Hi,

I have just installed Netmetal AC over 12km link. Signal strength is -36 which I feel very strong signal and I'm having problem in upload speed and CCQ is also not stable. Can any one suggest how to change power in netmetal ac? At present it only except default.

I have used UBNT Rocket AC with NetMetal AC.
NV2.
CCQ - 70/60 not stable.

UDP test :- 150/100Mbps.
Ping:- under 15ms.
Data rate :- 585/390Mbp

Regards,

JP

just change mode to 802.11

Hi,

I have just installed Netmetal AC over 12km link. Signal strength is -36 which I feel very strong signal and I'm having problem in upload speed and CCQ is also not stable. Can any one suggest how to change power in netmetal ac? At present it only except default.

I have used UBNT Rocket AC with NetMetal AC.
NV2.
CCQ - 70/60 not stable.

UDP test :- 150/100Mbps.
Ping:- under 15ms.
Data rate :- 585/390Mbp

Regards,

JP

-36db is too much.
Low the power to get value from -45 to -55 db and also try to change frequency.

Can you share with me more info about your 802.11 config mode.

AP Bridge.
5GHz only AC.
Prot:- 802.11
Datarate: uncheck all supported and basic. VHT left default.
Under Nstream :- Enable polling and disable CSMA selected. Best fit 4000 frame limit.
Tx power:- all rate fix 17.

Remote:
Station bridge.
Rest of configuration same as above.

Rocket Ac + NetMetal = bad results
Use the same manufacturer.
In fact what is the problem? I see you only ask how to change Tx power on NetMetal.

The problem is I cannot change tx power. The only option it accept is default & all rate fixed. Both are not working well.

Try to set all rates fixed to 3db.

Can you share with me more info about your 802.11 config mode.

AP Bridge.
5GHz only AC.
Prot:- 802.11
Datarate: uncheck all supported and basic. VHT left default.
Under Nstream :- Enable polling and disable CSMA selected. Best fit 4000 frame limit.
Tx power:- all rate fix 17.

Remote:
Station bridge.
Rest of configuration same as above.

ap side :
801.11
20/40/80 Ceee
data rate VHT : support : 0-7 / 0-7 --- basic : 0-7
hw retries : 6
ch 0 & 1 : enable
ASMPDU : 4096
Distance : 15 km

Hi,

I have used UBNT Rocket AC with NetMetal AC.

Are you saying you have RocketAC on one side of the link and Netmetal on the other?

The problem is I cannot change tx power. The only option it accept is default & all rate fixed. Both are not working well.

All rates fixed is working fine with the AC gear, it just doesnt show in the current tx power...

Set it to 3dB as suggested by ibm

Hi,

I have used UBNT Rocket AC with NetMetal AC.

Are you saying you have RocketAC on one side of the link and Netmetal on the other?

i think he said :
CCR ______ NetMetal5+RocketAC Dish <........12KM ........>Rocket AC Dish+Netmetal __________ CCR

Hi,

I have used UBNT Rocket AC with NetMetal AC.

Are you saying you have RocketAC on one side of the link and Netmetal on the other?

i think he said :
CCR ______ NetMetal5+RocketAC Dish <........12KM ........>Rocket AC Dish+Netmetal __________ CCR

Ahh, Rocket AC dish I see now then its ok

Your signal is not the issue. Its the protocol you use;
If you use 802.11ac without ubnt's airmax or mikrotik's NV2 you are using plain 802.11ac that is still vulnerable for interferences from other 802.115Ghz frequencies. It degrades the link a lot.

Using airmax one one and and NV2 on the other is not working neither. Both protocols (tdma) are proprietary and are not compatible between the two.

So set this right and first see (show us) what your connection rates are and the signals on each chain.

I haven't seen a metal yet but if these ROS operated MT radios work like any other MT unit you can set it in winbox under
Wireless menu / wireless tables / interfaces / click on the radio interface (usually wlan1) and in the next 'Interface' you go to the "Tx Power" tab. (If this is not there to be seen, go first to the right hand buttons menu and chose 'Advanced Mode'. Now the "Tx Power" tab becomes available and by clicking on the down-arrow beside the "Tx Power Mode" field you pick 'default' if you want the default setting, or pick "card rates" and set any level you'd like to try.

