How to measure SNR??

[hasiehasie]

Hi guys

long time reader, only registered now.

I was reading up abit, as I want to improve my signal.

I have a 8ta stick 5 promo.

How does one measure the SNR?

I see people mentioning they use the same dongle as me (E367), and MDMA gives them SNR readings, however I cannot find a way to measure it. MDMA gives me the RSSI and CellID and a few other values but no SNR. Not too clued with this type of stuff.


I bought the 3g combo from uniterm direct, havent installed it, just played around with it.(after consideration I maybe must have gone for the omni directional, because I allready have a semi decent signal, would just like to improve the SNR or stability if i can). On a quick test the signal seemed to be the same, allthough the connection is better on speedtest without the antenna! Ive read a previous post on here with someone having the same situation with the external antenna giving slower speedtests.


I also have no idea where the closest 8ta tower is in Pta North so that is not helping.


So booking time this weekend for myself and on the roof (roof of a second story) and see whats the best combination I can get between Speedtest, the signal, and the SNR(if someone can advise me how to test it).

Any help will be greatly appreciated



Another thing, general consensus of deciding between directional or omni directional.
As i understand at this moment, if you have very bad signal, then a directional antenna will be better, but if you allready have a semi decent signal, then one should opt for a Omni directional? Is that correct?


Many th

 

[sajunky]

MDMA gives me the RSSI and CellID and a few other values but no SNR.

Use the latest mdma beta. Works well on E367.

As i understand at this moment, if you have very bad signal, then a directional antenna will be better, but if you allready have a semi decent signal, then one should opt for a Omni directional? Is that correct?

Not. Directional antenna is always better for connection stability. Omnidirectional should be used only if you have multiple towers in mind and want to switch between towers. If signal is too strong attenuator should be used.

- post your RSSI to get advice.
- check if your onmi antenna covers proper 3G frequencies.

 

[hasiehasie]

Use the latest mdma beta. Works well on E367.

Thanks, in my mind I had the latest one. Got it and can safely see SNR values

Not. Directional antenna is always better for connection stability. Omnidirectional should be used only if you have multiple towers in mind and want to switch between towers. If signal is too strong attenuator should be used.

So every differnt CellID is a different Tower (excluding the last digit showing which side?)

- post your RSSI to get advice.

My RSSI varies from -77 -> -83db in the room where the router will stand.. Without any external antenna. just the E367.(As i understand its not too bad?)

yesterday did a handheld test with the antenna, and the speedtests were slower as soon as i plugged it in. weird
Will do a better quick test tomorrow, now that I know where one of the towers are. Will do it with MDMA and let you know.

I took the laptop with mdma tonight and drove around until I found one of the towers, i connect to its CellID in the room where the tests were started (CellID 10161 -> 10166). The cellid is not the one that i see most of the time, but i see i have connected to it. I take it thats the same tower from different directions.(hte whole 1&3 = North, 2&5= SouthEast, 3&6=SouthWest)

I cant find the one which I usually connect to(CellID 12751->12756), as soon as i drive out of the yard, i lose it and get another cell id (the one i found 1016X).

I am on a second floor, but dont have straight line of vision of the tower, theres A house in the way, and maybe a block of flats if i guess correctly.

I did 10 tests of each, but im not gonna post results when the modem was in the router as today was very weird. My first 10 speedtests I did through the TP Link 3420 , I got a Max download of 8MB. And a Avg of 6MB. (Allthough if I do pingtest it still gives a D or F rating). Cant see the CellID because its though the router. The most I have ever gotten on this stick in any way was 7.2. And it went past it twice, so dont know why. Few minutes later it was much lower all the time.

With the E367 straight on the PC Immediatly after that I got between 3MB and 4.5MB, here are the averages after 10
tests with MDMA and speedtest.net:
RSSI : -77.2
SNR_A : -80.4
SNR_B : -7.8
Download : 3.5MB
Upload : 2.5MB
Ping : 37
Pingtest Rating : D


This was weird cause why was the router so much faster. I then plugged back the router and then the speeds were +\- PAR with the averages I listed.

- check if your onmi antenna covers proper 3G frequencies.

Mine is the directional 11db one from uniterm.


Does this info tell you anything?


Sorry, im such a noob with this.

Appreciate all the help
Thanks
Cheers

 

[sajunky]

My RSSI varies from -77 -> -83db in the room where the router will stand.. Without any external antenna. just the E367.(As i understand its not too bad?)

Not too bad, but it can be improved. However with no tower in direct view, positioning antenna can be tricky. Reason is that if you work on reflected beam result can stabilize after couple minutes or even hours. Remember that tower use beam forming to optimise your connection.

I trying to get a clue from your post, there is many things you raised without specific details, many goes wrong. One thing is clear that antenna doesn't improve speed, it is even worse. But you don't say whether antenna tunes to the same Cell_ID, as only then results are comparable. I asked also the question about antenna characteristic, but it seems Uniterm Direct doesn't publish detailed specs. I saw some antenna's marketted for 3G, but there are designed for different services and only adopted for 3G. Such equipment don't work well with 3G. I think you should drive to the place of direct view of the tower, compare results and if problem persist raise question with Uniterm Direct. There is also possibility that modem antenna connector is damaged. Look closely at damaged one I posted here: http://mybroadband.co.za/vb/showthr...maged-antenna-connector?p=8826241#post8826241

Your other issue is getting faster speed from the router. This is possibile, read about power issues and how to combat it.

