Can you help me interpret this spectrum analysis noise plot?

It was my senior year in college. The minimum wage was $2.00/hr and I only worked 20hrs/wk (all they would give me during the school year). It's a good thing a glass of beer was only $.25. ;-)

They're gettin' antsy, though.

Hi Mark,

Here is a picture of the mount when I had installed it:

The radio itself is the classic Ubiquiti Rocket M2:

That radio is clipped directly onto the back of the dish:

One mistake I made was to use non-shielded cable, so, there's about 75 feet of the plenum stuff outside going from the roof to the first floor office where the router and POE lie.

From a practical (noise) standpoint, how much do you think that matters?

Hi, B4 you ask that question think about shield? Why do you think we use shielded cable in AF or RF cabling(wiring) CAT 7 spec. ethernet cable is best one. I believe your noise is radio related unlikely Ethernet cable related. Looks like your radio is repeater? All in all if you have reliable connection, what's the worry? One way of increasing signal strength will be using a bigger dish or did you picked best height of the antenna. The higher the better is a myth. Is it LOS install.?

Way back when I was working on microwave, troposcatter links biggest one was troposcatter link shooting from Nha Trang, S, Vietnam over to Thailand. Triple diversity, antenna was 120 ft. square parabolic pair. Nitrogen gas filled wave guides were used. No coax. SWR was around 1.2. One night VC rocket sharpnels made some holes in wave guides plumbing, SWR went upto about 1.4

Tri diversity is based on antenna spacing, frequency, polarization. RX was parametric cavity tuned preamp, and all tube based RX, TX was 10W TWT exciter to 10KW water cooled Klystron driving 100KW bigger Klystron also with water cooled. HV on them were around 27KV DC.

This all became obsolete soon when RCA launched first Commsat. That was end of my career in RF telecomm. going into Digital Data comm, main frames, etc. Worked on autovon/autodin, Arpanet(origin of Internet). Now I am no good any more, things changing too fast to grasp all.

I think the shielding shorts out EMI/RFI interference; but, if that interference is, say 60Hz, it won't be the cause of my noise.

My noise is at 2.4GHz.

So, I *think* the shielding isn't really shielding against *my* noise. But it could shield against 60Hz noise (which I'm not measuring).

That's what I think - but I don't really know the answer...

Hi Tony,

The radio is not set up as a repeater. It's set up as a router:

As you noted, I do have a reliable connection to the 2.4GHz WiFi access point about 3 miles away, but, since my service is unthrottled, the lower I can get noise, the faster I can get the speeds!

For example, here are side-by-side screenshots of the Rocket M2 speeds versus the Nanobridge M2 speeds the day I switched them out:

Notice, with *nothing* changed other than the dish & radio, on the smaller Nanobridge, I got 78Mbps but on the larger Rocket, I got 117Mbps.

I don't really want to go *bigger* than the Rocketdish, so, that's why I'm asking how to lower the noise.

Note: I see the noise on the Nanobridge was -97dBm, and that of the Rocket is -89dBm, so, it's interesting that the Rocket is noiser than the Nanobridge, considering the beamwidth of the Rocket is far narrower (but the gain is far higher).

I should explain that, while I can *fit* a bigger unit on my roof, the RocketDish is already 24dBi and the Rocket M2 radio is already

28dBm, so, the maximum (claimed) gain is 52 dBi (158 Watts), while the maximum legal gain is 36dBm (4 Watts), so, the Rocket is already throttling itself automatically just to keep within legal limits.

So I don't think a bigger dish is really the solution.

The solution is to figure out how to lower the noise...and, the RF Armor seems to be the best answer (albeit pricey).

Hi, What is operating power of the two radios and what is the sensitivity of two RX? They are built on same design specs? Is it real LOS install or radio horizon install?

Hi, If you rfeally hae to there iss no such thing as this the problem and this is solution. You need many hours "cut and try type approach". First off like using best cables you can get, even POE injectors are not created equal.

Hi Tony,

The access point is a bit more than 3 miles away, line of sight.

NANOBRIDGE M2: Measured signal strength = -68 dBm Measured noise floor = -97 dBm Measured speed = 78 Mbps / 52 Mbps (i.e., 2-channel MIMO) Claimed transmit power = 23 dBm (max) Claimed antenna gain = 18 dBi Calculated EIRP = 41 dBm (12 Watts) Claimed sensitivity = -93 dBm (max) SPEC:

ROCKET M2 + ROCKETDISH 2G-24: Measured signal strength = -64 dBm Measured noise floor = -89 dBm Measured speed = 117 Mbps / 117 Mbps (i.e., 2-channel MIMO) Claimed transmit power = 28 dBm (max) Claimed antenna gain = 24 dBi Calculated EIRP = 52 dBm (158 Watts) Claimed sensitivity = -95dBm (max) SPEC:

Hi Tony, With my cabling all less than 100 feet, it really doesn't matter what POE I use with my three radios.

My Bullet came with a half amp 15 volt POE, and my Nanobridge came with an amp 24 volt POE, and my Rocket came with an amp 24 volt POE with its own reset switch (which is really handy since the rocket is on the roof).

In practice, since my cabling is so short, the POE does not matter, as far as I can tell.

Something else *does* matter though ... :)

Nonsense, when you quit learning you're dead. You know how to learn which is a quality many people lack. ^_^

TDD

If they run the front end AGC using data from the back end ADC, I see two scenarios. One, if there is sufficient filtering on a per channel basis (not likely), then the out of band energy will not make it to the ADC, but the front end could be overloaded. Assuming crappy front end filtering, the strong signal is setting the front end gain, making it likely that the desired signal gets a reduced level.

I bet a WISP would benefit from a good front end low loss filter (excluding channel overlap), but it would no longer be frequency agile.

Yep. The front end usually has a ceramic bandpass filter that passes the entire 2.4 to 2.483 GHz band. Anything within the bandpass can overload (desensitize) the receiver front end. After the A/D converter, the digital filter and demodulator will only decode the desired channels (usually 3-4 channels for direct sequence spread spectrum) which eventually drives the AGC. If the interfering signal is outside of these channels, the AGC never kicks in.

I've seen a few block diagrams (which I can't seem to find again), that show a digital AGC detector immediately after the front end A/D converter. It was a simple digital level detector, where anything over some present digital level, would start AGC action. As such, any signal that makes it through the front end bandpass filter will produce AGC gain reduction. In other words, a strong wi-fi signal anywhere in the band will produce effects similar to desensitization without actually causing the front end amplifier to conduct or go non-linear.

I'm not sure which of these two methods is considered best practice.

Jeff, Do you think my noise, in the Santa Cruz mountains, is low?

I have no idea if that's high, low, or typical for your location. Actually, I'm surprised that Airview doesn't show more junk, noise, interference, and other stations at your altitude, so I guess it would be "low".

OK. Thanks. I don't have any idea what's a "good" Airview report, versus a noisy one. It *seems* low to me; but I wouldn't even know what low is.

These might help. Compare the Airview screens with yours. The middle graph (waveform view) is the one that shows other stations and interference best.