Noise question

[netrosas]

I.m new using Star-Os and I.d like to know if the following measures are apropriated for distance of 2 miles.
Quality:46/92
Signal:-50dBm
Noise:-95dBm
(Infra)

thank.s

[tony]

Those readings are very good, I would say you have an excellent link!

[netrosas]

thank.s for your answer, but, isn.t -95dBm a very bad number for noise??

[tony]

No, quite the opposite. -102 represents no noise, and -35 represents close to 100% signal noise.

Signal on the other hand is the opposite, where as -102 is no signal, and -35 is perfect signal. Quality on Lucent cards will rarely go over 50 even at very close range.

Your link stats should be the envy of many people. :)

[go.fast]

So is -82 signal and 10 quality is not very good.. ?
George

[tony]

So is -82 signal and 10 quality is not very good.. ?
George

-82 signal means the signal is quite low. 10 quality is also low. You want the signal to be at least -69 or better, and the quality (SNR) of at least 20, which is quite good.

[go.fast]

so it looks like on one of my sectors 1/3 are in the 60's most the rest are in the 70's with a few in the 80's.
Is this what most people see? Or is it possible to get all in the 60's?
George

[lonnie]

It is tough to get all 60's, the low 70's are fine. The 80's are not good. You have an acceptable signal but not for 11 mbit. The cards will most likely be running at 5.5 or 2 mbit. They will be solid, but are slowing your on air speed.

[bairdc]

I'm curious about exactly what everyone is using to get those kinds of signal levels, and at what distances you're getting them. Is everyone using amps? For the last three years, I've been using amped systems. However, I've recently been trying to move away from amplifiers because of the problems they create. Specifically, I've had trouble with the amps picking up and amplifying noise from other sources. So, while I've eliminated most of my noise issues by not amping (and using horizontal polarity), I've ended up with low signal levels. Customers within about a mile of the tower seem to be fine with signal levels in the -70s and noise around -97. However, I just put a customer on that is about 2 miles away. His signal averages around -84, and noise is around -97 using an Orinoco card and a Pacific Wireless 21dB grid.

Is there anyone running un-amped systems? What kind of signal levels are you getting at what distances? Maybe I need to use higher power cards than the Orinocos at those distances...

Craig

[TheBarron]

Well.. I'm still beginning testing, but I'm going to do my best to avoid amps.. which will prolly be nigh impossible for my shots. At any rate, I'm using Lucent cards (Ruby) and YDI 24db grids (was gonna order PacWireless, but anyway). I'm also hoping to make some Primestar antennas in the hopes of getting in the range of 30dB gain out of them...

Trying to get the most power..err.. signal with the least money :)

[bobbyc]

We try to get the customers outputted signal at ~-69 on our amped towers, because our amped towers average -85 noise, so tha gives us 16dB SNR partner reading, while their station hears us around 30dB.
On our unamped microcells, we'll shoot for -79, since the noise floor on our unamped towers are around -95~-99.
We don't like to amp the customer end unless they are far far away. We have one customer who doesn't want to pay for an amp who is 13 miles away from our amped tower who has SNR's of 17/9 on average. He gets about a half megabit (we throttle at 768kb). We put a orinoco silver in a WET11 (it's legal, it's the pcmcia card in the wet11 that has the FCC ID, not the WET11 board) with a YDI 24dB grid and mounted the WET11 at the antenna so he didn't have any cable loss. Then we changed the max distance on the hilltop to 13 miles. The customer doesn't upload anything and just downloads so he is happy with his speeds.
Bob C

[TheBarron]

What sort of upload speeds can he get?

[Jester]

We have 3 60 degree sectors all three are currently amped and it has improved performance across the board. We are using the TT CPE units. We were having some issues with customers and didn't want to amp them individually. I haven't noticed any detriment to their traffic. Most customers are only using 256k - 512k so it hans't been a problem.

[bobbyc]

I think he gets 384kb upload or thereabouts. Still pretty good, just not near our advertised rates. But we told him that it would be slower before we agreed to install him.