My advice is, get problems eliminated one by one, not all at once, it would give you a nightmares.

Cheers

 

[hasiehasie]

Firstly, Many thanks for the help.

Not too bad, but it can be improved. However with no tower in direct view, positioning antenna can be tricky. Reason is that if you work on reflected beam result can stabilize after couple minutes or even hours. Remember that tower use beam forming to optimise your connection.

Remarkably 8ta just got back to me within 2 days, with the exact closest tower location. With a google map image. 800m away on a mountain and I think I might have a straight line of sight if im on the roof. Will check this weekend. I saw the towers but when driving around I did not see the tower nor pick it up on MDMA. And somebody told me they are there for something else. I suspect this might be the tower Im connecting to mostly.

I trying to get a clue from your post, there is many things you raised without specific details, many goes wrong. One thing is clear that antenna doesn't improve speed, it is even worse.

Yes, sorry I didnt make that very clear. The test I did with the antenna was just a few speed tests outside holding the antenna with my hand, just for a quick test. So the test was not really proper.So I will properly test it with the antenna this weekend. I was just mentioning immediatly after I hooked up the antenna I saw reduction in speed. Will do it with the laptop and through MDMA this weekend.

But you don't say whether antenna tunes to the same Cell_ID, as only then results are comparable.

I will do that this weekend when Im on the roof, then I will have better info to work on.

I think you should drive to the place of direct view of the tower, compare results and if problem persist raise question with Uniterm Direct.

Now that I know where the tower is I know where to point it. I know of two towers that are close. And I suspect the one on the mountain to be the one I usually connect to. Will report back.

There is also possibility that modem antenna connector is damaged. Look closely at damaged one I posted here:

Im using the cheap stick on with rubber band universal one from Uniterm Direct. The Poynting one looks to be much better quality.

On the E367, have you found the external aeriel connection with pigtail to be better than the inductive coupler?

Your other issue is getting faster speed from the router. This is possibile, read about power issues and how to combat it.

I was just mentioning it, I think it was just for that time. Later on speeds seemed quite the same.

My advice is, get problems eliminated one by one, not all at once, it would give you a nightmares.

The coupler from uniterm just looks like it might not work too well.....
Anybody have experience with their coupler?


Thanks

Cheers

 

[sajunky]

Firstly, Many thanks for the help.

You are welcome. Nice to see progress, you are on the right track.

Yes, sorry I didnt make that very clear. The test I did with the antenna was just a few speed tests outside holding the antenna with my hand, just for a quick test.

It is important to hold the antenna on the pole, like you see on various installations. Pole must be at least one meter long, do not touch active antenna elements, keep it away from any objects.

I don't have experience with Uniterm inductive coupler. Don't bother with, as your modem supports proper external antenna connector. You loose 6dBm on the coupler. Get not expensive pigtail from the WEB or Otto Marketting. If connector comes to tight, you should adjust tension with a knife or small flat scredriver to prevent damage to the connector like it happened to me.

Cheers.

 

[Tim the Techxpert]

Hi There,
Sounds like you are well on your way to getting a better connection. To just add my comments.
An omni directional antenna should enable you to connect to whichever tower is around whereas the uni directional aimed at your identified tower is not going to do so. This may be a problem if the tower you are sighted to is down.

I have heard very good comments about Poynting antennas so when you make your final decision I would consider them

Regards

Tim

 

[hasiehasie]

Many thanks

I allready bought this directional one, so will have to try make this one work :/

I put up a 3m pole this weekend just finished and it started raining, still thinking about how to earth it cause its on the roof on the 2nd storey.

Will get my pigtail tommorow and mount the antenna.

When this tower does go down, I will just plug out the antenna, and live with the maybe bit less stable connection.

As mentioned, I allready have kind of a good signal, just want to make it more stable. Hopefully the Directional Antenna will help with that.

Wont be too happy after the effort if its the same, or worse hehe...

But like junky said, will try everything one step at a time. And see where I end up.


many thanks

 

[hasiehasie]

got a pigtail today.

Only got home late and had little light left. Had about 30min on the roof. I will do proper tests tomorrow, I just had time to do a few tests as light caught me and the mosquito were killing me on the roof. I WILL DO PROPER TESTS TOMMOROW, this post is just for a little more insight concerning my SNR and how to see what is good and bad. Maybe to have more knowledge when I get on the roof tomorrow.

Was on the roof and had weird results.
The tower that 8ta advised me to point the antenna at(Lets call this Antenna A), I have line of sight. its 810m from me. When I point the antenna towards it, it improves the signal abit (from -75db without antenna to -69). Also stayed the same when turning the antenna in small increments looks like no matter what way in that direction it stayed on 69db.

