wAP 60G experience

Insider · Mon Dec 18, 2017 4:12 pm

Hello, what are test results of this equipment?

Our experience is quite surprising, keep on mind this is WiFi Renesans..

There are very few technical information in menu. No Signal strength information, just quality and MCS modulation method. On RF spectral analyser we saw something like 2GHz channel width, but we need to get pre amplifier to see more. Estimated POut ERP is 100 mW, the gain of antenna is unknown, but we estimated something from 13-20 dBi. Anyway with this we are still linking on 200m, tests will follow! Premature Merry Christmas!!

mag2020 · Mon Dec 18, 2017 10:13 pm

Hello, what are test results of this equipment?

Our experience is quite surprising, keep on mind this is WiFi Renesans..

There are very few technical information in menu. No Signal strength information, just quality and MCS modulation method. On RF spectral analyser we saw something like 2GHz channel width, but we need to get pre amplifier to see more. Estimated POut ERP is 100 mW, the gain of antenna is unknown, but we estimated something from 13-20 dBi. Anyway with this we are still linking on 200m, tests will follow! Premature Merry Christmas!!

What results are you getting?
Still expecting some kits I ordered for a test.
But in all, I expect a good result for backhauling of WiFi traffic around clusters of buildings.
A 200m LOS/nLOS distance is good enough in my use case.

Insider · Mon Dec 18, 2017 10:22 pm

200m is over the physical limit quality 30-40 points, which is almost no L2 link.. capacity is below 30mbit

mag2020 · Mon Dec 18, 2017 10:50 pm

200m is over the physical limit quality 30-40 points, which is almost no L2 link.. capacity is below 30mbit

Did you say below 30Mbps? At what distance do you get the full rated capacity from your test?

MulderSk · Tue Dec 19, 2017 12:13 am

320m signal 80 MCS 8 full speed but one side wAP is inserted into LHG XL

lazna · Tue Dec 19, 2017 2:50 am

Anybody know if signal level and/or MCS could be readed via SNMP? RouterOS 6.40.5

thankx, L.

blingblouw · Tue Dec 19, 2017 8:36 am

320m signal 80 MCS 8 full speed but one side wAP is inserted into LHG XL

More info! How exactly did you do this?

Insider · Tue Dec 19, 2017 11:39 am

200m is over the physical limit quality 30-40 points, which is almost no L2 link.. capacity is below 30mbit
Did you say below 30Mbps? At what distance do you get the full rated capacity from your test?

Thats not the goal at the moment, we need to find sensitivity edge level for at least MCS4 with that patch antenna, which we believe is somewhere about 140-150m, than we will move forward to concentrate signal either to horn or to waveguide.
We are now on edge of MCS0 which is BPSK, the QPSK is + 3dB, the 16QAM is + 9dB, this is overall performance windows from zero to hero..

We are now exactly 230 m far, so the 100m distance should be +6dB and more which means MCS8 = almost full speed

Other part of team is now checking up params of radio part, at least oscillator can do quite better = better sensitivity.

Insider · Tue Dec 19, 2017 11:44 am

Hello, what are test results of this equipment?

Our experience is quite surprising, keep on mind this is WiFi Renesans..

There are very few technical information in menu. No Signal strength information, just quality and MCS modulation method. On RF spectral analyser we saw something like 2GHz channel width, but we need to get pre amplifier to see more. Estimated POut ERP is 100 mW, the gain of antenna is unknown, but we estimated something from 13-20 dBi. Anyway with this we are still linking on 200m, tests will follow! Premature Merry Christmas!!
What results are you getting?
Still expecting some kits I ordered for a test.
But in all, I expect a good result for backhauling of WiFi traffic around clusters of buildings.
A 200m LOS/nLOS distance is good enough in my use case.

200m is behind the edge for using. We are expecting range of 140-160m is the max with wAP, as the gain of antenna and estimated 20dBm ERP. THis makes sense, that with antenna 1 feet you can gain next 40dB ! of gain with correct EIRP parameter, we will know more soon

Insider · Tue Dec 19, 2017 11:47 am

320m signal 80 MCS 8 full speed but one side wAP is inserted into LHG XL

It is blow. with such apperture and quality antenna you will get complete MCS12 on 1,5km

mag2020 · Wed Dec 20, 2017 10:54 am

Hello, what are test results of this equipment?

Our experience is quite surprising, keep on mind this is WiFi Renesans..

There are very few technical information in menu. No Signal strength information, just quality and MCS modulation method. On RF spectral analyser we saw something like 2GHz channel width, but we need to get pre amplifier to see more. Estimated POut ERP is 100 mW, the gain of antenna is unknown, but we estimated something from 13-20 dBi. Anyway with this we are still linking on 200m, tests will follow! Premature Merry Christmas!!
What results are you getting?
Still expecting some kits I ordered for a test.
But in all, I expect a good result for backhauling of WiFi traffic around clusters of buildings.
A 200m LOS/nLOS distance is good enough in my use case.
200m is behind the edge for using. We are expecting range of 140-160m is the max with wAP, as the gain of antenna and estimated 20dBm ERP. THis makes sense, that with antenna 1 feet you can gain next 40dB ! of gain with correct EIRP parameter, we will know more soon

Thanks for sharing. Sure the antenna size should be small at 60GHz. And that gives the size advantage but with higher propagation losses I doubt if your signal can go far.

vlatko · Wed Dec 20, 2017 11:18 am

im already writing about this in viewtopic.php?f=7&t=128251
obviously its possible to mount wap60 board inside lhg housing, i think its possible to mount in ldf housing also,
only question is how mehanical fix 60ghz antenna on top of the board before replacing housing.
further mounting ldf on a larger parabola is simple task.
dont forget big problem about aiming on 60ghz.
big question for mikrotik is why we dont have more data about 60ghz board and why we dont have more control on that boards?
there are some logic for wap60 regarding his simplified role in network.
but, how do they think we can use and setup lhg60 as PtP without detailed information about the signal,rates?
how to examine and find problem when it arises on 60ghz links if we do not have any information?

Insider · Wed Dec 20, 2017 11:25 am

Hello, what are test results of this equipment?

Our experience is quite surprising, keep on mind this is WiFi Renesans..

There are very few technical information in menu. No Signal strength information, just quality and MCS modulation method. On RF spectral analyser we saw something like 2GHz channel width, but we need to get pre amplifier to see more. Estimated POut ERP is 100 mW, the gain of antenna is unknown, but we estimated something from 13-20 dBi. Anyway with this we are still linking on 200m, tests will follow! Premature Merry Christmas!!

What results are you getting?
Still expecting some kits I ordered for a test.
But in all, I expect a good result for backhauling of WiFi traffic around clusters of buildings.
A 200m LOS/nLOS distance is good enough in my use case.
200m is behind the edge for using. We are expecting range of 140-160m is the max with wAP, as the gain of antenna and estimated 20dBm ERP. THis makes sense, that with antenna 1 feet you can gain next 40dB ! of gain with correct EIRP parameter, we will know more soon
Thanks for sharing. Sure the antenna size should be small at 60GHz. And that gives the size advantage but with higher propagation losses I doubt if your signal can go far.

All the factors are just great. Huge propagation loss (12dB/km) are keen to dense repeating of band, the narrow beams will be ok up to 1 km. Right now we made several tweaks, making 500m with MCS 8.

Insider · Wed Dec 20, 2017 11:43 am

im already writing about this in viewtopic.php?f=7&t=128251
obviously its possible to mount wap60 board inside lhg housing, i think its possible to mount in ldf housing also,
only question is how mehanical fix 60ghz antenna on top of the board before replacing housing.
further mounting ldf on a larger parabola is simple task.
dont forget big problem about aiming on 60ghz.
big question for mikrotik is why we dont have more data about 60ghz board and why we dont have more control on that boards?
there are some logic for wap60 regarding his simplified role in network.
but, how do they think we can use and setup lhg60 as PtP without detailed information about the signal,rates?
how to examine and find problem when it arises on 60ghz links if we do not have any information?

Hi, totally agree there are very few user interface information fir those who use it in professional basis, but enough for dummies. I believe MikroTik will release more tweaks for this band, it is not necessary to mess with band this way. Channel width is about 2 GHz, which wastes band, and capacity on 16QAM is 4,5GBPS. This needs to be chopped to subbbands like 225MHz, 500MHz or 1GHz channels to make better location planning which will be enough to carry 500Mbit in 16QAM, or 1Gbit in 256QAM etc.. We see big underdevelopment in radio part and antenna. There are also 4 nests for modules, which will bring expansion for yx MIMO. This is complete new revolution of WISPs !

Wed Dec 20, 2017 11:59 am

Hi, totally agree there are very few user interface information fir those who use it in professional basis, but enough for dummies. I believe MikroTik will release more tweaks for this band, it is not necessary to mess with band this way. Channel width is about 2 GHz, which wastes band, and capacity on 16QAM is 4,5GBPS. This needs to be chopped to subbbands like 225MHz, 500MHz or 1GHz channels to make better location planning which will be enough to carry 500Mbit in 16QAM, or 1Gbit in 256QAM etc.. We see big underdevelopment in radio part and antenna. There are also 4 nests for modules, which will bring expansion for yx MIMO. This is complete new revolution of WISPs !

