We have an older Anritsu benchtop spectrum analyzer that I would like to set up to take ambient readings in the lab.
Can anyone recommend what additional hardware is needed to do this?
For example, the type of external amp and antenna to use to achieve the widestpractical response within the 300MHZ to 3GHz range.
Mark Harris
Could you disclose the model number? I want to lookup the sensitivity.
Of course, this Usenet. Anyone can recommend anything at any time.
Assuming you don't have any accuracy and requirements, a discone antenna should work: etc. Google for "wideband discone antenna".
You can also use "bow tie", "biconical", or "log periodic" antennas. For example:
Not enough detail to recommend a specific antenna.
The RF amplifier is a little tricky. You probably want better sensitivity, without creating internal mixes (intermod). See section on sensitivity starting at Pg 12. Adding lots of gain at the front end will increase the base line noise level while simultaneously decreasing the usable dynamic range (Fig 15). I would want to know the model number before offering a specific recommendation. If your unspecified model spectrum analyzer happens to have sufficient sensitivity, just the antenna might be sufficient. If you're testing an RF screen room or anechoic room, you'll probably need multiple RF amps or a tracking filter in order to obtain sufficient sensitivity to see weak signals.
Incidentally, you can easily tell if your front end is overloaded and producing spurious signals. Just increase the input attenuation by
-10dB. If the signal drops a corresponding 10dB, it a real signal. If it drops more then -10dB or disappears, it's an internally generated mix. If you want to see further down into the "grass" (noise), narrow the IF bandwidth, which will require a corresponding decrease in sweep time.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
Don't forget input protection. Somebody might stick their walkie-talkie in the focus of your antenna.
I hung a 12" banana test lead out the front of ours and got this:
which I partly blame on this:
-- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
Thank you for your reply. The model is: ANRITSU MS2661C - 9kHz-3GHz.
I should also mention the lab has a metal roof, so the antenna may need to be located outside.
Ideally, it should be non-directional. So that probably rules out log periodic. Discone sounds good. Have you seen any DIY plans to make these?
What level of sensitivity do you believe is required before an external RF amp is not required for ambient readings?
If the one above is not up to the task, it might be simpler to buy a new (used) analyzer with better input specs.
Mark Harris
First of all connect an external antenna to the analyzer directly. I would expect to see at least some FM BC and TV signals. Make a note of the signal levels of the strongest signals. Compare these to the maximum signal levels supported by the analyzer. From this you should be able to determine the headroom i.e. what is the maximum gain the preamplifier could have.
The preamplifier maximum output level (1 dB compression point) should be greater than the analyzer maximum input level. If no compression point data is available, a preamplifier with a high operating voltage (12 V or greater) and a big current consumption (over 50 mA) tend to have quite good compression points.
I would expect that in urban areas FM and TV stations would dominate, but a cellular phone in the room would of course dominate. If you are interested in some low level signals, you really should keep at least the predicable broadcast signals out of the preamplifier and analyzer to avoid overloading. This can be done by selecting the antenna frequency response or in the worst case using some notch filtering.
If the UHF TV signals are strong in your area, I would suggest using two separate antennas, one for frequencies 300-470 MHz and the other
800-3000 MHz, which also simplifies antenna bandwidth requirements. If you still are going to use a single antenna, at least make sure that the response is quite low at frequencies around 100 MHz (FM BC).It would also be quite helpful to install a switchable attenuator in front of the preamplifier. If you suspect that some of the spectral lines shown are intermodulation products and not real signals, adding a say 6 dB attenuator will attenuate the real signals by 6 dB, but the spurious will drop 18 dB and hence can be identified.
Thank you for your advice.
Are you saying it may be possible to get away without an external preamp in populated areas?
Please note, we are not interest in particular signals. We just want to monitor what is there at any given time, without regard to directionality.
Can these be purchased off-the-shelf in a non-directional (eg. not log periodic) configuration?
Is it acceptable to wire them in parallel to cover the entire range without switching?
Mark Harris
Only if you are interested only in local radio and TV broadcasts.