You wouldn't set it higher than the radio default; i.e 30dBm or 33dBm (in case of the 2000mW version). Set it higher and you might end up burning the radio.
Set it lower until you'll find a signal strength on the other end of the link somewhere in the range of 35/40 to 45/50dBm
to get the highest s/n ratios. (And thus the highest connection rates = higher data throughput)

Some mentioned "3dB" ??? That makes no sense. If it was referred to the power setting, such a low setting would make the radio card very unreliable and probably the link to fail.

I suggested 3db first of all because if he use Rocket AC Dish, this has 31db gain and in many country the eirp limit is 30db and he can't get over it, second reason is that for 12Km it's unuseful to saturate the radio module...

Your signal is not the issue. Its the protocol you use;
If you use 802.11ac without ubnt's airmax or mikrotik's NV2 you are using plain 802.11ac that is still vulnerable for interferences from other 802.115Ghz frequencies. It degrades the link a lot.

Using airmax one one and and NV2 on the other is not working neither. Both protocols (tdma) are proprietary and are not compatible between the two.

So set this right and first see (show us) what your connection rates are and the signals on each chain.

I haven't seen a metal yet but if these ROS operated MT radios work like any other MT unit you can set it in winbox under
Wireless menu / wireless tables / interfaces / click on the radio interface (usually wlan1) and in the next 'Interface' you go to the "Tx Power" tab. (If this is not there to be seen, go first to the right hand buttons menu and chose 'Advanced Mode'. Now the "Tx Power" tab becomes available and by clicking on the down-arrow beside the "Tx Power Mode" field you pick 'default' if you want the default setting, or pick "card rates" and set any level you'd like to try.

You wouldn't set it higher than the radio default; i.e 30dBm or 33dBm (in case of the 2000mW version). Set it higher and you might end up burning the radio.
Set it lower until you'll find a signal strength on the other end of the link somewhere in the range of 35/40 to 45/50dBm
to get the highest s/n ratios. (And thus the highest connection rates = higher data throughput)

Some mentioned "3dB" ??? That makes no sense. If it was referred to the power setting, such a low setting would make the radio card very unreliable and probably the link to fail.

Well he was complaining the TX power settings doesnt work, this is the way to prove it actually works As stated by ibm, if hes in EU its way over limit. Still 12km with tx power 3dBm should give around -65 which is still a reliable link maybe not for 256QAM but reliable.

-65 is not a reliable link. Is at the limit for a good client connection. A reliable link is at ~ -50...-55

-65 is not a reliable link. Is at the limit for a good client connection. A reliable link is at ~ -50...-55

So if you have noise floor for example -105 (which is pretty noisy) then you have SNR 40dB. This is a reliable link for full modulation on 802.11n... maybe not for AC and 256QAM but dont say me that -65 is not ok. I have plenty of links with -70, they are not on full modulation but they are reliable and still passing 70Mbps without any problem on 20MHz channel.

-65 is not a reliable link. Is at the limit for a good client connection. A reliable link is at ~ -50...-55

So if you have noise floor for example -105 (which is pretty noisy) then you have SNR 40dB. This is a reliable link for full modulation on 802.11n... maybe not for AC and 256QAM but dont say me that -65 is not ok. I have plenty of links with -70, they are not on full modulation but they are reliable and still passing 70Mbps without any problem on 20MHz channel.

I only consider links with -45 to -55 to be possibly very stable. And noise floor given by ROS is only a small part of what is out there.
I've seen several occasions where noise floor on a channel is reported as being -107 or -115 (I rarely see better) but when I scan in the same frequency I'm being hit with -60dBm signals from other make radio's that are not mine...
The noise floor therefore should be now -60dBm isn't it?

I've also have links with -50~ -55 dBm signals and -119 dBm nf that still perform very poor. MCS 13 800ns is the best I can get where only mcs 11 or 12 shows me stable CCQ levels. This while the S/N is around 70dB!!
Try to pump data over that link and I'm happy to get some 40Mbps.