Looks like the chipsets are limited at the moment. Ignitenet (peraso chipset) does not allow to change 2GHz channelwidth, too. Esp where lower bandwidth is enough it would save spectrum. And I guess smaller channels are more stable, too. 60GHz Wifi seems to be at the beginning but will be a very valuable tool.

vlatko · Wed Dec 20, 2017 12:42 pm

we cant (without enormous hassle) setup and maintain 60ghz link over more than 500m without rssi signal level.

Insider · Wed Dec 20, 2017 12:46 pm

Hi, totally agree there are very few user interface information fir those who use it in professional basis, but enough for dummies. I believe MikroTik will release more tweaks for this band, it is not necessary to mess with band this way. Channel width is about 2 GHz, which wastes band, and capacity on 16QAM is 4,5GBPS. This needs to be chopped to subbbands like 225MHz, 500MHz or 1GHz channels to make better location planning which will be enough to carry 500Mbit in 16QAM, or 1Gbit in 256QAM etc.. We see big underdevelopment in radio part and antenna. There are also 4 nests for modules, which will bring expansion for yx MIMO. This is complete new revolution of WISPs !
Looks like the chipsets are limited at the moment. Ignitenet (peraso chipset) does not allow to change 2GHz channelwidth, too. Esp where lower bandwidth is enough it would save spectrum. And I guess smaller channels are more stable, too. 60GHz Wifi seems to be at the beginning but will be a very valuable tool.

I dont think so..they are may be limited with modulation method as the LO for multiplier is really bad parameters, but about bandwidth I am completely sure, they can do what ever they want, there is I a Q baseband with 800MHz clock ducted to balance mixer which has harmonic 8 time mulitplier, so if you will put lower baseband the lower bandwidth will be available.

Insider · Wed Dec 20, 2017 12:50 pm

we cant (without enormous hassle) setup and maintain 60ghz link over more than 500m without rssi signal level.

If you use standard mesh antenna, which is fine for 5 GHZ you cant make fortune on 60 GHz, the wavelength is 1/2 cm, I am surprised you made more..

lazna · Thu Dec 21, 2017 4:52 pm

Only Signal and MCS are displayed (in 6.40.5). Lowest signal is 40 (MCS 4) than link disconnect. by Signal 45 is MCS 6 and by Signal 50 is MCS 8. I am not able to read both values from SNMP nor parsing web interface (to put values to my own monitoring system). Anybody have idea?

server8 · Thu Dec 21, 2017 6:04 pm

320m signal 80 MCS 8 full speed but one side wAP is inserted into LHG XL

Please can you post some pics about this hack?

Thank you

Fri Dec 22, 2017 9:35 am

wAP60G uses an complex algorithm to select best elements from patch array. With reflector it needs to be re calibrated to be able to work correctly.

Fri Dec 22, 2017 9:42 am

In addition to what Antons says, the LHG grid is not optimal to work as reflector for 60GHz. It will only reflect a small part of the signal. Need some more solid stuff

Insider · Fri Dec 22, 2017 1:08 pm

wAP60G uses an complex algorithm to select best elements from patch array. With reflector it needs to be re calibrated to be able to work correctly.

Are you able to provide us interface to enable certain groups of elements of phase array grid?

Insider · Fri Dec 22, 2017 1:12 pm

In addition to what Antons says, the LHG grid is not optimal to work as reflector for 60GHz. It will only reflect a small part of the signal. Need some more solid stuff

We would be happy tp interact more with your developer team on this issue a present whole bunch of new solutions, is there any way ?

Fri Dec 22, 2017 3:23 pm

We would be happy tp interact more with your developer team on this issue a present whole bunch of new solutions, is there any way ?

You are welcome to email support@mikrotik.com with your ideas, suggestions and observations

server8 · Fri Dec 22, 2017 5:33 pm

Normis please ETA for the LHG60?

Thank you

marcin21 · Sat Dec 23, 2017 12:23 am

@Insider: do You think that in 120m distance Wap60 is strong enough for AF24 replace (max300mbits of traffic) ?

MulderSk · Sat Dec 23, 2017 4:36 am

320m signal 80 MCS 8 full speed but one side wAP is inserted into LHG XL
Please can you post some pics about this hack?

Thank you

first I tested about 1 month when wAP 60 in LHG only with tape but today I recieved another units so I make new version

foto https://ulozto.sk/tam/_PQHTNYRtxU2w
second unit will make today I try better work

Insider · Sat Dec 23, 2017 12:18 pm

@Insider: do You think that in 120m distance Wap60 is strong enough for AF24 replace (max300mbits of traffic) ?

Should be OK

vlatko · Sat Dec 23, 2017 1:02 pm

obviously its possible to replace wap60 board into lhg box relatively easy. Still I am convinced that the lhg mesh reflector isnt best solution for 60ghz wavelenght.
i will repeat it again. Probably would be the best to replace somehow wap60 board inside ldf box and then mount ldf60 on any appropriate parabolic dish.
I'm pretty sure that mikrotik not planned lhg60 for reliable p2p link over 500m. i dont say its not be a possible but there is regulatory limitations and also price range.
it may be easier to replace lhg60 board inside ldf and then reach a greater distance with parabolic dish especially taking into account the problems with the antenna inside wap60.
I will repeat it again, all this is torture and big hasle if we can not further adjust the parameters of the 60ghz connection, as well monitoring the same ones on the active p2p link.

vlatko · Sat Dec 23, 2017 1:12 pm

i dont want to vacillate,
but watching pictures from the assembly wap60 board inside lhg box although you have very elegantly resolved board mounting,
I dont think you have positioned the antenna exactly in the middle of lhg box as it should be for optimum results
please consider this

server8 · Sat Dec 23, 2017 2:31 pm

first I tested about 1 month when wAP 60 in LHG only with tape but today I recieved another units so I make new version foto https://ulozto.sk/tam/_PQHTNYRtxU2w
second unit will make today I try better work

thank you a lot

Insider · Sat Dec 23, 2017 2:31 pm

As it was written before LHG is not the type of antenna which is eligible to work properly with the phase array antenna by the first, then the mesh is not proper surface. Except the phase distraction, the surface is not fully exposed to feeder. At the moment you need some surface with f/D about 0.7 - 0.9, the LHG has something like 0.3-0.4

MulderSk · Sat Dec 23, 2017 4:27 pm

obviously its possible to replace wap60 board into lhg box relatively easy. Still I am convinced that the lhg mesh reflector isnt best solution for 60ghz wavelenght.
i will repeat it again. Probably would be the best to replace somehow wap60 board inside ldf box and then mount ldf60 on any appropriate parabolic dish.
I'm pretty sure that mikrotik not planned lhg60 for reliable p2p link over 500m. i dont say its not be a possible but there is regulatory limitations and also price range.
it may be easier to replace lhg60 board inside ldf and then reach a greater distance with parabolic dish especially taking into account the problems with the antenna inside wap60.
I will repeat it again, all this is torture and big hasle if we can not further adjust the parameters of the 60ghz connection, as well monitoring the same ones on the active p2p link.

And what cooling ? how make cooling inside lfd ? wAP60 has very big cooler

vlatko · Sat Dec 23, 2017 9:25 pm

its ok to mount radio on metal cooling plate, i just say that antenna should be fixed to center of lhg box.

vlatko · Sat Dec 23, 2017 9:34 pm

mikrotik say, antennas in wap60g a phased array beamforming. i understand why is that important in p2mp conecting several wap60 to one from different directions.
but i dont understand when we have simple wap60 configuration or two wap60 in p2p how that work, meaning whitch patch antennas coresponding inbetween.
i doubt that type of connection include funcionality of all phased array patch.
possibly a key role is in some algorithm in device hardware, so i find its easily solvable but depends from good will from mikrotik.

lazna · Sat Dec 23, 2017 9:34 pm

And what cooling ? how make cooling inside lfd ? wAP60 has very big cooler

It should consume, by datasheet, only 5W. Also supplied power source is only 24V/0,38A. So I guess wAP60 is overcooled for some reason...

server8 · Sun Dec 24, 2017 10:34 am

I have some questions about hardware:

1) which kind of connector is for the antenna pigtail? Ufl?
2) the communications from board to antenna is base band or 60ghz?
3) is it possible to use longer pigtail?

Thank you

mistry7 · Sun Dec 24, 2017 12:00 pm

And what cooling ? how make cooling inside lfd ? wAP60 has very big cooler
It should consume, by datasheet, only 5W. Also supplied power source is only 24V/0,38A. So I guess wAP60 is overcooled for some reason...

In Idle WAP60g Uses 2,5 W at 23,5V
With max Best running over them it is 3,2 to 3,3 W

mistry7

lazna · Sun Dec 24, 2017 4:35 pm

I have some questions about hardware:

1) which kind of connector is for the antenna pigtail? Ufl?

Looks more likely like UMCC, but not verified...

vlatko · Sun Dec 24, 2017 11:03 pm

its not 60ghz antenna pigtail, its look like 60ghz radio or downconverter is integrated on back of phased array pcb,
first cable is power source,
pigtail possible have some rf role, downcoverting to lower freq? or connecting to baseband?

