There isn’t really a H and V to the radio as it has no way to know, nor does it care, how you have the antenna oriented. You could connect it to a antenna that is oriented H and V or you could connect it to one where the pols are tilted 45° or you could rotate your H/V antenna 45° .
I’m not sure it even matters if both ends are the same with ePMP as I believe the 2.4Ghz ePMP sector is slant 45° while the subscriber radios are not and cambium said the ePMP radios have magic sauce that allows them to figure out what is what or not care or something.
I am sure that those who say there is no such thing as H or V are less experienced in field work. Because people here most say that there is no way to differentiate between H or V. And I am new to the Cambium equipment problem, but I have tried many radios from other Brands. And no matter it says Ch0, Ch1 or something there must be a sign to know H or V if we want to install using an external antenna. Ummm But i apologize for my rush to ask here, because from the cambium side it turns out that it has posted to differentiate H or V between Ch0 and Ch1.
This is the last time I’m going to comment on this. From a technical/RF standpoint it doesn’t matter which port/polarity you choose. From a reporting/UI/troubleshooting standpoint it’s helpful, but not required, to try to get your ports and polarities matched up.
ePMP3000L utilizes benefits of 802.11ac standard. You will not see them when using with 802.11n subscriber modules. ePMP1000 and Force 130 are both 802.11n.
While the 1000 might allow more clients technically, in practice you’d never want to put that many clients on it, and while the 3000L allows less clients, you could most likely put more clients (assuming all 802.11ac clients) then you could realistically put on the 1000.
The F130 (and all Cambium 802.11n based radios) can connect to the 3000L. The 802.11n clients won’t be able to take advantage of some of the features that the 3000L and other 802.11ac radios have to offer, most notably 256QAM.
Have had 3000L at over 100m and as long as the cable is perfect and the surge protectors aren’t cambium they drop a few ethernet packets here and there but work ok. Make sure your cat cable is 110% Perfect. If you have surge protectors (especially cambium surge protectors) try bypassing them and see if that fixes. You can also try turning off Configuration > Network > Smart Speed. Also check you power supply / power source, Anywhere there is a patch cable or surge suppressor is suspect.
Or the the ethernet port could just be bad.
The 2000 and 3000 are very finnicky about the ethernet cable/connectors so the cable / connector has got to be absolutely perfect.
20M is not a long cable, the spec for cat5e at 1gpbs is 100M and several of us are using cables that are longer!
As Brubble has pointed out, Cables are your #1 culprit after smart-speed. Turn smart speed off, and most definitely bypass the Cambium SS equipment if installed.
The type of cable only matters if you are using CCA (Copper Coated Aluminum) as is is not suitable for any use in WISP/Server environments!
Your shielding needs to be single point grounded, make sure you did not make any ground loops. This is more important at longer cables but it is a good practice to keep.
Good quality crimp ends using a very good quality crimper. Ebay can be your friend if you know what to look for, but nothing beats Leviton connectors and Kline tools for quality.
For the time being, make sure you are using firmware 4.6.1 (most stable in our network) and make sure your tower switch is better than an $25 Walmart junker!
Hi,
I know all you have write.
Cable is ok beacuse if I change 3000L with a 2000, all works at 1G full duplex.
If I replace 3000L with a 3000… problem is the same.
If I replace 3000L with a MikroTik NetMetal, speed is 1G full duplex.
I Think the ethernet chips, or transformers, is poor than epmp2K or MikroTik devices…
Anybody can confirm ethernet negotiation problem with cable over 20mt and a little RF environment?
Thanks
I can not confirm a negotiation issue with the 3000 (e3k) and 3000L (e3kL) on cables 20m and longer. We have several e3kL radios on longer cables working just fine at full gigabit (auto negotiated) and a couple e3k radios at 68m cables just fine. Granted though that these are powered by a Packetflux Sync-injector and connected to a Cisco ws-c2960s-24ts-s (modified for dc only use) and the radios are running firmware 4.6.1 (4.6.2 and 4.7-rc7 had other issues for us).
Since you are adamant that this is a radio side issue, open a support ticket. But be aware that you will need to verify the cable with a proper analyzer and use the cambium poe brick to a good quality network switch. Only support can verify a problem radio.
Me, personally, I would be looking at Smartspeed to disable it, checking my cable (since you claim it doesnt matter which e3k radio, you seem to have a spare) by making another one and bench testing using as much of the same equipment that you have at the tower. It is important to actually have a test setup that is as close to identical as the tower setup to compair against. Be detail oriented and record what single thing you changed and the effect. If you change everything at once then you will have no clue as to what to look for on the tower.
Look at:
cable - the wire itself can be good but how about connectors and the actual termination? just because one radio works doesnt mean another will. So you need an apples to apples comparison not that a ubnt or mkt work but the e3k will not.
configuration on the radio
configuration on the switch
power source injector
surge arrestors
grounding (avoid grounding loops)
the switch itself! some switches use block magnetics, which means that you have several ports that are sharing a group of magnetics. you need to move to the far end of the switch to guarantee your on a different set of magnetics. Also, use an enterprise managed switch not a sudo managed switch.
Also, be more detailed when asking for help. W can only give you suggestions based on what you tell us. Simply asking us to confirm your suspicions about an entire product sub family line is not going to get you very far since a simple forum search would yield that. And be more open minded about what we suggest since your answers will direct the suggestions.
Hi - I realize you’ve done the common-sense testing - if you replace a 3000L with a 2000 and it links at 1GB right away, then it feels like it’s a 3000 issue, and not a cable or router issue.
But, at the same time… I can tell you that we have lots of 3000L and 3000 MU-MIMO AP’s on 100 - 140 foot towers, and on 200 foot grain elevators, with no ethernet issues.
So, I don’t really know what to suggest to help exactly - but I wanted to chime in and let you know that not everyone is having issues over 20 meters. We generally have no ethernet issues.
