07-12-2018 10:17 AM
We just did a link of 19 km using 2 RFS antennas of 4 ft with 40 db gain using 2 Ubiquiti radios. We weren't able to get more than somewhere in mid -60dBm even though calculations tell that we should be in low 50s. As you can see we even applied more power to the radios, respectively 23 dBm instead of 18 dBm. With 18 dBm we should have achieved low 50s. The weather was perfect, humidity low and clear path, so no reason to get maximum result. Unfortunately we didn't get it and the team of aligners are very experienced with hundreds of links done for all major carriers. They spent over 4 hours on this link having two men at each end and as much as they tried they couldn't get anything over mid 60s,
They suggested that it is posibly that the radio doesn't actually put the power that is written in the software. They said that radios should be run with rezistive load on their RF connectors, while in the manual of AF-11X says that you assemble the duplexers and you power on the radio to configure it. That's what we did. We configured both of them at ground level and insure they connect when put close to each other and then we took them up the tower. Is that actually true. Is it posible that the radio RF connectors to be affected for running them without a rezistive load?
Another cause that I suspect is the waveguide converter from WR90 to N-type. We have RFS antennas, but I was sold some chinese converters made of aluminum from www.ainfoinc.com that even had the securing holes for screws smaller than the screw, so we had to enlarge them using a drilling bit. If such a precission device has imperfections at this level I am wondering if it doesn't have issue on other levels, so that's the reason I am also suspecting those.
Please let me know if the AF-11FX radio could be affected by running it without a rezistive load. Here is the link photo:
07-12-2018 02:16 PM - edited 07-12-2018 02:19 PM
My first guess after reading this is the alignment is off. I see bad alignment very often.
But, since you bring the other equipment into question, I'm not sure.
I have seen several people link the AF11FX without duplexers or antennas, so if you are asking if you damaged the radios by linking them up, I sincerely doubt it.
To my knowledge, I have not seen an RMA on the AF11FX due to a low signal.
07-12-2018 02:53 PM
So basically I need an official reply from Ubiquiti that the radios are OK to run without resistive load without breaking down or causing an RF port issue that could lead to lower performance in the future. If Ubiquiti can confirm this I will go ahead and look into replacing that other part and if that doesn't change the situation it is clearly a poor alignment and in that case I can bring it back to the alignment team, but please consult with other UBNT employees that are knowleageable in the product and can confirm that AF-11FX radios can run without having an antenna connected to their RF ports and brunning like this will not cause any harm on their RF ports, or affect their effective power emission in the future.
07-12-2018 03:10 PM
I am looking basically for an engineering response as a certainty. I cannot go back to this alignment guys and tell them that I have doubts that such operation will cause harm to the radio. I have tot have a correct and precise answer as I am on the hook for several thousand dollars for a job that they pretend they executed perfectly while I have doubts about it.
I hope you understand my concern. Of course I will check also the other part that I suspect, but I have to rule this one out so please have one of your RF engineers involved to get a pre ise answer. The other option would be just to RMA the products and have them replaced. They are only 2-3 months old.
07-12-2018 04:58 PM
I see no evidence that you have anything to RMA here. If you are using 40dBi antennas (I don’t doubt that a 4 foot aperture would create that sort of gain) then you are looking at something less than a 1.5degree 3dB beamwidth. 1.5 or 2 degrees of error on both sides could impact link budget by 6 to 10dB. These are very difficult antennas to aim and the structure holding them would need to be very robust. That is the most likely source of error here in my opinion. Secondly you refer to some sort of OMT/N-connector adapters that appear to have some issues/unknowns. Without the ability to measure the actual return-loss of these adapters it is difficult to ascertain how these may impact performance.
As far as running radios without loads; it is generally a very bad idea due to the possibility of reflecting too much energy back into the high power amplifier stages, which can lead to failure of those devices. In this case if you were running the radios with the duplexers then the return-loss the duplexer presents to the RF amplifier certainly helps to protect things. This radio also has an automatic feature built-in which will reduce the bias applied to the high power RF stages if the radio senses too much power being dissipated as the result of a bad mismatch. The radio is very robust and is not likely to have suffered permanent damage. Even with all these caveats, it is never a good idea to run any type of transmitter without a proper load. You also cannot directly connect two radios together directly without appropriate attenuation that simulates real-world path-loss. If they were directly cabled together then you most likely would have damaged the radios.
07-12-2018 05:54 PM
“These are very difficult antennas to aim”
Sorry, but this just made chuckle since we designed and installed literally thousands of 11 GHz links over the years and all of them before 2006 were using 6’ and much larger antennas. Your comment on tower twist and sway is dead on, especially with larger antennas.
07-12-2018 08:21 PM
When you say that connecting the radios directly with cables would have damaged them, does it mean that they wouldn't work at all, or they would just performed under specs, like mine, respectively put out less power than they should.
Thank you for clarification.
07-12-2018 08:29 PM - edited 07-12-2018 08:31 PM
@keefe007 @UBNT-GaryIt is very hard for me to convice a veteran installer that installed hundreds of links over the last 30 years of activity in RF world that I am right and he is wrong, repsectively his alignment is wrong, so I either prove him by replacing these radios with 2 known that are in good operating condition, or otherwise I have no way of commenting that his alignment is off. He even brought the subject with a person from Hutton Toronto who has no doubts in his aligning skills and that person didn't even question his skills and agreed that the radios might have an issue. Hutton does an advanced RMA for their customers, but I did not purchase the radios from Hutton. I got them through Streakwave in USA.