Insider · Wed Dec 27, 2017 12:08 pm

its not 60ghz antenna pigtail, its look like 60ghz radio or downconverter is integrated on back of phased array pcb,
first cable is power source,
pigtail possible have some rf role, downcoverting to lower freq? or connecting to baseband?

Cable contains base band and telemetry, it is not uFL neiter ipex, it is smaller

vladimirslk · Wed Dec 27, 2017 10:39 pm

one of the options, by the way the dimensions are in pdf as well available..

https://www.pasternack.com/images/Produ ... 881-24.pdf

lazna · Wed Dec 27, 2017 11:28 pm

one of the options, by the way the dimensions are in pdf as well available..

https://www.pasternack.com/images/Produ ... 881-24.pdf

QUANTITY PRICE
1-24 $1,182.91

vladimirslk · Thu Dec 28, 2017 12:49 am

who told you to buy, i think it is very simple to make it by your own, dimensions are available, construction is very simple... placing the tx-rx unit in the center and using "rouper" should give an directional improvement... i did not checked my thoughts, but it might work

who wanna try/ proove mathematically?

lazna · Thu Dec 28, 2017 2:00 am

who told you to buy, i think it is very simple to make it by your own, dimensions are available, construction is very simple... placing the tx-rx unit in the center and using "rouper" should give an directional improvement... i did not checked my thoughts, but it might work who wanna try/ proove mathematically?

Man, I do not know much about microwaves, but what I know is enough to imagine the price is equal of what you get in this case. If you want to DIY, than (without irony) good luck. So much variables are in play....

Insider · Thu Dec 28, 2017 11:33 am

who told you to buy, i think it is very simple to make it by your own, dimensions are available, construction is very simple... placing the tx-rx unit in the center and using "rouper" should give an directional improvement... i did not checked my thoughts, but it might work who wanna try/ proove mathematically?
Man, I do not know much about microwaves, but what I know is enough to imagine the price is equal of what you get in this case. If you want to DIY, than (without irony) good luck. So much variables are in play....

Sorted out..

server8 · Fri Dec 29, 2017 10:39 am

We have placed a link test the distance is 101 meters the link is dropped 4 time in 3 days with light rain. We have a meteo station close to the installation that measured 23mm/hour of rain.during last drop.

With no rain the singnal is 80

vlatko · Fri Dec 29, 2017 11:00 am

please give us details about link drop during a rain. Has it preceded degradation or is just link interrupt with the beginning of rain.if the smaller rain interrupts the link to 100 meters it will be great problem. I have already mentioned several times to the mikrotik how to assess the problem if we can not read signal rssi on devices.

Fri Dec 29, 2017 11:37 am

We have placed a link test the distance is 101 meters the link is dropped 4 time in 3 days with light rain. We have a meteo station close to the installation that measured 23mm/hour of rain.during last drop.

With no rain the singnal is 80

Please upgrade to latest version to support rates under MCS4 and check alignment.

server8 · Fri Dec 29, 2017 2:12 pm

please give us details about link drop during a rain. Has it preceded degradation or is just link interrupt with the beginning of rain.if the smaller rain interrupts the link to 100 meters it will be great problem. I have already mentioned several times to the mikrotik how to assess the problem if we can not read signal rssi on devices.

WIth very very light rain the signal level degrade quick to 50, the worst signal that I see is 30 then the link drop. If I put a speedtest sometimes the signal back to 70 or 80 even with rain, I can't explain why

server8 · Fri Dec 29, 2017 2:22 pm

Please upgrade to latest version to support rates under MCS4 and check alignment.

I am running 6.41.rc66.

Which vversion 6.41 or 6.42rc2 ?

vlatko · Fri Dec 29, 2017 4:12 pm

my 2 cents. i think signal drop is not about rain attenaution at least not on 100m. my opinion is that durring rain rf path enviroment change dynamically, so originated rf, expirience numerous reflections from rain drop and confuse phase array antenna algoritmh.if this is reason that leads us to next generation of mikrotik wap60, device for p2mp where i expect to solve simillar problem. also we must conclude mikrotik must rearange algoritmh on p2p wap60 device to p2p,its not need complex array algoritmh mehanism except for achieving wider beam coverage, and on the end that improved p2p device will be better solution to mounting on dish antenna

lazna · Fri Dec 29, 2017 5:32 pm

latest version to support rates under MCS4 and check alignment.

Just testing 6.42RC in my room: MCS could fall down to 1 by signal level 35/40, latency going up to 200ms time by time on low signal... When one unit was turned a bit, two signal LED on both units light-up but pings still timed out. There si a few seconds delay from the moment wehn signal leds light-up to pings start replaying. Expect to signal/MCS algorithm still need some improovements. Lowest signal I could achieve is 25.

Is there any way to read signal level and MSC from device programaticaly (SNMP or SSH)? Have custom monitoring system and want to include this boxes for testing.

MonkeyDan · Fri Dec 29, 2017 8:45 pm

Is there any way to read signal level and MSC from device programaticaly (SNMP or SSH)? Have custom monitoring system and want to include this boxes for testing.

You can get it via SSH using "/interface w60g monitor 0"
We really need SNMP support, though, for integration.
Awesome product otherwise!

lazna · Sat Dec 30, 2017 12:52 pm

Is there any way to read signal level and MSC from device programaticaly (SNMP or SSH)? Have custom monitoring system and want to include this boxes for testing.
You can get it via SSH using "/interface w60g monitor 0"

Thanks, I cant find it...

server8 · Tue Jan 02, 2018 11:00 am

New data from our outdoor test enviroment:

client 1: 97m
client 2: 105m
client 3: 200m wag60g in the focus of cambium450d prime focus parabola

the clients are on the same line (hidden nodes problem) few degrees of difference:

client 1: signal from 50 to 80
client 2: signal from 40 to 60
client 3: signal form 50 to 100

ih the last 24 hours wre have had some rain/wind events here the pics from meteo station the wap60g reports all clients dropping at:

22:21 jan/1/2018
22.44 jan/1/2018

no dropping in the heavier rain event in the mornig

mistry7 · Sat Jan 13, 2018 1:55 pm

on China MUM in Wireless Wire Presentation was Info of Antenna gain, 13,5dBi when i´m remember correct...
I missed the Hardware Presentation... hope the Slides are online next week.....

Lupin · Sat Jan 13, 2018 7:52 pm

I've made a simple 85mt. link with wAP 60G.
Signal is 65, MCS 8.. but when I start to pass TCP traffic across the link signal drop and link go down
If I test with UDP BWTest across the link still remain up

Why??

I'm running 6.41

Hesse · Sat Jan 13, 2018 9:45 pm

@ Lupin
Try max. MTU 1500 (L2MTU 1600) on Wireless-Interface not higher

server8 · Sun Jan 14, 2018 10:35 am

I can confirm this issue during heavy load (speetest udp both directions) the signal floats from 40 to 60 with loads until 200 Mb/s both direction the signal no drop and it stay stable at 80

Upgrade from 6.41 to 6.42rc6 not solve the problem.

Lupin · Sun Jan 14, 2018 11:10 am

I've made a simple 85mt. link with wAP 60G.
Signal is 65, MCS 8.. but when I start to pass TCP traffic across the link signal drop and link go down
If I test with UDP BWTest across the link still remain up

Why??

I'm running 6.41

Problem in my case is MTU.
I lowered from 1700 to 1600 and problem disappear

Sun Jan 14, 2018 7:11 pm

Can you please monitor CPU core (not CPU, but individual CPU core) usage during this TCP test and report back to us if it hits 100% when problem appears? You can see this usage under System/Resource/CPU menu.

server8 · Sun Jan 14, 2018 10:05 pm

Making test from wag60g
cpu0: 100%
cpu1: 0%
cpu2: 0%
Cpu3: 0%

Making several test from from HEX gigabit to rb3011 passing wag60g sometime the signal is stable at 80, sometime floats a little bit or sometime floats a lot. cpu0 load is always around 50% and one of the other 3 cpu goes aorund 65%

Ros version is 6.42rc6

mistry7 · Sun Jan 14, 2018 10:11 pm

One of the long issue list with ROS6

Bandwidth test is not capable for multi Core

Please don’t make Test from Device themselv!

mag2020 · Mon Jan 15, 2018 1:16 pm

I suggest that SFP port be added to this device in future models. That will make it even more superb. My concern is that the board may not have enough space for this.
With SFP included, it will be easy to use the wap g60 to bridge gaps on a fiber transmission path.

ndor · Sun Jan 21, 2018 9:45 am

Some pictures with the product , we can see ?

djvolt1942 · Sun Jan 21, 2018 3:28 pm

Yes, testing model: https://trzepak.pl/viewtopic.php?p=480703#p480703

Final product will be with 8 year guarantee for protection without rust and some modifications.

mag2020 · Sun Jan 21, 2018 4:44 pm

Yes, testing model: https://trzepak.pl/viewtopic.php?p=480703#p480703

Final product will be with 8 year guarantee for protection without rust and some modifications.

Nice work if you can achieve 1km. But, what are your antenna parameters like? What beam width do you achieve on the horizontal axis?

djvolt1942 · Sun Jan 21, 2018 8:23 pm

I don't have spectral analysis so I can't do this, but I think about 20 degrees, because wAPs have 60 degrees and we have more free space to align this antenna for 600 meters.

