Assuming a 30 GHz bandwidth, what categories of ambient broadcast signals would one be _unable_ to accurately analyze on a good quality spectrum analyzer?
For example, I have heard frequency analysis of spread spectrum and FHSS is not possible with standard SA's, and that current models are designed not to display certain frequencies used by the government.
For experimental reasonsI would prefer not to have anything excluded. Are work-arounds for these "missing" capabilities?
Charles Lind
I have never heard of an analyzer that wouldn't display frequencies within its bandwidth (and I work with analyzers on "government" frequencies all the time). The ability to properly allow quantitative display and analysis of spread spectrum would be limited by the resolution bandwidth of an analyzer not be some mythical "missing" capabilities.
Ron
If the analyzer is a sweeping type and not an FFT type, there is a way in which it can miss some signals.
The analyzer has a local oscillator that is swept thereby sweeping the part of the spectrum being looked at. If your signal is also sweeping, and your luck is about normal, a goodly fraction of the time the generator will be at a frequency that the analyzer isn't looking. You can solve this by fiddling with the knobs. Shifting the rate the analyzer sweeps at and watching the screen will make these show up.
On many analyzers, the front end has only a limited dynamic range. This can make a strong signal mask a weak one.
A short pulse with a wide bandwidth looks the same as general noise. This can make it hard to find signals like UWB and pulse position modulated signals.
snipped-for-privacy@cybernet.com (Charles Lind) wrote in news: snipped-for-privacy@news.tpg.com.au:
I think that is a load of crap. I know the gov't itself uses off-the-shelf commercial SA's.
-- Jim Yanik jyanik at kua.net
You can almost certainly do certain types of analysis of those sorts of signals with a standard swept-frequency analyzer, but may have trouble looking at the signals in detail.
This is due to the fact that a traditional SA uses a swept LO, mixer, and narrow filter to scan a whole range of frequencies... in effect it's a sampled system. Since it only looks at one small slice of the spectrum-under-examination at any given moment, it will miss whatever happens outside of its filter position and bandwidth. For example, a FHSS signal might be bouncing around in ways which would always (or almost always) cause its carrier and sidebands to be where the SA wasn't looking at that instant. If you want good frequency resolution when you do see something, you have to use a narrow filter, and a relatively slow sweep.
The same problem could exist for any pulsed modulation, not just spread-spectrum.
A spectrum analyzer which digitizes a whole spectrum-slice and then uses FFT to do the analysis would not have this particular sampling problem, but might well have other limitations depending on how continuously it can capture and process and display data.
Certain other types of SS might show up just fine and be easy to analyze on a swept-frequency analyzer. For example, an OFDM signal carrying continuous intelligence would show up like gangbusters since all of its carriers would be "on" and modulated.
Haven't heard that. I suppose it's possible, but it seems rather futile. All you'd need to get around this would be an external oscillator and mixer... and that's just how my old Systron-Donner extends its response up to multi-GHz range (external mixers, to combine the incoming signal with harmonics of a reference oscillator that the SA provides).
I imagine that some commercial high-speed analyzers do have discontinous frequency coverage, but I'd expect that's a money-saving design issue (concentrate on the specific bands for which there are commercial customers) and not one intended to hide anything deliberately.
--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
I doubt the government blackout thing. But some spread-spectrum signals are indistinguishable from wideband noise, and could well be below the sa's noise floor.
John
Patent: 7,010,481 (Spread Spectrum) System and Method for Generating Ultra Wideband Pulses ;-)
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine Sometimes I even put it in the food
government blackout thing is for older cellular phones, on scanner receivers, not SAs. Its a moot point now that 845 mhz cellular is digital as of last year.
I have never seen a SA with gaps at 1.7, 7, or 8 ghz Etc
Steve
I've heard that said of scanners, but not SAs
Actually, the analog spectrum analyzer has a swept filter.
They make tools that are more suitable for actual demod of signals. They are basically spectrum analyzers with either IF output or perhaps some demod. I have an old Microtel 1200 that takes just about anything and puts it on a 30MHz IF for further demod. A more modern version, though still stone age is the MSR904A
My Microtel is so freakin' old that the receiver lock is done with a servo motor. For wireless use, you can get a service monitor. Damn expensive even used. In fact, damn expensive even to rent. I assume but don't know for sure that a cellular service monitor can er um well monitor. ;-) So your phone calls are not safe from Bill Gates.