But I know, its full with 5Ghz using radios there. Every single 20Mhz is been used by me or the competition so the noise levels should be very high.

And what about the channel bandwidth? If a radio works in 20 or 20/40Mhz band its scan results and noise floor level is only is for that band width. There can be a 10 or 5Mhz radio hitting with -30dBm and still I see a noise floor of -110dBm! That can't be right.
The noise is there, it's only not 'seen' by OS.
(The radio physically still picks it.up..)

And what about none 802.11 signals in the band?

Opposite; My Astra Satellite dish receipt is ruined by my tower with about 6x 30dBm 802.11an radio's (on directional antennas). A sattelite technician spend 2 days in troubleshooting and replacing everything we could think of to make it work. Only when I suggested the only thing we didn't change was the dish location we moved it to the side of the house, in the 'signal shade' from my tower and everything works like a charm again..... But, theoretically the 5Ghz of my radio's should not al all interfere with the ??Ghz used by the satellite!!!... Well, it still does......

The bottom line is that in troubleshooting radio links due ill performances where all the hardware and maths should be ok means you still have to look in other sources of disturbances...
The only way to find if your prospected radio channel is really 'free' and clean should be done by the use of a proffesional spectral scanner.... Have you seen the price tickets? .... pffff

-65 is not a reliable link. Is at the limit for a good client connection. A reliable link is at ~ -50...-55

So if you have noise floor for example -105 (which is pretty noisy) then you have SNR 40dB. This is a reliable link for full modulation on 802.11n... maybe not for AC and 256QAM but dont say me that -65 is not ok. I have plenty of links with -70, they are not on full modulation but they are reliable and still passing 70Mbps without any problem on 20MHz channel.

I only consider links with -45 to -55 to be possibly very stable. And noise floor given by ROS is only a small part of what is out there.
I've seen several occasions where noise floor on a channel is reported as being -107 or -115 (I rarely see better) but when I scan in the same frequency I'm being hit with -60dBm signals from other make radio's that are not mine...
The noise floor therefore should be now -60dBm isn't it?

I've also have links with -50~ -55 dBm signals and -119 dBm nf that still perform very poor. MCS 13 800ns is the best I can get where only mcs 11 or 12 shows me stable CCQ levels. This while the S/N is around 70dB!!
Try to pump data over that link and I'm happy to get some 40Mbps.

But I know, its full with 5Ghz using radios there. Every single 20Mhz is been used by me or the competition so the noise levels should be very high.

And what about the channel bandwidth? If a radio works in 20 or 20/40Mhz band its scan results and noise floor level is only is for that band width. There can be a 10 or 5Mhz radio hitting with -30dBm and still I see a noise floor of -110dBm! That can't be right.
The noise is there, it's only not 'seen' by OS.
(The radio physically still picks it.up..)

And what about none 802.11 signals in the band?

Opposite; My Astra Satellite dish receipt is ruined by my tower with about 6x 30dBm 802.11an radio's (on directional antennas). A sattelite technician spend 2 days in troubleshooting and replacing everything we could think of to make it work. Only when I suggested the only thing we didn't change was the dish location we moved it to the side of the house, in the 'signal shade' from my tower and everything works like a charm again..... But, theoretically the 5Ghz of my radio's should not al all interfere with the ??Ghz used by the satellite!!!... Well, it still does......

The bottom line is that in troubleshooting radio links due ill performances where all the hardware and maths should be ok means you still have to look in other sources of disturbances...
The only way to find if your prospected radio channel is really 'free' and clean should be done by the use of a proffesional spectral scanner.... Have you seen the price tickets? .... pffff

If you have interference problem, no matter how high the SNR is, the link reliability will be poor... I personally set up N links in the -60 range if possible and they usually perform great. If they do I dont see any reason why to burn the radios on -40,-50. Only if in PTMP the clients are very close and the power cant be lowered, in these cases I sometimes tweak the noise floor manually to get to -60s... The first links I started to tune around -50 were AC links.

I agree that the noise floor readings could be sometimes wrong but I probably do not have so bad interference problem even we have some competition around.