Antenna is renew NanoBridge 400m with more modifications.

marcin21 · Mon Jan 22, 2018 3:31 pm

Hello @Djvolt1942 could I ask for contact? I'm interested in some of your equipment.
Moze byc po polsku chetnie zakupie rozwiazanie nawet w fazie testow dla ptmp.

djvolt1942 · Mon Jan 22, 2018 4:14 pm

Please send email to: office[at]multimediahd.pl

djvolt1942 · Tue Jan 30, 2018 1:01 am

Dears

Our solution is finished.

Please check our website: http://siec.multimediahd.pl/60ghz#400

Products 300mm, 650mm and special wAP holder very soon.

At this time we preparing also Sector PtMP with special holder and housing for 120 degrees

Kind Regards

mybuzi · Sat Feb 03, 2018 10:30 am

Speed Issues with both firmware... i'm testing now the last release candidate, but i had the same problem with ufficial.

I've a link less than 100mt with 2 wap 60G

I've signal 80 with mcs 8, when i start bandwidth test with tcp signal go down to 50 and mcs to 4-5 and come up at 7 with signal 70-80 sometimes. Bandwidth is max 300mps.
If i test with udp protocol signal will be better and transfer rate will be 1gbps.

if i test a link with real transfert of a file from 2 servers i reach max 7mbps.

Please help me.. l2 mtu is 1600 and mtu 1500.

ARP is enabled
Cliend isolation is enabled
Channel on auto.

Now i've signal 50 and mcs 8 (bandwidth test is off).

I align the link not well? what's happend?

Mon Feb 05, 2018 3:21 pm

Please read previous posts testing. Bandwidth test runs on one core, currently You are testing one core packet generation performance, not link throughput.

djvolt1942 · Tue Feb 06, 2018 12:57 am

Our tests:

660m - https://www.youtube.com/watch?v=87kzOolah8A
1050m - https://www.youtube.com/watch?v=SR6lDmP49Q8

Bergante · Tue Feb 06, 2018 1:10 pm

Just received mine, I'm impressed

I have just tried them inside the office. The link length was about 5 metres.

In order to generate more bandidth with bandwidth-test I have ran two tests in transmit mode, one from each unit. Using UDP it sustained around 900 Mbps
in full duplex.

One problem, though. I have managed to break the locking tab of both table stands.

djvolt1942 · Wed Feb 07, 2018 1:43 am

Our new test on our Sector 120 degree with special Shield and also Precision 300 Station [300mm with special precision holder to milimeter align HV]

3 points = https://www.youtube.com/watch?v=z6c4JrmGkk0 [to show 120 degrees]
400m = https://www.youtube.com/watch?v=BMWKk_TVtWw

Steveocee · Thu Feb 08, 2018 4:38 pm

Does anyone have any clarity about UK requirements for these?
From my reading that whilst 60Ghz is license exempt OFCOM require a minimum 30dB antenna to give narrow beams. Reading above it seems the antenna is about 12-14dB so this would make the product unusable in the UK?

***Edit just re-read the distributors web page and they clearly state they are for indoor use only. Such a shame.

Bergante · Fri Feb 09, 2018 9:39 am

Does anyone have any clarity about UK requirements for these?
From my reading that whilst 60Ghz is license exempt OFCOM require a minimum 30dB antenna to give narrow beams. Reading above it seems the antenna is about 12-14dB so this would make the product unusable in the UK?

European regulators haven't made a decision yet, although it seems they are studying it. This report from Ofcom suggests it:
https://www.ofcom.org.uk/__data/assets/ ... report.pdf

But for now the only allowed usage without license is, at least in Spain (and I think the EU) is indoors, considering the 802.11ad standard.

However, I guess they will reconsider it eventually. The oxygen absorption peak severely limits the maximum achievable link distance but
at the same time it makes frequency reuse much easier.

Time will tell.

djvolt1942 · Fri Feb 09, 2018 5:13 pm

All in Europe: ETSI EN 302 217-2 = 06.2016 and ETSI EN 302 567 12.2016

ubikrotik · Sat Feb 10, 2018 11:06 pm

MESH wap 60 possible in the future? Sounds like more appropriate for this band. Short range, large spectrum space. What do you think ?

Mon Feb 12, 2018 10:47 am

MESH wap 60 possible in the future? Sounds like more appropriate for this band. Short range, large spectrum space. What do you think ?

For what kind of clients?

mag2020 · Mon Feb 12, 2018 6:32 pm

I suppose he is reffering to backhaul of traffic from one AP to another AP forming a meshed network. The huge capacity available on the wap60 makes it a candidate for traffic backhauling.

mag2020 · Mon Feb 12, 2018 6:33 pm

I suppose he is reffering to backhaul of traffic from one AP to another AP forming a meshed network. The huge capacity available on the wap60 makes it a candidate for traffic backhauling.

ubikrotik · Wed Feb 14, 2018 5:54 am

Not realy, urban deployment. from neighbors to neighbors.
not for backhauling.

p3rad0x · Thu Feb 15, 2018 10:34 am

I have one installed about 100m away and it works like a charm.

Getting 100 up and down even in the rain, haven't replaced the poe's yet to get a gig link

ubikrotik · Tue Feb 27, 2018 4:17 am

mesh implement in the future?

Pante1eev · Fri Mar 02, 2018 4:16 pm

Dear Sirs,
pls inform if you have any plan to include into software of model wAPG-60 support of NV2 protocol (for m2p networks)
Thanks

Pante1eev · Fri Mar 02, 2018 4:23 pm

dear sirs,
pls inform if you have any plan to include into software of model wAPG-60 support of NV2 protocol (for m2p networks)

Bergante · Mon Mar 05, 2018 1:00 pm

A suggestion.

I see that the standard MIB for the 60 GHz interface includes both discards and errors statistics. Are those variables actually updated? It would be helpful to know it for sure.
A note in the documentation would help.

lazna · Mon Mar 26, 2018 1:12 pm

Is this improovement:

-----------------------------------------------------------------------------
What's new in 6.42rc49 (2018-Mar-21 12:34):
*) w60g/Wireless-Wire - increased supported distance for wAP 60G to 200+ meters;
-----------------------------------------------------------------------------

somehow related to MCS 1-4 provisioning announcement in late december 2017? If not, how is greater operating distance achieved? By increasing of TX power?

Mon Mar 26, 2018 1:14 pm

By increasing of TX power?

No, by optimizing the driver.

lazna · Mon Mar 26, 2018 1:41 pm

No, by optimizing the driver.

So this man https://youtu.be/S9bMO4r9htI?t=1090 says no truth?

Mon Mar 26, 2018 2:04 pm

No, by optimizing the driver.
So this man https://youtu.be/S9bMO4r9htI?t=1090 says no truth?

Where does he say that the only thing done was boosting TX power?
Several improvements were done in the driver, including beamforming optimisations and others.

lazna · Mon Mar 26, 2018 2:56 pm

No, by optimizing the driver.
So this man https://youtu.be/S9bMO4r9htI?t=1090 says no truth?
Where does he say that the only thing done was boosting TX power?
Several improvements were done in the driver, including beamforming optimisations and others.

He ment greater distance togeather with provisioning MCS under 4 and says something like: "...will suffer with less speed, but distance could be greater", which could be misleading. Thank you for your clarification.

BTW: Could you more specific "and others"? Many people accross web forums experimenting are surely interesting in details.... Thanks

Mon Mar 26, 2018 3:16 pm

No we can't be more specific, sorry. This is internal workings of a proprietary technology.

lazna · Mon Mar 26, 2018 4:23 pm

No we can't be more specific, sorry. This is internal workings of a proprietary technology.

Understood. Thanks anyway....

NeoPhyTex360 · Wed Apr 04, 2018 4:15 pm

Have someone tried with some "light" obstacles between the aps? Like trees (not fat branches, yes light leaves), etc.

I'm thinking about to purchase them, now I have -40dBm (got -50dBm in summer with more leaves) on 100m link with 5.8Ghz 5AC 19dBi with 2-3 trees in the middle (most all leaves), and I don't know if would be possible to upgrade the link to 60Ghz with that soft obstacles (or wait to the more powerful and directional LHG 60G to awoid trees). I don't spect to get the full gig, but I'd like to improve my actual real (230~/230mbps) rates

Thanks!

lazna · Sun Apr 08, 2018 12:42 pm

There are T/Rx-sector items in

interface w60g monitor 0 once

command output.Anybody know what this last two lines exactly means?

connected: yes
       frequency: 60480
  remote-address: 30:07:4D:81:D0:B6
             mcs: 8
          signal: 80
       tx-sector: 54
       rx-sector: 96

Sun Apr 08, 2018 1:06 pm

RC releases will show more information on aligment by using sector calibration inforation

There are T/Rx-sector items in
Code: Select all
interface w60g monitor 0 once
command output.Anybody know what this last two lines exactly means?
Code: Select all
connected: yes
       frequency: 60480
  remote-address: 30:07:4D:81:D0:B6
             mcs: 8
          signal: 80
       tx-sector: 54
       rx-sector: 96

npero · Thu Apr 12, 2018 8:44 am

wAP 60G by the google maps 238m distance. In good weather condition

mcs: 4
          signal: 40
            rssi: -70
       tx-sector: 28
  tx-sector-info: left 3.8 degrees, down 3.8 degrees
       rx-sector: 96

In btest UDP have 600-800Mbps also have half hour with rain speed drop to 100Mbps in test.
For now wating rain again to see results.

juliokato · Mon Apr 16, 2018 8:39 pm

we have using this?
https://mikrotik.com/product/wireless_wire_dish

djvolt1942 · Mon Apr 16, 2018 8:47 pm

wAP 60G by the google maps 238m distance. In good weather condition
Code: Select all
mcs: 4
          signal: 40
            rssi: -70
       tx-sector: 28
  tx-sector-info: left 3.8 degrees, down 3.8 degrees
       rx-sector: 96
In btest UDP have 600-800Mbps also have half hour with rain speed drop to 100Mbps in test.
For now wating rain again to see results.

