Yes and yes. They are neat boxes, especially if you can find one for less than $1000-$1500. I wouldn't pay much more than that for one, though. Agi lent was careful not to allow them to compete with the "real" E4400 SAs. T he maximum span width is limited to a few MHz (10?) and there are a lot of blank menu entries and nonfunctional buttons.
On the upside, they have a lot of digital RBW settings, going down to 0.1 H z. I use mine when I want to look at port isolation or interconnect leakag e down to -120 dB and below. That's about the only time I turn it on. I n ormally prefer my old 8566 and 8568 boat anchors because of the operational limitations, subpar user interface (IMHO) and long boot time imposed by th e E4406A.
A few of my quick-and-dirty GPIB tools
m) can be used with it, including the phase noise and spectrum recording ap ps.
Looking at the datasheet, this isn't that much better than a spectrum analyzer. If it was a VNA, I'd be a lot more interested. But it does have a power meter built into it.
The noise spec is a bit high, but they spec it at a 1kHz BW. Most spec at the 100Hz BW.
The Agilent box can input IQ. I presume it can display a constellation, but I haven't read the manual.
A spectrum analyzer can be just that, i.e. a spectrum analyzer, or it could incorporate a power meter. Better yet, it could have a tracking generator to sweep filters.
If you want to test antennas, you need a dedicated device like an Anritsu Site Master, or a network analyzer with impeance bridge. I suppose if you want to characterize any port, you would need a VNA. Products like a Site Master are very noisy, since they are intended to be used for cable testing, impedance sweeps, etc., i.e. using the built in source. A VNA is generally as low in noise as a spectrum analyzer, but more expensive.
It might make more sense to indicate what you want to accomplish.
As John said, its first limitation as compared to a classical spectrum analyzer is its maximum span : you can analyze no more than 10MHz of spectrum at a time. The second disavantage is that it is a quite complex equipment, therefore you will need to spend a little more time to use it efficiently (and if not you could get wrong measurements, for example with pulsed signals, as it is FFT-based and not sweeping). Lastly it covers from
7 to 314MHz and from 329MHz to 4GHz, so check that you have no interest in the intermediate small band...
Now the advantages : as a vectorial signal analyzer such a product can do things that a classical SA can't. In particular it can demodulate signals and display I/Q signals, constellations, etc. However for anything else than basic applications you would need to have the associated software options installed on your 4406A : GSM/EDGE/CDMA analysis, etc.
Have fun with it, nice equipment and available at reasonnable costs as plenty of telecom manufacturer bought it to test their transmitters...
In many cases this gets down to what the instrument class does. A VNA will measure two port parameters of most any arbitrary DUT, over a defined frequency range in fixed steps (linear and maybe logarithmic). An SA will measure the general amplitude of (quasi-) steady signals over a defined range with a defined sweep speed and receiver bandwidth. The sweep speed and RBW interact.
That said what things do you wish your instrument to do?
I think people may be confusing max instantaneous bandwidth with sweeps. You need to check if it sweeps. I'd be surprised if it didn't but I suppose it is possible (older VSAs did not include SA operations, but they usually do these days).
For example, look at the image under NI 5665 14 GHz VSA Analysis Time, Preselector Enabled
That instrument has a 50 MHz max instantaneous bandwidth, but there you see a sweep of essentially 14 GHz.
Spectrum/signal analyzers have always had limited instantaneous bandwidth, of course. If the displayed spectrum is beyond the max instantaneous bandwidth, then it is swept, whether the sweep is analog or stitched digital.
A late reply, but the E4406A doesn't have the (downconverter) frequency pla n of a conventional spectrum analyser. It downconverts to a first IF of 321 .4MHz with no preselection so there is no image rejection and also a hole i n its tuning range at 321.4MHz. It will also suffer from other significant spurious terms. eg subharmonics of 321.4MHz will have limited rejection. Yo u are supposed to use it for vector signal analysis under controlled condit ions rather than as a spectrum analyser.
lan of a conventional spectrum analyser. It downconverts to a first IF of 3
21.4MHz with no preselection so there is no image rejection and also a hole in its tuning range at 321.4MHz. It will also suffer from other significan t spurious terms. eg subharmonics of 321.4MHz will have limited rejection. You are supposed to use it for vector signal analysis under controlled cond itions rather than as a spectrum analyser.
That is the case with the NI 5663 too, if I remember the model number corre ctly. It is also true of some Aeroflex models. However, I will point out that some old SA's---that did not have "signal analysis" in any sense that we mean it now---also did not have image rejection. I think I used to test with a "Cushman" that had the problem, but my memory is suspect these days . (It was at a place I did not work at for very long, in the '80's.)
The 5663 is really intended for signal analysis in a "controlled environmen t" only. I mean, one can do spectrum analysis with it, but has to be aware of the limitations.