[bairdc]

Hmm... I feel like I'm sort of stuck between a rock and a hard place. As I mentioned in my previous post, I'm trying avoid using amps wherever possible. Here's the reason. My biggest POP is on top of a 500 foot cliff overlooking town This is a 180* sector with a YDI 250 mw amp. There is a residential neighborhood right smack at the bottom of this cliff. Over the past couple of years, we've seen the noise problems get worse and worse. It started with 2.4 GHz cordless telephones in this neighborhood, causing our noise floor to go from -87 to -70 whenever the owner used the telephone. More recently, we noticed that our noise level had suddenly gone from -87 to a constant -79. So, I went searching for the source of the noise. Turns out that a little 95 year-old lady about 1/4 mile from the POP recently put in a 2.4 GHz video surveillance system (actually, her son put it in for her) because someone had been stealing firewood from her wood pile. We changed the channel on her surveillance system and cleared up the problem. Now, it looks like we're back in the same boat again! Someone must have gotten some kind of new 2.4 GHz toy for Christmas because ever since then, our noise floor has gone from -87 to -80. I'm getting tired of trying to resolve noise issues, so my solution has been to gradually move away from using amps and switch to horizontal polarity.

However, it looks like that's not the greatest idea either, because if I don't amp, I can only seem to get about 1.5 miles before my signal level is in the mid -80s. I want this network to be capable of 11 Mbps, so I don't want my signal levels that low.

Anybody got any magic that I can use to both avoid the noise and keep my signal levels high?

Craig

[georgew]

My biggest POP is on top of a 500 foot cliff overlooking town

Ah ha! You just discovered why being in a "really good" location is not so good.

Being at the top of a hill over a large population means you see every noisemaker. What is worse, if anyone else is transmitting from the hill,
you are dealing with your competitor's customers focusing their noise at a spot close to your tower.

If you go to 3 60* sectors, you will reduce the noise to 1/3rd on average by reducing your noise reception capability. The problem is you also use up all of the spectrum.

Another approach is to use the site as a relay only, by slowly moving your customers to smaller cells closer to them, you will reduce the customer's issues. Your subordinate cells can have more gain pointed
at the tower to get over the noise, and you only have to set up the cells for this rather than each customer.

Divide and conquer I always say. Lots of ways to divide. You can do both of the above methods, and take it to the maximum level possible...
at the expense of your competition. Too many competitors on a hill can shut the hill down, of course.

When I received my microwave training, I was taught to use the lowest point on the tower that got the connection established. You will notice that on many towers, the microwave dish's are not at the top. This is to reduce the distance to the horizon to just enough to get to the other end of the link and no further. This is part of the spectrum management. When we were doing licensed links, even the height was part of the license for this reason.


George

[bobbyc]

Ok before you completely give up, I would get a 200mw card from demarctech and mount it in a nema box on the tower right at the sector antenna.
See if your noise floor is in the -85~-90 range and if it can still emit and pick up enough signal from your distant customers.
Bob C

[bairdc]

Ah ha! You just discovered why being in a "really good" location is not so good.


Yup. What I wouldn't give to have this same location, only about a mile away from the nearest house. I'd be able to use an amp without picking up all the noisemakers...

Craig

[bairdc]

Ok before you completely give up, I would get a 200mw card from demarctech and mount it in a nema box on the tower right at the sector antenna.
See if your noise floor is in the -85~-90 range and if it can still emit and pick up enough signal from your distant customers.
Bob C

I've considered this, but I've been wondering about something. Isn't 200mw the transmit gain? What is the receive gain of these things? Without a boost in received signal, I've still got the same problem. Of course, if I boost my received signal, I also boost my received noise, right? Same problem as if I were using an amp.

Craig

[georgew]

Well, if you keep in mind that the amp adds noise, in theory the amps don't do a lot of good.....

On the other hand, I have seen amps improve receive sensitivity on most cards in low-noise environments.

You have an environment I haven't tested an amp in...


George

[dkii]

We put a orinoco silver in a WET11 (it's legal, it's the pcmcia card in the wet11 that has the FCC ID, not the WET11 board)

Do I understand you saying you replaced the card inside a wet11 with an orinoco silver card and it works? :shock: I'm confused as to how that would work without some hefty modifications to the firmware on the wet11.