On what ROS? Newest r59 or 6.41.4?

There are same wAP60G link without any our shields etc?

brasileottanta · Tue Apr 17, 2018 5:36 pm

Hello ,

our first test on real link. Distance about 240m , perfect LOS without particular aim.

We using traffic generator for check bandwidth and latency. Test using tcp packet of 1500.

Very nice

brasileottanta.

ps . please make a note on rc59 changelog for Ap Bridge BROKEN !! :-9

npero · Tue Apr 17, 2018 7:35 pm

On what ROS? Newest r59 or 6.41.4?

There are same wAP60G link without any our shields etc?

rb56 for now, planing upgrade to final version in Friday. Just wAP60G unit from Wire kit package without anything.
Also today is raining have some link drops but they has been just for 1s that say in log.

Wed Apr 18, 2018 2:54 pm

Ap-bridge mode should be fixed in latest current and RC releases.

Insider · Thu Apr 19, 2018 1:05 pm

Sorry for delay I had long period of heavy metal

Too busy to make things which make sense..

So.. Our tests.. With the regular body of wAP we are getting stable link for 150m, anything more make pain during rain (yes rains came finally)

The good story is we are making with wAP now links for more then 500m, with regular AP and modified clients..
Here we go 240 m with modified client, last prototype:
connected: yes
frequency: 58320
remote-address: 30:07:4D:84:A0:52
mcs: 8
signal: 95
rssi: -56
tx-sector: 27
tx-sector-info: right 3.8 degrees, down 3.8 degrees
rx-sector: 96

If we will make both clients same way it will have on same link rssi:-48, we will see if MT will go for this solution..

jarod · Mon Apr 23, 2018 8:15 pm

Hello,

Question;

With RouterOS 6.42 i saw w60g > status: MCS: 9/10/11.
Now with RouterOS 6.42.1 i only see MCS: 8 or lower. Is this a bug of feature?

jarod

Insider · Thu Apr 26, 2018 12:43 pm

Multiple instalallation with in close range on pole brings unstability, we are now installing shields to try if performance will be better (1 site, 3APs, 20 cm far each from other, different frequencies, 3 clients on each)

sup5 · Thu Apr 26, 2018 5:37 pm

Hello,

Question;

With RouterOS 6.42 i saw w60g > status: MCS: 9/10/11.
Now with RouterOS 6.42.1 i only see MCS: 8 or lower. Is this a bug of feature?

jarod

Seems to be a feature.
http://rfmw.em.keysight.com/wireless/he ... ncepts.htm
MCS8 is more than enough to carry 2Gbps.
More is not necessary, because 1000Base-T is the limiting factor here.

luksik · Tue May 08, 2018 3:50 pm

Hi,
I've recently tested p2p on wAP 60G with tagged transmission to switch and vlans. Unicast is working without ant problems but i've noticed that multicast (iptv) streams has problem with continuity.
I'm generating (using iperf) tcp traffic between two machines that are using 60g connection and problems appears but only if traffic is from same side that multicast stream is coming.
When i'm generating (about 1Gbps) in backard direction there is no continuity errors.
Any ideas ? (link has about 10m - set on table).

Lukasz

Bergante · Thu May 10, 2018 9:57 am

Hi,
I've recently tested p2p on wAP 60G with tagged transmission to switch and vlans. Unicast is working without ant problems but i've noticed that multicast (iptv) streams has problem with continuity.
I'm generating (using iperf) tcp traffic between two machines that are using 60g connection and problems appears but only if traffic is from same side that multicast stream is coming.
When i'm generating (about 1Gbps) in backard direction there is no continuity errors.
Any ideas ? (link has about 10m - set on table).

I wonder... Do these units lower the MCS when transmitting multicast or broadcast traffic, like other wireless protocols? For wireless point to point links that shouldn't happen in my opinion.

If the link is very short and you are running a recent version check the signal levels, you might be suffering signal overload. I had to lower power for a 20 - 30 m link because it got a bit unstable.

luksik · Thu May 10, 2018 12:44 pm

Hi,
I've recently tested p2p on wAP 60G with tagged transmission to switch and vlans. Unicast is working without ant problems but i've noticed that multicast (iptv) streams has problem with continuity.
I'm generating (using iperf) tcp traffic between two machines that are using 60g connection and problems appears but only if traffic is from same side that multicast stream is coming.
When i'm generating (about 1Gbps) in backard direction there is no continuity errors.
Any ideas ? (link has about 10m - set on table).

I wonder... Do these units lower the MCS when transmitting multicast or broadcast traffic, like other wireless protocols? For wireless point to point links that shouldn't happen in my opinion.

If the link is very short and you are running a recent version check the signal levels, you might be suffering signal overload. I had to lower power for a 20 - 30 m link because it got a bit unstable.

I've tested this setup on short link (table) and outside (link about 90m). Same problem in both cases,

Lukasz

hrcho · Fri May 11, 2018 4:17 pm

Can I expect full throughput at 70 m distance through one tree using the Wireless Wire Dish (RBLHGG-60ad kit)?

Fri May 11, 2018 4:25 pm

You cant expect anything with a tree.

Lakis · Tue May 15, 2018 12:18 am

Anyone tested wAP 60G (with no modification) on 210m can I get >150mbit, should I expect disconnect on light rain?
Thanks.

Bergante · Tue May 15, 2018 10:14 am

One suggestion: It would be great to be able to graph information such as RSSI, signal, MCS, tx-sector and rx-sector (or better yet, tx-sector-info) and Phy rate these interfaces.

Given that some of it is not available over SNMP and that even for people with SNMP management software it can be tricky to add custom OIDs I think it would be really helpful!

Lakis · Thu May 17, 2018 12:20 am

Anyone tested wAP 60G (with no modification) on 210m can I get >150mbit, should I expect disconnect on light rain?
Thanks.

Link is up and running distance 210m, throughput is 350mbit (btest is made from wAP to wAP), 24h no disconnects for now looks promising, just waiting for the rain.
Only one minor problem, I can't understand why signal is fluctuating so much from 30-60, every second signal is changing 30-35-40-50-55-60 than again 60-50-45-35-30 all the time
Why is that ?

Oh yes I m using 6.42.1 OS.
Thanks

djvolt1942 · Thu May 17, 2018 12:31 am

You should use our shields: http://24.multimediahd.pl/60ghz/2-shiel ... wap-s.html or: http://24.multimediahd.pl/60ghz/96-shie ... s-cnc.html

Bergante · Thu May 17, 2018 9:12 am

Link is up and running distance 210m, throughput is 350mbit (btest is made from wAP to wAP), 24h no disconnects for now looks promising, just waiting for the rain.
Only one minor problem, I can't understand why signal is fluctuating so much from 30-60, every second signal is changing 30-35-40-50-55-60 than again 60-50-45-35-30 all the time
Why is that ?

Multipath? What does the environment look like?

With the latest releases you can try to fix the "tx-sector" parameter.

Lakis · Thu May 17, 2018 1:00 pm

djvolt1942 shields for what?
Bergante Multipath? - don't get it!, What does the environment look like? ?!like moisture, LOS... ??

marcin21 · Thu May 17, 2018 1:06 pm

60ghz-6.41rc52vs6.42.1.jpg

not sure, should I tweak something after upgrade ?

Thu May 17, 2018 1:22 pm

How far is the link? Please provide CLI output from both ends:
/interface w60g monitor wlan60-1

marcin21 · Thu May 17, 2018 1:29 pm

/interface w60g monitor wlan60-1
connected: yes
frequency: 60480
remote-address: 30:07:4D:
mcs: 8
signal: 50
tx-sector: 63
rx-sector: 96

interface w60g monitor wlan60-1
connected: yes
frequency: 60480
remote-address: 30:07:4D:
mcs: 8
signal: 50
tx-sector: 24
rx-sector: 96

Link length is around 130m.
but I downgraded them to rc52 to have more stable link now.

Thu May 17, 2018 2:11 pm

Try to switch used frequency to 58320 and upgrade to 6.42.1

djvolt1942 · Thu May 17, 2018 2:17 pm

We use 58GHz for far link like 600, 800 meters

Lakis - shields for far links up to 300meters or better stability in snowy/rainy day, also for reduce SNR from wap to wap or MultiPoint...

Bergante · Thu May 17, 2018 5:44 pm

Bergante Multipath? - don't get it!, What does the environment look like? ?!like moisture, LOS... ??

It's a long shot, but depending on how you have installed the units you might be getting multiple reflections? One of my
units is installed on a window (outside of course) and it's in a corner with three walls around. That's a scenario
that can be troublesome in some cases.