At this point I don't know what to do. I would like to test 2 new radios that are know in good condition to convince myselft who is right and who is wrong. At least that way I can rule out the radios from the equation. If the new radios perform the same I just return them as it is clear the radios are not an issue and I will keep my existing ones. On the other hand if with the new radios I get a better llink it means that my radios are defective, and at that point I would sent them back. Would that be posbile @UBNT-Gary ?
07-12-2018 09:00 PM
This is pointing to an alignment issue. The only way to know for sure is if your installers used an AAT (antenna alignment tool). Four foot dishes are difficult to align. A half turn of the wrech can miss the whole main lobe.
Lets rule some things out.
What's the exact model of your dishes?
What kind of WR90 adapters did you use?
What kind of jumpers did you use? 11 Ghz jumpers are hard to fine. If, for example, you used something like LMR600 it would be incorrect becasue it has a frequency cutoff at 10.3 Ghz.
Please also post some pictures of your setup.
07-12-2018 09:29 PM
@keefe007 I told my guy about the AAT, but he is at around 60 years old with over 30 years of experience in RF world and he told me he does not need an AAT as he tried it and it can put you off in certain situations. To be honest that would probably be my next stept, respectively to rent an AAT from ATEC and try to do a better alignment myself. You are right with regards to the half of turn. The guy knows that and he literally went in quarter of a turn or even less when he came to this numbers. He was around these nunbers for over an hour tewaking up, down, left, right, both antennas but however he tweak them he couldn't get more out of them.
The exact model of dish is RFS UXA4-107DC . I don't use the original RFS WR90 adapters, but a "compatible" chinese adapter from ainfoinc.com . Here is the adapter:
I got the jumpers from Data Alliance and they are rated up to 12.5 Ghz. they are LMR200. here is the link for the cable of 3 ft I purchased: https://www.data-alliance.net/n-male-to-n-male-cable-3ft-6-5-ft-double-shielded-indoor-outdoor/?src=...
I don't have photos from up the tower yet the tower is CSA-37-S13 certified, meaning is a carrier grade tower which has a calculated load and at maximum wind speeds in the area should not sway more than 1 degree with maximum load on it. I am way below maximum load. I bearely have half the load the tower accepts.
Let me know your opinion.
07-12-2018 09:50 PM
I recall you mentioning that you had/have two links, is the other working correctly ?
If so what is different hardwarewise ?
If not are they using the same waveguides ?
07-12-2018 09:59 PM
No, the other one is even worse, but over there we have that antenna issue with the alignment of the brackets. I sent you the photos of it. The 24 km link using a 6 ft and a 4 ft bearily gets into mid 70s Here is the link:
On this one we know we have a problem with antenna or at least antenna bracket which when tighten pulls the antenna in a different plane. I know it is fixed in 3 points yet 3 points define a plane and that plane should stay the same when the bolts are losen or when you thighten them. In our case when we tighten the securing bolts it puts the antenna in a different plane and the whole alignment goes haywire...My concern though for now is the link that both antennas seem to be OK (with the exception of the CPRG90 to N-Type connector on which I have some doubts). Why isn't that one performing well?
07-13-2018 07:15 AM
I am not familiar with those "compatible" waveguide adapters. There are two separate adapters and each is feeding a separate orthogonal polarization on a waveguide feed?
I can’t really give you a definitive answer on whether or not you damaged the radios because I would need to know a lot more information on how they were configured (power levels etc). Since it appears you did directly connect the radios together with cable then I would say there is a high probably that you damaged something. In this situation it would most likely not be the TX power amplifiers but instead it would be the low noise RX amplifier that would suffer. This would result in a radio with very poor RX performance and this would not be considered eligible for an RMA since it would be considered owner abuse.
07-13-2018 08:54 AM
Are you 100% sure you have your duplexors set right?
Even though those cables say that N-connectors support up to 12 ghz doesn't mean the pigtails you bought are tested and designed for 11 ghz.
Times Microwave, experts in this field, said they would need to custom manufacture and test jumpers for me in 11 ghz.
07-13-2018 09:11 AM
There are so many unknowns flying around here it is brutal, it might be informative to buy a pair of the ubnt dishes and setup the radios with them on a 'test range' being careful of course not to nuke somebody else' link while testing.
My gut says waveguide is suspicious, following that the installers, if the 6 foot dish is distorting then it is being done wrong, if the elevation is free to move when the azimuth (u-bolts) are tightened there is no way it can exert any force on the dish shape, if the mount is as goofy as you describe it may create an azimuth rotation if not done carefully.
07-13-2018 09:46 AM
@UBNT-Gary I did not connect them directly with cable but I did operate them on the bench without resistive load and my alignment guy says that by that alone the radio might be damaged. I also agree with some other people's opinion on the forum that radios have no issues and it is an alignment problem, but I have nothing in my sleve to prove that, respectively I would have to have a pair of radios in known good working conditions, to install them on the tower to prove that nothing changes in the link. You have to understand that this guy is known by all major wireless providers like Alliance, Hutton and so on and recognized as one of the bbest aligners in Toronto area, and he says that based on his experience and his tools the antennas are properly aligned as it is not the first time he does such alignments. He insists that radios are the problem and even one of his Hutton representatives agrees with that. I am literally in the man in middle and the one that takes the blow.
I am willing to pay for shipping back and forth for these replacement radios to have them for a day, put them on the tower to prove what needs to be proved and based on results we will determine who is right or wrong.
I have to pay this guy a considerable amount of money, as as I mentioned before he is one of the best in Toronto area and his rates are the same, and I don't want to pay him for a well done job only to realize he rushed it and he did a mediocre job, which for now I have no way of proving.
Please let me know if you could help me to perform this test and what steps need to be taken in this direction.