No surprise that such low cost devices could affect 10-12GHz satellite signal... Are you using metal boxes for the RB and antennas with side shielding?

For the spectral scanner, did you see the latest Ubi AC airview? Dont you think it can display the similar data like the professional device? Their noise floor readings is way better too IMO.

If you have interference problem, no matter how high the SNR is, the link reliability will be poor... I personally set up N links in the -60 range if possible and they usually perform great. If they do I dont see any reason why to burn the radios on -40,-50. Only if in PTMP the clients are very close and the power cant be lowered, in these cases I sometimes tweak the noise floor manually to get to -60s... The first links I started to tune around -50 were AC links.

Well, I have some PTMP situations were clients are so close to AP their receive signal is -25 to -30. It still runs for years..... I don't think the radios burn that easy. There is also no option to lower signal from AP because it also still needs to serve distant clients that I try to keep around -60.

I agree that the noise floor readings could be sometimes wrong but I probably do not have so bad interference problem even we have some competition around.

Recently just found out that the competition aimed a sector with a 10Mhz bandwidth radio towards my working range of one of mine AP's with exactly the same centre frequency as my 20Mhz channel..... its only because all my 25+ stations are in the -55 to -65 range they still have good speeds. It also depends a lot on the actual situation of the CPE antenna (all SXT's). Recently had one that can't even visually 'see' that intruder's AP and only had a nlos to my AP. Although the signal read -62 to -65 the speed was crap. 2-3Mbps at best.... Had to relocate the SXT to a more distant ap, but on another frequency and everything is fine again.....

No surprise that such low cost devices could affect 10-12GHz satellite signal... Are you using metal boxes for the RB and antennas with side shielding?

Yep, all in metal boxes, all high gain directional antennas. Actually I suspect more the 'other end' of one of the links, that also had a high gain (26dB dish with side shield+27dBm radio over only 300 meters) dish aimed at this mast only 3-4 meter away from the sat dish. Obviously that remote antenna's signal would hit the dish as well. Even under a relative small angle. Maybe that more been the problem. By relocating the sat dish It is completely in the shade of the remote unit as well....

And I would not call Mikrotiks boards and radios in combination with Jirious dish antennas "cheap". They might not cost a lot compared to some high end antennas but I think they perform very well.
Just once bought myself a twice as much costing antenna but saw no real difference. But he ho, indeed I'd wish the SXT's would come with a bit more shielding from the housing. The ubnt dish CPE's look very good to me but to be honest, I yet have to see a test between similar gain microstrip antenna and dish CPe to see which of the two is really better in rejecting unwanted signals..... My impression is everybody thinks the one he'd chose for is the best. But its never really scientifically proven...

For the spectral scanner, did you see the latest Ubi AC airview? Dont you think it can display the similar data like the professional device? Their noise floor readings is way better too IMO.

No, I didn't Would it even pick up non 802.11 related radio energy? And still, I need a scanner on each AP location again and again. I don't want to setup an ubi antenna everywhere for just that....

In fact, my strategy is to setup many small cell AP sectors with as their remotely controlled power feed and possible backup system an OmniTik. They than can also be regularly used to do a site scan to see if the projected frequencies of the sector AP's are still 'free'. And its not only the competition that starts to interfere with my frequencies.... I've seen the first domestic wifi routers already hitting some of my antennas with -65dBm of signals.
And this is only part of the problem. Knowing and avoiding interferences at the AP is one thing..... but what about the client?
Although an AP is relatively free from interferences it doesn't mean that one of its clients is not been hammered by stronger interfering signals from just another AP. I've had an issue with one client is such respect were I could not get his radio to work stable. Even with signals in the -50 range his CCQ stayed low and connection rates were all over the place. We replaced his unit about 4 times for new ones. We swapped twice an SXT that worked somewhere else fine but not with him.... Finally I decided to scan in the 10Mhz band... I found the competition placed one AP in 10Mhz band only 10Mhz away from my center frequency of my 20Mhz band AP. and its was only some 50 meters away from this client! He was hit with -45dBm of signals from their AP were he would "only" 'see' my AP with -55..... I split the AP in two new higher gain sectors (SXT-CA) and set them to 10Mhz as well, in another freq. range. He has -40 from me now and the problems are gone... (I don't know about 'his'. )

I am very curious what speed capable of this equipment on 12 km.
My experience is that in point point backbone network needs to achieve at least -60dBm signal strenght.