In theory that might sometimes confuse the beam selection algorithm, and, if that was the case (sorry, it's really
hypothetical but I haven't seen your installation) selecting a fixed beam configuration might help.

Anyway the experiment is easy. If it's just a point to point link look at the tx-sector values when the
link is performing best and configure them statically. if that solves the problem then you had something
disturbing the beam selection algorithm.

marcin21 · Thu May 17, 2018 5:45 pm

Try to switch used frequency to 58320 and upgrade to 6.42.1

why should I change freq? It seems to work quite good with rc52. there is no noise or other 60ghz link visible under scan. besides if it works worse with newer routerOS
I doubt Your advice is to solve problem looks rather as a workaround.
Sad.

Bergante · Thu May 17, 2018 6:25 pm

Try to switch used frequency to 58320 and upgrade to 6.42.1
why should I change freq? It seems to work quite good with rc52. there is no noise or other 60ghz link visible under scan. besides if it works worse with newer routerOS
I doubt Your advice is to solve problem looks rather as a workaround.

Are the tx-sector and rx-sector numbers changing often? If they are, fixing tx-sector could be beneficial.

My link (very short, around 30 metres crossing a street) actually worsened a lot when Mikrotik improved the antenna patterns. But the only reason was signal overload. Reducing power solved it.

Monitor tx-sector regularly and let's see.

Thu May 17, 2018 6:50 pm

Try to switch used frequency to 58320 and upgrade to 6.42.1
why should I change freq? It seems to work quite good with rc52. there is no noise or other 60ghz link visible under scan. besides if it works worse with newer routerOS
I doubt Your advice is to solve problem looks rather as a workaround.
Sad.

Different frequencies has different attenuation over air. Channel 2 isn't picked up automatically on free spectrum if "frequency=auto"

MonkeyDan · Mon May 21, 2018 8:46 pm

I had the same problem with a ~30m link. New firmware was just too hot. Lowering tx-power to 10 on both sides helped tremendously

Try to switch used frequency to 58320 and upgrade to 6.42.1
why should I change freq? It seems to work quite good with rc52. there is no noise or other 60ghz link visible under scan. besides if it works worse with newer routerOS
I doubt Your advice is to solve problem looks rather as a workaround.
Are the tx-sector and rx-sector numbers changing often? If they are, fixing tx-sector could be beneficial.

My link (very short, around 30 metres crossing a street) actually worsened a lot when Mikrotik improved the antenna patterns. But the only reason was signal overload. Reducing power solved it.

Monitor tx-sector regularly and let's see.

MonkeyDan · Thu May 24, 2018 11:29 pm

And 6.43rc17 has vastly improved one of our longer links (~170m)
Waiting for a final release to upgrade to the rest.

marcin21 · Fri May 25, 2018 1:10 am

And 6.43rc17 has vastly improved one of our longer links (~170m)
Waiting for a final release to upgrade to the rest.

what FW was before upgrade to 43rc17 ?

almdandi · Fri May 25, 2018 1:20 am

@MonkeyDan
What software are you using for the grath visualization?

marcin21 · Fri May 25, 2018 1:27 am

after upgrade from
6.42rc52 to 43rc17
i've got ERROR WRONG PASSWORD after trying to login into ( using winbox 3.11)
?!?!
is it some kind of very late aprilfool joke?
PLEASE mikrotik support help me solve my problem

npero · Fri May 25, 2018 9:13 am

Look viewtopic.php?f=21&t=133420&start=100#p663507 at the end of post say need to use new Winbox 3.14 in rc17.

Fri May 25, 2018 11:34 am

new version will have distance measurement tool, it would be cool if we could get some feedback to support@mikrotik.com

marcin21 · Fri May 25, 2018 2:09 pm

@npero thanks, with this beta winbox and webbox password works ok.

@antonsb where Can I find dist. measurment tool ?

Bergante · Fri May 25, 2018 2:55 pm

@antonsb where Can I find dist. measurment tool ?

On the AP/master: /interface w60g station monitor 0

[admin@w60-master] /interface w60g station> monitor 0
       connected: yes
          remote: 30:07:4D::
             mcs: 8
        phy-rate: 2.3Gbps
          signal: 80
            rssi: -54
       tx-sector: 21
  tx-sector-info: left 11.4 degrees, down 11.4 degrees
       rx-sector: 96
        distance: 18.34m

And on the slave,

[admin@w60-slave] /interface w60g> monitor 0
       connected: yes
       frequency: 58320
  remote-address: 30:07:4D::
             mcs: 8
        phy-rate: 2.3Gbps
          signal: 80
            rssi: -53
       tx-sector: 20
  tx-sector-info: left 3.8 degrees, down 11.4 degrees
       rx-sector: 96
        distance: 18.33m
-- [Q quit|D dump|C-z pause]

Before someone gets nervous with the signal levels at a distance of 20 metres: I lowered tx-power to 10 for both units.

marcin21 · Fri May 25, 2018 4:34 pm

strange on 6.43rc17 i get this :/

int w60g station monitor 0                
  connected: yes
     remote: 30:07:4D:
        mcs: 8
   phy-rate: 2.3Gbps
     signal: 80
       rssi: -70
  tx-sector: 63
  rx-sector: 96

MonkeyDan · Fri May 25, 2018 5:41 pm

@MonkeyDan
What software are you using for the grath visualization?

rrdtool https://oss.oetiker.ch/rrdtool/
Specifically a Perl script to do an snmpwalk and write the values to an RRD which can be graphed with rrdgraph.

MonkeyDan · Fri May 25, 2018 5:41 pm

And 6.43rc17 has vastly improved one of our longer links (~170m)
Waiting for a final release to upgrade to the rest.
what FW was before upgrade to 43rc17 ?

6.42.1

brasileottanta · Sat May 26, 2018 10:55 pm

after upgrade to 43rc19

djvolt1942 · Sat Jun 02, 2018 12:50 am

w60g_1421 .jpg

Test in Rainy day, Light Rainfall, on 1200 meters link Rock Solid Signal 80 mcs 8 everytime
New model of aluminium antenna with draft shield on the front soon

https://www.youtube.com/watch?v=78On7sLnv_k
http://siec.multimediahd.pl/60ghz#400

Insider · Mon Jun 11, 2018 11:38 am

Hi guys, we finalised a EM concentrator solution for ranges up to 600m.. It is compact solution for current housing with extra gain 10 dB for each end..

connected: yes
frequency: 58320
remote-address: 30:07:4D:87:84:16
mcs: 8
phy-rate: 2.3Gbps
signal: 80
rssi: -51
tx-sector: 27
tx-sector-info: right 3.8 degrees, down 3.8 degrees
rx-sector: 96
distance: 231.2m

Shortly we will have ready market solution

We found some bug, that by this level of signal device disconnects time to time

Mon Jun 11, 2018 12:32 pm

This will be fixed in next RC version

Insider · Mon Jun 11, 2018 1:37 pm

OK, I lied you.. Should be fine up to 1 km with -63
We will build a link shortly for 973 m to test it..
Here it is how it looks like in prototypes, it has simple aiming tool (which is completely correct)

Next week we will improve the LHG.

djvolt1942 · Mon Jun 11, 2018 2:13 pm

This Horn is made from Plastic or Metal?

You can reach on this about 600m?

Insider · Mon Jun 11, 2018 2:14 pm

This Horn is made from Plastic or Metal?

You can reach on this about 600m?

What you prefer, plastic or metal ? We can make you limited edition

We will reach 1km with this.

Mon Jun 11, 2018 2:20 pm

This Horn is made from Plastic or Metal?

You can reach on this about 600m?
What you prefer, plastic or metal ? We can make you limited edition

We will reach 1km with this.

You try to sell your ugly stuff in the MT-Forum? Please keep away with this. Order a booth at a MUM if you want to make money or ask MT to sell you AD-Banners on the website.

djvolt1942 · Mon Jun 11, 2018 2:29 pm

ok ok, the wAPs will be original with this or we will do something with original plastic case?

I think Aluminum will be much better than plastic...

Please send me price on PM

Insider · Mon Jun 11, 2018 2:34 pm

You try to sell your ugly stuff in the MT-Forum? Please keep away with this. Order a booth at a MUM if you want to make money or ask MT to sell you AD-Banners on the website.

I am not trying, to sell anything to anybody. In business, there are certain conditions which are specified offer and demand, it is reasonable to know price for buyer, and the seller should offer it somewhere something to sell. Those two condition were not met, thats why you get probably confused about, once you ask.

Insider · Mon Jun 11, 2018 2:38 pm

ok ok, the wAPs will be original with this or we will do something with original plastic case?

I think Aluminum will be much better than plastic...

Please send me price on PM

I am not trying to sell, we are developers and we sell solutions to large players, you will find soon some relevant offers in distribution chains, depending who will take care about opportunity.

jarod · Fri Jun 22, 2018 9:35 pm

Hello,

Question;

With RouterOS 6.42 i saw w60g > status: MCS: 9/10/11.
Now with RouterOS 6.42.1 i only see MCS: 8 or lower. Is this a bug of feature?

jarod
Seems to be a feature.
http://rfmw.em.keysight.com/wireless/he ... ncepts.htm
MCS8 is more than enough to carry 2Gbps.
More is not necessary, because 1000Base-T is the limiting factor here.