Then as mentioned earlier in this thread the other night I drove around and found a antenna with MDMA (Lets call this Antenna B). It is about 1.5 - 2km away, i dont have line of sight. The are plenty of trees and buildings in the way. Just for a test I turned the antenna that (guessing)way. And the signal went down to -63db! (was wow! it doesnt get any better than this i thought)

So to summarize:
Without antenna : -75db
Antenna A -69db - Antenna 800m from home, clear line of sight, nothing obstructing.
Antenna B -63db. - Weird, have no line of sight. Many things in the way. Just guessing the average direction.


Then I started doing speedtests. And then I found that even with the antenna turned in any way, average results were still better without any antenna? How is this possible? On average the antenna speed tests (took about 10 each) were 1 -3MB Slower, and the ping was higher ms every time?

Is this then 8ta? According to me -63db is very good signal? Is the signal good just the speed from 8ta not good?

What should I be looking at when pointing the antenna? RSSI only or SNR as well?
I see there is a SNR_A and a SNR_B. Dont know the difference between the 2.

Must the SNR be higher or lower? What is a good SNR to work with?

Sorry for all the questions, just want better idea of what to look for when im on the roof tommorow.

many thanks
Appreciate all the help

When I align the antenna,

 

[hasiehasie]

Here is a better idea of the setup.

Excuse my artistic skills! hahha

EDIT :REsize PIC

 

[ginggs]

So to summarize:
Without antenna : -75db
Antenna A -69db - Antenna 800m from home, clear line of sight, nothing obstructing.
Antenna B -63db. - Weird, have no line of sight. Many things in the way. Just guessing the average direction.

Take note of the Cell IDs. Which base station do you attach to with no antenna?

Is this then 8ta? According to me -63db is very good signal?

-63dBm is a very good signal.

What should I be looking at when pointing the antenna? RSSI only or SNR as well?
I see there is a SNR_A and a SNR_B. Dont know the difference between the 2.

Show us the RSSI and the SNR. The SNR_A and SNR_B figures aren't documented by Huawei, so we aren't sure of the difference either.

Must the SNR be higher or lower? What is a good SNR to work with?

SNR is the signal-to-noise ratio. You want the strongest signal with the least amount of noise, which would give a high SNR.
I'm not sure how Huawei are reporting that value though, so let's see your figures.

Here is a better idea of the setup.

Excuse my artistic skills! hahha

Nothing wrong with your artistic skills , just make the pic a little smaller please.

Edit: See this thread:
http://mybroadband.co.za/vb/showthr...l-antenna-does-not-improve-but-degrades-speed
in particular post #9 by Ununoctium on SNR values and the discussion which follows

 

[hasiehasie]

Thank you very much for the info.

I read that thread quickly. Very informative thank you

As I understand then, for a quick reference for myself. i can look at the RSSI for a quick good position, and then adjust further looking at SNR_B. Should give me a better indication.

I dont care about getting mega speeds. Would just like the connection to be abit more stable. Just saying that if the two dont go hand in hand.

I will take a few logs tonight will CellIDs ect before I post again, so I can see exactly what happens. Yesterday was just a quick test(you know one always wants to play with the new stuff hehe).

When you say the SNR should be higher. I Assume:
SNR = -2
is better than:
SNR = -10?

Just making sure, confusing what people mean when the neg(-) comes in, technically its so, but maybe it comes out different when people mention it.

Many thanks

will report back what I find with IDs and relevant info

 

[sajunky]

As I understand then, for a quick reference for myself. i can look at the RSSI for a quick good position, and then adjust further looking at SNR_B. Should give me a better indication.

Yes, assuming it picks up the same Cell ID. As ginggs pointed out, comparing also Cell ID the modem picks up using internal antenna would give some clue.
I think if 8ta recommends tower, they wouldn't point to the tower with known issues. Tower B gives you bigger RSSI, but is not in direct sight and we don't know what maximum power is allowed to transmit and associated coverage area (and number of active users). Therefore I would focus on the tower recommended by 8ta, unless quick tests shows otherwise.

Yes, less negative is better:

SNR = -2
is better than:
SNR = -10?

 

[hasiehasie]

Hi guys

Ok played with it abit yesterday. The tower 8ta advised me to point to was by worst of them all.
I found a different tower, and this is the one the stick connects to without any antenna. Just for reference lets call this Tower C. I dont know where it is , just the direction. All I know is that the CellID is the same as the one that the stick connects to on the roof.

Here is a Summary of the results:
I wanted to just give averages, but I dont know which minimums and maximums will be usefull.
So this post will be very long. please advise me if I must rather just give a summary, or if I must rather just give a Excel document.

EDIT***
Its much too long.
Is it possible for me to maybe compile a complete excel document and mail it to you?

Would really help me out.
Or will Minimum and Maximum figures be fine?

many thanks for the help, appreciate it!

 

[sajunky]

Typical lab method is to reject any extraordinary samples which do not match typical response (if present), then calculate average. I'd like to see summary with your comments, please.

 

[ginggs]

So this post will be very long. please advise me if I must rather just give a summary, or if I must rather just give a Excel document.

Maybe make it public in Google Docs?

 

[hasiehasie]

Sorry for the delay.

I had to go away for work urgently. Will update as soon as I can.

Many thanks