An analog TV receiver requires about 1 mV of signal, that is +60 dBuV or about -40 dBm.
Usable monophonic radio reception requires about 1 uV or 10 dBf or
-110 dBm. For best stereophonic reception, you require 45 dB more. Check your analyzer specifications and at what bandwidths they are specified.
Discones should be usable, the largest dimension determines the lowest frequency.
I do not know about 3000 MHz, but ordinary TV/satellite splitters go up to 2200 MHz, just use them in the reverse direction. Note that these are usually 75 ohm designs, not 50 ohms.
Around urban areas, you'll see a LOT around 2.5GHz, from Wifi, Bluetooth, etc. AND microwave ovens which splatter like crazy. For what it's worth, you'll find that the microwave ovens 'fire' at peak AC mains rates, so you'll see the predominantly 3, a little 6, phases floating around modulating all those 2.5GHz bursts. Plus, activity is extremely related to time of day - peaking at 11 to 12 for restaurants, 5, 6 to 7 as homes fire up uwave ovens. [I have 'recordings' of the 2.5 GHZ band taken at different times of day from various locations around the Bay Area, which would be useful as representative interference if anyone wants to simulate wireless link for developing/testing protocols. Interestingly, somebody had one powerful spread spectrum transmitter, not hopper, but true spread. Believe it was a 'thermal' furnace for drying located in the industrial area. Periodically PG&E, I think the origin, would blast around]
You should also see a lot of GSM(?) 900 to 1GHz cell phones. Again, super time of day related. If you're near a cell phone tower LIMIT your signal, else you can zap a SA. Or, reverse, a guard's walkie talkie, or an energetic cell phone puts out some power, too.
[at least to me] you haven't said what noise floor you're after. THAT will determine everything. From memory SA's have huge NF's, in the range of 34 to 36dB, which translates to NEED AN AMPLIFIER! or very sensitive antenna, which, of course, will be narrowband.So back to question
Exactly WHAT signal levels do you wish to monitor? Do you wish to look betwen/around FM, or AM areas, too.
Mark,
This question suggests you need a LOT of education in the RF/microwave area.
Don't take offense, but this suggests [at least to me] that once you obtain the information you won't be able to do anything meaningful with it, except maybe make some colored spectral display, like done with music to light displays.
You need to DEFINE with some numbers what you're trying to do. That way, you'll know when you're done.
How old is this instrument ? The data sheet was printed only 10 years ago, but for such instrument, the specifications are not quite spectacular.
If I am reading the specs correctly, the 1 dB compression point occurs already at -5 dBm, ie. 35 dB above usable analog TV reception and 1dB CP at -35 dB with preamplifier on, suggesting that the built in preamplifier gain is 30 dB (my guess).
With the stated 7 dB noise figure, there is not much point in using an external preamp instead of the internal preamp, since it would saturate the mixer anyway.
One should remember, that the current urban RF environment is a _hell_ and measuring it correctly is quite a problem.
Something that was acceptable measurement tool 10 years ago, might be badly outdated today.
While I would not be quite that harsh, but try to understand that with current technology, it is not possible to make a DC to daylight receiver/scanner/spectrum analyzer in a single unit, especially if both weak and strong signal processing is needed.
With a more realistic requirement, you will get more usable answers.
Data Sheet:
Does yours have the Option 08 20dB preamp installed?
Without the preamp, overload (1dB compression) at about -5dBm with dynamic range about 109dB at 1GHz. That should give a noise floor of about -114dBm which is quite good. Of course, that's at the narrowest IF bandwidth, which may require a really slow sweep time. Good enough.
With the Option 08 20dB preamp, the overload point is -35dB with a
-132dBm noise floor. The numbers make no sense, but that's what the data sheet says. In any case, you don't need that kind of sensitivity, so you can probably live without the preamp.