Just to make the story complete.
In 6.42.4 i noticed MCS 10 and MCS 11 again.
Not sure if it got changed back or something prevented me from going 'full' speed before.

Testing with winbox 3.15 and Bandwidth Test 6.42.4, Wireless Wire.

jarod · Sat Jun 30, 2018 11:52 pm

Update 6.42.5:
*) w60g - added 4th 802.11ad channel (CLI only);
*) w60g - added distance measurement;
*) w60g - general stability and performance improvements;
*) w60g - improved maximum achievable distance;
*) w60g - properly report center status under "tx-sector-info";

Only seeing MCS 8 again.
Which has a PHY Rate of 2.3 Gbps, which of course is more than enough.
Cool updates! Hope to see the 4th channel in the GUI soon.

And in the future, can you use 1 Wireless Wire as 60 GHz WiFi AP?
Or maybe you already can? Still learning a lot.

Sun Jul 01, 2018 10:54 pm

Update 6.42.5:
*) w60g - added 4th 802.11ad channel (CLI only);
*) w60g - added distance measurement;
*) w60g - general stability and performance improvements;
*) w60g - improved maximum achievable distance;
*) w60g - properly report center status under "tx-sector-info";

Only seeing MCS 8 again.
Which has a PHY Rate of 2.3 Gbps, which of course is more than enough.
Cool updates! Hope to see the 4th channel in the GUI soon.

And in the future, can you use 1 Wireless Wire as 60 GHz WiFi AP?
Or maybe you already can? Still learning a lot.

You can if You upgrade license to level4, or You can get wAP60G AP:
https://mikrotik.com/product/wap_60g_ap

lazna · Sun Jul 08, 2018 7:00 pm

Introduction of QAM16 in 6.43.42 was not succesfull. MCS 10 but performance close to dial-up networking :-/

lotnybartek · Mon Jul 09, 2018 9:49 pm

Need your advice.

I just installed wAP60g kit but horizontally. Left is MASTER, right is SLAVE. Is this supported? Should I cover any holes where water could potentially leak? With what? Silicone? Anyway, it works well. Even with heavy rain - and I mean heavy as fcuk.

Distance is 72m and stats as seen here:

npero · Tue Jul 10, 2018 11:01 am

The cable must ti oriented down, rotate you device that cable is down, in that case not need any other method to make it water proof.
Mounting like on picture is just mater of time when water or condensed water will fill up case and fry electronic inside.

MUST BE installed vertically if you need to work long time

.

lotnybartek · Tue Jul 10, 2018 12:46 pm

Slave is mounted to the window sill and this is the only way - house owner do not allow us to mount it in any other way. I mounted it using quickmount pro, but I'll try to mount it using quickmount pro lhg - it should be in correct position.

lotnybartek · Wed Jul 11, 2018 10:13 pm

Fixed

Everything works as expected

jarod · Thu Jul 12, 2018 7:27 pm

https://mikrotik.com/download/changelogs

What's new in 6.42.6 (2018-Jul-06 11:56):

*) bridge - improved packets processing when bridge port changes states;
*) crs3xx - fixed bonding slave failover when packets are sent out of the bridge interface;
*) crs3xx - fixed LACP member failover;
*) crs3xx - improved link state detection when one side has disabled interface;
*) defconf - fixed bridge default configuration for SOHO devices with more than 9 Ethernet interfaces;
*) package - free up used storage space consumed by old RouterOS upgrades;
*) snmp - fixed w60g "phy-rate" readings;
*) supout - added "ip-cloud" section to supout file;
*) w60g - fixed random disconnects;
*) w60g - general stability and performance improvements;
*) winbox - added 64,8 GHz frequency to w60g interface frequency settings;
*) winbox - show "sector-writes" on devices that have such counters;

Important update! 4th channel/frequency, in case you may need it

But most importantly, the random disconnects are fixed. And i thought i did something wrong.
Very happy with this update, good job!

Update:
After a night of sleep, 7 link downs. Testing in the living room.
Now off to work, see how it goes.

Update 2:
After about 10 hours, no more link downs. Who knows, maybe my living room is haunted during the night.

Will keep testing!

Update 3:
Link downs from 8 to 23 during day time.
So yeah, the fix is not 100% yet.
Hope the developers read this.

cwachs · Mon Jul 16, 2018 12:49 am

Installed a LHGG-60ad link a few days ago. Couple questions for those that have some of these up:

1) Is there an SNMP OID for the RSSI? We are graphing signal and MCS but I prefer to capture RSSI.

2) What are you doing to aim? I have aimed Siklu, IgniteNet and other millimeter wave gear but the aiming for these is different. I feel like there is beam steering on the AP and it makes it very hard to aim with beam steering on. I'd like a more responsive RSSI meter or another method for aiming. We keep hitting side lobes but the numbers are not very responsive. Are there tricks or can you turn off beam steering (if it's on)?

3) Where do I adjust TX power? We are a 233 meter link and I think "default" power is too strong. We see lots of fluctuations with signal.

Firmware is 6.42.6 Not wanting to run RC's on a production link...

djvolt1942 · Mon Jul 16, 2018 12:55 am

in CLI:

/interface w60g set wlan60-1 tx-power=1...15

server8 · Mon Jul 16, 2018 10:13 am

Someone knows the default tx poxer of lhg and wag60?

djvolt1942 · Mon Jul 16, 2018 12:08 pm

LHG60: I think 15

wAP60: I think 5

server8 · Mon Jul 16, 2018 1:18 pm

Theorically default tx power of LHG is less than WAP60

In old thread I find this

salamon · Wed Aug 15, 2018 6:33 pm

Hi Mikrotik dudes, do you plan add an SFP interface on wAP 60G ?

jarod · Thu Aug 23, 2018 5:04 pm

What's new in 6.42.7 (2018-Aug-17 09:48):

MAJOR CHANGES IN v6.42.7:
----------------------
!) security - fixed vulnerabilities CVE-2018-1156, CVE-2018-1157, CVE-2018-1158, CVE-2018-1159;
----------------------
...
*) w60g - added distance measurement feature;
*) w60g - fixed random disconnects;
*) w60g - general stability and performance improvements;
*) w60g - improved MCS rate detection process;
*) w60g - improved MTU change handling;
*) w60g - properly close connection with station on disconnect;
*) w60g - stop doing distance measurements after first successful measurement;
...

Link has been online all night and day after this update.
Testing in my living room. Wireless Wire that is.

So very advised update! Good stuff.

Update:
Last Link Up Time: Aug/23/2018 04:59:43 (now Aug/25/2018/06:05:00)
Link Downs: 0

Very happy with this update.

server8 · Tue Sep 11, 2018 4:39 pm

Just curious to know if someone has put 8 in clients on WAP 60?

djvolt1942 · Tue Sep 11, 2018 4:42 pm

Yes, total 18 clients for 3 sectors with wAP60G, Todays max 200meters form Sector...

server8 · Tue Sep 11, 2018 4:48 pm

Have you tried to run 8 speedtest from the clients? Can you share the performance?

salamon · Tue Sep 18, 2018 6:19 pm

Is there some tests completed on RBLHGG-60ad with frost cover on the reflector ?
What can we expect, how it will behave in winter?

Elliot · Fri Oct 05, 2018 4:05 pm

Is there some tests completed on RBLHGG-60ad with frost cover on the reflector ?
What can we expect, how it will behave in winter?

I also like to add a question how many clients AP support? with new LHG 60G lite for clients it's datasheet says it can support maximum of 3 clients per AP which seems quite low imho.

Fri Oct 05, 2018 4:13 pm

All 60GHz devices in AP mode currently supports up to 8 clients.

server8 · Fri Oct 05, 2018 4:23 pm

@antonsb in the next future we can expect 16 o more clients in AP mode?

All 60GHz devices in AP mode currently supports up to 8 clients.

ciberica · Sun Oct 28, 2018 5:55 am

Hi, I have problems with my mikrotik. I can not get more than 200 symmetric Mbs

If I do a btest only with Tx = 596 and Rx = 0 Mbs
If I do u btest only reception TX = 0 and RX = 450 Mbs
I do the teas with Tx and RX and the result is Tx = 145 and Rx = 98 - 120 (it has fluctuations)
The distance <600 m

attached screen capture to help me

thank you

server8 · Sun Oct 28, 2018 8:13 am

check the cpu cores load of both antennas during the speedtest, from terminal:

system resource cpu print interval=1

mistry7 · Sun Oct 28, 2018 9:13 am

check the cpu cores load of both antennas during the speedtest, from terminal:

system resource cpu print interval=1

And don’t forget that LhG60 has Quad Core CPU and Bandwidth Test only utilized one core.....
So if you see 25% CPU using, it is maxed out!

WirelessRudy · Mon Nov 19, 2018 5:35 pm

How can I have one wAP60G client have a signal of 45 while MCS 6 or 7 and a second wAP60G client have signal 55 but only MCS3?

Would higher more signal not mean higher MCS?

Bergante · Tue Nov 20, 2018 5:55 pm

How can I have one wAP60G client have a signal of 45 while MCS 6 or 7 and a second wAP60G client have signal 55 but only MCS3?