There are no holes in test equipment, even though they can monitor things that the feds think should be private.
=A0and that current models are
=A0 =A0 =A0 =A0 =A0 =A0 =A0 AE6EO
rThe FFT by default gives you the whole spectrum. You can choose not to display the whole specturm. To just sample a range, you need a chirp- z.
To be really clear here, a spectrum analyzer doesn't have to have a source. A network analyzer does, but a spectrum analyzer can strictly just do spectral analysis.
I'm not entirely convinced the analog spectrum analyzer is totally blind to spread spectrum, or vastly inferior to FFT. Since you can set the sweep rate of the bandpass, given a slow enough sweep, some of the spread spectrum energy should make it through the bandpass filter. Normally the sweep rate is chosen based on the group delay of the bandpass, but there has to be a way to override the sweep rate to compensate for the delay in the DUT.
Lastly, getting back to the FFT analyzer, would it be possible to set up a waterfall display and see the spread spectrum energy move around. Perhaps more likely to be seen with a frequency hopper than direct sequence.
Slipped a digit somewhere?
John
Oooops! Should be 7,010,056. Looks like my cut-and-paste editing was sloppy ;-)
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine Sometimes I even put it in the food
That would be a very unusual design. Making a narrow filter sweep is a lit harder than mixing to an IF frequency and then filtering there. I guess there may e advantages too.
n
=A0 =A0and that current models are
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 AE6EO
ior
Re the last paragraph, of course that's true of frequency-hopping spread spectrum, but not of spread spectrum done through modulation with a (pseudo)random signal. That is very difficult to distinguish from noise, and if the amplitude is low, can very easily be missed or mistaken for a source not carrying any useful information.
For a system capable of not only finding but identifying interesting signals from an antenna (for example), see
Cheers, Tom
Although a modern spectrum analysis system designed to find and classify signals from (typically) an antenna has no forbidden frequency ranges, and in general can see spread spectrum (though may not be able to demodulate it, of course), there are limitations. One key one is that receivers for narrowband communications systems can easily have noise figures well under one dB, and often are used in systems with directional and specifically polarized antennas, a spectrum analyzer used for broadband survey work will have an extremely difficult time maintaining a very low noise figure while also accomplishing all its other front-end tasks (e.g. very wide dynamic range), and of course with no a priori knowledge about the antenna system used in some communications system, may end up relatively blind to communications in that system. You can overcome the noise figure issue by using a narrow-band preamp for specific bands of interest.
The full specs in the data sheets of commercial signal analyzers and signal analysis systems should tell you exactly what performance to expect. At least that should be true of reputable manufacturers. It should be clear in the data sheet which specs are guaranteed and which are typical. Pay attention to the applications literature, too, to understand the limitations and capabilities. If you're serious about putting out money for an SA, sales reps should be willing to demo them for you and let you try them on your signals of interest.
Cheers, Tom
he
aion
=A0 =A0and that current models are
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 AE6EO
rrior
ll
!Direct sequence IS the use of psuedo random modulation,
I haven't seen that particular HP snooping box, but I saw a similar HP setup. [Clearly the NSA and governments buy those racks.] The receiver I saw had GPIB on it. It had a mess of varactor diodes for use in an agile front end narrow band filter.
d s
r,
the
er
e, a
ution
al
ce
=A0 =A0 =A0and that current models are
. rng
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 AE6EO
warrior
will
ds!
o t e . da.org/wiki/DSSS
Well, it's not HP, it's Agilent. They don't have anything you'd call a front end narrow band filter, though the N6830 that's mentioned has a set of very low distortion preselection filters for HF reception. For example, since the AM MW broadcast band is commonly not very interesting but has lots of large signals, there's a switchable filter specifically to kill it by at least 50dB. I suppose anything tuned with varactors would have way to much distortion to be useful in such a system.
Cheers, Tom
Tom,
Speaking of the differences between swept analyzers and those performing FFTs, do you happen to know of any publicly-available literature/application notes that discuss the various methods of splicing together multiple data sets collected prior to each FFT? It seems as though you would definitely want some overlap from data set to data set as soon as you apply a (non-rectangular) windowing function to the data.
---Joel
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.