Thank you guys for all wonderful reply.

Both the side i have used UBNT AC dish with Netmetal.

UDP test i'm getting 150/100Mb

Live backhual traffic it can pass 80/40Mb on 12km link.

I cannot get more upload.

I have installed another 20km link with RocketDish30dBi with Netmetal. See below screenshot.

Your CCQ is not someting good, it is should be 90-100%. Have you tried other channels ?

Try only-N with 20 mhz channel and you will get nearly same throughput

Gesendet von meinem HTC Flyer P510e mit Tapatalk 2

SET 802.11ac (plain protocol)


nor NV2 nor nstreme works with 802.11ac

SET 802.11ac (plain protocol)


nor NV2 nor nstreme works with 802.11ac

NV2 works fine on my AC links... so this is not true, 802.11 has too high latency under load

I would definitely try 20MHz as finding clear spectrum for 80MHz is almost impossible. If 20MHz works fine, try increasing to 40MHz and see.

Hi
we have same problem with netmetal and 30Dbi dual chain Dish,
our config
AP :
5GHZ ac only
20/40/80 Ceee
data rate VHT : support : 0-7 / 0-7 --- basic : 0-7
hw retries : 7
ch 0 & 1 : enable

Distance : 10 km

signal : 55/56 tx/rx
CCQ: 20/30

bandwidth test udp:
maximum 20mbps

what is the problem????

Hi
we have same problem with netmetal and 30Dbi dual chain Dish,
our config
AP :
5GHZ ac only
20/40/80 Ceee
data rate VHT : support : 0-7 / 0-7 --- basic : 0-7
hw retries : 7
ch 0 & 1 : enable

Distance : 10 km

signal : 55/56 tx/rx
CCQ: 20/30

bandwidth test udp:
maximum 20mbps

what is the problem????

Interference?

Hi
we have same problem with netmetal and 30Dbi dual chain Dish,
our config
AP :
5GHZ ac only
20/40/80 Ceee
data rate VHT : support : 0-7 / 0-7 --- basic : 0-7
hw retries : 7
ch 0 & 1 : enable

Distance : 10 km

signal : 55/56 tx/rx
CCQ: 20/30

bandwidth test udp:
maximum 20mbps

what is the problem????

Interference?

yes, probably

our old link here(same direction, same height) with 433ah+28Dbi flat antenna work fine.
bandwidth: 40mbps;

our old link here(same direction, same height) with 433ah+28Dbi flat antenna work fine.
bandwidth: 40mbps;

Well, here's your answer then... Now you have two links running side by side, from same towers? They interfere with each other...
Do a proper site survey (both end) in regard of the spectrum to use. And make sure antenna's are not too close to eachother. And to test one link, disable the other otherwise one end just penetrates with his signal into the other link's other end.....
Making good working radio link is all about tuning, fine tuning and more tuning.
Everybody can make a link in the Sahara Desert.

our old link here(same direction, same height) with 433ah+28Dbi flat antenna work fine.
bandwidth: 40mbps;

Well, here's your answer then... Now you have two links running side by side, from same towers? They interfere with each other...
Do a proper site survey (both end) in regard of the spectrum to use. And make sure antenna's are not too close to eachother. And to test one link, disable the other otherwise one end just penetrates with his signal into the other link's other end.....
Making good working radio link is all about tuning, fine tuning and more tuning.
Everybody can make a link in the Sahara Desert.

excuse me for bad information
we replaced old link by new netmetal+32DBi dish. with no difference.

You should stop using 80Mhz channels.

You should stop using 80Mhz channels.

tanq for all replay.

after 2 days we now have about 40/80 mbps TX/RX bandwidth.
it's enough but in reverse direction, if we have 80/40 mbps it's good for now.