Would higher more signal not mean higher MCS?

Could be multiple reflections? Difficult to know without having a look.

I have been following the early versions and although my link is short (18 m, just crossing a street) it proved to be troublesome with multiple reflections. The last
beta version (6.44beta28) is rock solid for me. However, make sure to run the same version in the clients and AP units. Otherwise I think that the beam training features can fail.

gemb · Tue Nov 20, 2018 9:18 pm

LHG60 @4,7km

Of course its unstable, just for testing purposes

4km.png

cohprog · Tue Nov 20, 2018 11:31 pm

When will we get MCS12?

1300 meters link with latest Metrolinq firmware:

r00t · Wed Nov 21, 2018 3:33 am

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....

mistry7 · Wed Nov 21, 2018 4:15 am

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....

100% approval

cohprog · Wed Nov 21, 2018 8:21 am

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....

I could be wrong but I think regulations mandate the use of 2Ghz channels in that band.
(and the Metrolinq has a 2.5G ethernet port)

mistry7 · Wed Nov 21, 2018 9:18 am

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....
I could be wrong but I think regulations mandate the use of 2Ghz channels in that band.
(and the Metrolinq has a 2.5G ethernet port)

I know Ignite since 2015, but Mikrotik doesn’t have any 60G Device with More then 1gig port.

Channel widths could be Hardware problem but no regulation problem!

WirelessRudy · Wed Nov 21, 2018 12:08 pm

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....

100% agree. If we would be able to halve the bandwidth we could still serve 4-8 clients per AP but since we now could have 8 different frequencies in total we can make an 8 AP tower serving a max of 64 clients. By smart placing the AP's and distribution of the channels we would still have a undisturbed (interference) spectrum for the clients and would still be able to deliver 100Mbps or more to the clients.

WirelessRudy · Wed Nov 21, 2018 12:33 pm

How can I have one wAP60G client have a signal of 45 while MCS 6 or 7 and a second wAP60G client have signal 55 but only MCS3?

Would higher more signal not mean higher MCS?
Could be multiple reflections? Difficult to know without having a look.

I have been following the early versions and although my link is short (18 m, just crossing a street) it proved to be troublesome with multiple reflections. The last
beta version (6.44beta28) is rock solid for me. However, make sure to run the same version in the clients and AP units. Otherwise I think that the beam training features can fail.

Well, don't know. We have 4 wAP60G AP units in a mast, using all the 4 frequencies.
We have one AP with 1 client at 132 meter and both units running 6.44beta28.........rock solid
But the other 3 wAP60G AP PtMP networks were all on 6.44beta28 as well, all with 2 or 4 clients, all with 20 to 150 meter distance, and all loads of cuts.Special with light rain loads of cuts. Even more then when we have heavy rain which only sees the signals en MCS rates drop.
Two of these wAP60 AP's we already brought back to 6.43.4 which made the drops much less.
I have seen several saying 6.44beta28 is rock solid, to us in general it doesn't seem so....

autostoper76 · Wed Nov 21, 2018 12:35 pm

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....
100% agree. If we would be able to halve the bandwidth we could still serve 4-8 clients per AP but since we now could have 8 different frequencies in total we can make an 8 AP tower serving a max of 64 clients. By smart placing the AP's and distribution of the channels we would still have a undisturbed (interference) spectrum for the clients and would still be able to deliver 100Mbps or more to the clients.

100% agree. 1+
The larger antenna is the larger range + Radom
LARGE antenna LHG XL 60 GHz 550mm ( LHG 60Ghz 330mm )

http://24.multimediahd.pl/104-298-thick ... g-60ad.jpg

the range with the 600mm antenna would be over 8km

WirelessRudy · Wed Nov 21, 2018 12:53 pm

There is no point having MCS12 when we are limited by GigE interface (I'm sure current hardware can do it, but there is no point why go that high). Instead of making cheap 100Mbit/s CPEs, Mikrotik should release SFP+ version or something like wired 5Gb/s. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....
100% agree. If we would be able to halve the bandwidth we could still serve 4-8 clients per AP but since we now could have 8 different frequencies in total we can make an 8 AP tower serving a max of 64 clients. By smart placing the AP's and distribution of the channels we would still have a undisturbed (interference) spectrum for the clients and would still be able to deliver 100Mbps or more to the clients.
100% agree. 1+
The larger antenna is the larger range + Radom
LARGE antenna LHG XL 60 GHz 550mm ( LHG 60Ghz 330mm )

http://24.multimediahd.pl/104-298-thick ... g-60ad.jpg

the range with the 600mm antenna would be over 8km

Did you actually test that antenna?
There is a physical limit for the signal range unless you boost the radio output. Next you are going to tell us that with a 2 meter dish you can do 20km in 60Ghz. No way.... attenuation will still break down the signal. Maybe if you put a 500Watt booster on the radio..... everything has to be metal by then....

nimbo78 · Wed Nov 21, 2018 2:01 pm

dear MikroTik guys... plz add 60GHz products to your Wireless Calculator!

cohprog · Wed Nov 21, 2018 2:54 pm

Channel widths could be Hardware problem but no regulation problem!

The larger antenna is the larger range + Radom
LARGE antenna LHG XL 60 GHz 550mm ( LHG 60Ghz 330mm )

the range with the 600mm antenna would be over 8km

Right, 500hz width channel at 66Ghz+ frequency with 50+ EIRP for the win! Seems nobody even pretend to operate legal links here. Oh, yeah, I forgot: it's just for testing!

cohprog · Wed Nov 21, 2018 3:07 pm

There is no point having MCS12 when we are limited by GigE interface. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....

Oh, and by the way, you want 500Mhz channel and don't see why MCS12 would be useful, even with full duplex Gbps interface?

PS: Personally, I don't care about 100Mbps clients. I only use 60Ghz for backhaul. The LHG are mainly PtP anyway.

r00t · Wed Nov 21, 2018 6:30 pm

There is no point having MCS12 when we are limited by GigE interface. Also reducing modulation speed to get more channels would be nice, there is no point wasting entire current 2GHz channel to serve a few 100Mbit/s clients....
Oh, and by the way, you want 500Mhz channel and don't see why MCS12 would be useful, even with full duplex Gbps interface?

Sure, with narrow channel high MCSs would make a sense. But if anything, it seems to be a hardware limitation of used 60GHz radio chipset to only support 2GHz channels. So if anything, I expect some better interface options from Mikrotik. Wouldn't be surprised if they already have some prototype with SFP+ but it usually takes a long time for a new product to appear and even longer to be available...
Most flexible option would be SFP+ and one or more GigE ports. You could use GigE for POE and data for the building while connecting two 60G bridge units back to back with SFP+ DAC cable for a high capacity backbone link.

cohprog · Wed Nov 21, 2018 9:34 pm

A 1/2.5/5/10Gb ethernet interface would be great as it would work on most existing cables. (I have 10Gb over 30m of cat5e cable, not even cat6)

But even before such interface, I would like a better antenna with radome so that 1km+ links would be stable under heavy rain or snow. Under best conditions, the LHG60 is still at least 10dB worse than a Metrolinq 60-35. In fact, what I would really like is the LHG CPU inside a Metrolinq antenna hardware (with RouterOS).

WirelessRudy · Wed Nov 21, 2018 10:37 pm

A 1/2.5/5/10Gb ethernet interface would be great as it would work on most existing cables. (I have 10Gb over 30m of cat5e cable, not even cat6)

But even before such interface, I would like a better antenna with radome so that 1km+ links would be stable under heavy rain or snow. Under best conditions, the LHG60 is still at least 10dB worse than a Metrolinq 60-35. In fact, what I would really like is the LHG CPU inside a Metrolinq antenna hardware (with RouterOS).

A radome makes no difference to 60Ghz with rain. With snow it might. But not with rain. It's the amount of rain drops that consume and reflect the radio energy over bigger distances that make the attenuation too big.
I also have ML antenas and replace them all by LHG60G since they are so much more stable... and for half the price....

djvolt1942 · Wed Nov 21, 2018 11:18 pm

You are wrong as many say so,..

if there is no water on the dish and radiator [feeder] means a much more stable link, because we have a very focused beam.

The same is how you set up a shield for powerbeam, it works much better and more stable than without a shield/radom...

cohprog · Wed Nov 21, 2018 11:49 pm

A radome makes no difference to 60Ghz with rain. With snow it might. But not with rain. It's the amount of rain drops that consume and reflect the radio energy over bigger distances that make the attenuation too big.
I also have ML antenas and replace them all by LHG60G since they are so much more stable... and for half the price....

Many people reported that, with LHG, the signal stays down even when it stops raining until the antenna dry somehow. I noticed the same thing myself. I don't have that problem at all with the Metrolinq that are protected by a radome.

I don't care if the LHG are half the price: they are fine bellow 1Km but above 1Km, they lose signal under heavy rain while the Metrolinq stay at full capacity. That's not really a surprise: at 1.2Km, LHG have a signal around -55dBm while Metrolinq have a signal around -44dBm. That's a much bigger margin. I would prefer to use LHG because their throughput is much higher with small packets but I can't accept to fully lose the links each time there is some heavy rain, even if it's only for a few minutes. So for everything above 1Km, I am sticking with Metrolinq.