Hint: Numbers are always nice. How far away is the antenna going to be located from the spectrum analyzer. The problem is that the coax cable is going to be rather lossy over the distance. For example, if you're using 100ft of the very best LMR-400 coax, you'll see 7.5dB loss at 3000 MHz compared to about 2.2dB loss at 300 Mhz. This will skew any signal strength or power measurements across the frequency range. 7.5dB is quite a bit of loss and will prevent you from seeing signals near the noise floor without a preamp. However, even adding a preamp doesn't solve the skewed (also known as "tilt") in the frequency response. You may need some type of equalizer network after the preamp in order to get a flat response.
A discone is one of the easiest antennas to build. At 300-3000 MHz, you're NOT going to build something out of rods that looks like a porcupine. For those frequencies, use copper sheet (also known as copper rain gutter and roofing flashing). See the 2nd photo at: Yours will be larger. Machine an insulator. Bend the copper sheet into a 60 degree cone and solder to the coax shield. Add a top hat and you're done.
Ideas:
I have no idea due to too many variables. You may already have a 20 dB gain preamp in your spectrum analyzer (Option 08). The preamp really belongs at the antenna to compensate for the coax cable losses. I also don't know what you're looking for or even if you're measuring anything. Strong interfering signals, or weak signals near the noise floor. Strong signals like radio stations and cell towers won't require a preamp. Weak signals will. I think the best idea is to leave a connector near the antenna where you can later insert a preamp, and just see if it is usable without a preamp. If not, you can add one later.
You must like to spend money. If you disclose what you expect to see on the spectrum analyzer, or what you're trying to achieve, there may be a better instrument for the purpose. General purpose and broadband spectrum analyzers are not very SIGINT/ELINT receivers.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
Yes, this is a learning process. We had the SA just sitting around.
The first step was to simply see what it would display in the 300MHz to 3GHz range at our location, rather than try to isolate or analyize any particular signal from the mix.
My original question related only to this. The replies so far seem to be more usage oriented, as one might expect from experienced technicians.
Please allow me to rephrase.
Ideally, I would like to use a single non-directional antenna, perhaps one that we can fabricate ourselves from wire or light tubing. It does not need to be calibrated.
The other part of the question was if anything was needed on the front end to boost or rcondition the entire bandwidth of ambient signals mentioned above.
Once having this up and running, it would hopefuly be the incentive to put in some time after work and learn more.
Mark Harris
Two antennas are not necessary if a broadband antenna can be found to cover the 3.3 octave frequency range. A discone, log periodic, bow tie, and biconical antennas will all cover the frequency range.
I provided some overpriced links for such commercial antennas in my first reply. Discones and biconicals are so easy to build, I think you should try building first.
No. Parallel, as in a Tee connector, will not work. Officially, you need a splitter/combiner in order to isolate the antennas from each other. Finding a splitter/combiner to cover the 300-3000 MHz frequency range might be tricky. Digging: This looks ok: $75 However, even that won't work. There's still a problem of creating a null if both antennas hear a signal. If the phase is the same from a single source, through both antennas, the two signals will cancel when they hit the splitter/combiner. You can sorta fix the problem by physically separating the antennas, but that creates another problem, where each antenna only see's part of the coverage area. Best to use a single antenna.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
A receiver connected to a room temperature resistor will have -174 dBm/Hz and if I understood the data sheet correctly, the narrowest bandwidth was 1 kHz, thus the noise power would be -144 dBm.
With 7 dB noise figure, you are not that far from -132 dBm noise floor with preamp and 114 dB without preamp, so we are talking about nearly believable numbers.
If you have the optional preamp installed, you do not need additional preamps. In the best case, you might get 3-6 dB more sensitivity, but most likely your dynamic range would be ruined.
My original idea of using two separate antennas to avoid UHF TV signals from overloading the analyzer preamp front end. If this is not a problem, go ahead with the full spectrum.
I would use notch filter to reduce overload on specific frequencies. The complications of using multiple simulataneous antennas is not worth the exercise. There are also more potential sources of overpowering signals between 300-3000MHz, such as the new land mobile and cellular data frequencies, which will likely be located closer than a broadcast TV transmitter: HSPA+: 850, 900, 1900, 1700, 2100 MHz LTE: 700, 750, 850, 1700, 1800, 1900, 2100 MHZ
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
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.