replacement, if at all. Having said that, hobbyists might do better with a service monitor because a service monitor gives you a handy jack of all trades at RF generation, detection, demodulation and parameter measurement in one box. A few years back, I got a clean IFR 1200S for $600 US. The top end is “only” 1 GHz and it isn’t the most portable instrument, though it does have a battery. The controls are sensibly organized and it is easy to operate. The IFR is a nice tool for adjusting duplexer cavities, checking filters and setting CTCSS levels and transmitter deviation, etc. Even though it is old, the IFR gets to stay on my bench for now.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.