RSSI chain imbalance on multiple SMs

Thank you for the reply. I had swapped radios and im pretty sure I didnt mess up reconnecting but I will double check them.

Its easy to do, I have done it. changing a radio is at the best of time a complex procedure, now add cold!

Yes exactly, nice putting everything together on the ground but yes such a pain up on the tower lol… thanks again… I have seen some posts on the forum that the second chain being off is a realignment issue but since I did swap the radio and I noticed the chain issue so its pointing toward cables.

a loose cable or a damaged connector or even an over tightened connector can cause RF loss.
There is too much that can cause similar issues.

connector tightness is hard to teach over forums, but is best described as finger tight plus a little to ensure it wont come loose. I use 3 to 5 inch pounds on connectors smaller than half inch diameter.

damage with the RP-SMA is fairly hard to do but not impossible, check the center pin and receiver for obvious damage.

1 Like

Will do… I connected them as you described but will double check.

Also - you can’t assume that ‘CH0’ and ‘CH1’ in the software (eg in eAlign) are necessarily related to CH0 and CH1 on the radios when you’re trying to diagnose what the issue with chain inbalances are. I reported in April 2020 that the Chains are crossed from the hardware to the software. There was a couple jokes made about it, but nothing more than that. :frowning:

Click this link below to go to this thread from April - there are pictures if you click this link.

Okay, thank you… the issue is I have one SM out of six that has no chain imbalance issue but the other 5 are ALL at -75 for Ch1… so is this a alignment issue… Im not too sure, could it be the RP cables, sounds more likely since my chains were not out of whack before I swapped the radio. Thanks!

This may be a stab in the dark but could it be i have my AP software 4.4.3 and its inactive bank at 4.5.6?

Also - as per my post on the chains… when you do climb and start looking at cables/pigtails DO NOT ASSUME THAT CH1 is CH1. You’ll have to disconnect each pigtail, and do the “process of elimination” to be certain where the issue is.

https://community.cambiumnetworks.com/t/ealign-chain-0-chain-1-weirdness/66368

Okay, let me get this straight… on the epmp 3000 radio from right to left is ch0,ch1,ch2,ch3 correct? on the epmp 3000 sector from right to left is ch0,ch1,ch2,ch3 correct? not sure how one could mess this up…unless we are talking about cambium mislabelling things?
On a epmp 3000L is the H ch0 and the V ch1?

I encourage you to test for yourself.

Cambium and others will say that the Chains don’t really matter (and they kinda don’t… connect your ePMP3000L CH0/CH1 to any dual pole or dual-slant antenna, in any order, and it doesn’t matter if they are crossed or not. The radios will figure it out, so that part doesn’t technically matter. That part is kinda accurate.)

BUT - were it does matter is that it’s important to get them correct, so when we were looking for what was wrong with CH0 signals (and we assumed that meant a problem with the CH0 connector or the CH0 pigtail or the CH0 feed on the sector) and it turns out the CH1 connector on the AP, is displayed as CH0 signals.

No one followed up when I reported it – just a couple jokes made and that was that. When I test and report, I’m not doing it just to complain, or to be grumpy about something… I’m actually doing it to help make things better. Anyway, I would encourage you test yourself (disconnect the pigtails one at a time and see what results you get) and report your findings / questions to Cambium as well.

1 Like

Sounds goo, I will thank you. So what was your resolution?

I have no resolution. I can’t make the Firmware display match up with the actual correct hardware connectors.

Physical radios do not align with the data chains. It is better to think of the physical ports as radio1 & 2 (3 & 4 if you have a 3000 AP).

Chains are data streams on a radio and are not fixed to a particular radio as they are subject to the channel width. Eg a 20Mhz channel has two 10Mhz sub-channels with one chain each but a 40Mhz channel has two 20Mhz sub-channels with two chains each.

Polarity of the cables does have an affect on the signal chains at the SM but this is more important for back to back APs. Though it is more likely that the radio is out of alignment slightly. I am focusing on the cables as a good possibility as everything was working properly prior to a radio swap.

Is this a reply to me? Or a reply to DigitalMan2020?

If you’re replying to me – for my 2c, I think it’s pretty reasonable that people will expect that if the connectors are labeled CH0 and CH1, that they expect those to match what eAlign displays as CH0 and CH1

If they are supposed to be backwards, or if they are supposed to be arbitrary, then why didn’t Cambium label them A and B, or label them V and H, or label them Fred and Joe? Don’t label them CH0 / CH1 and have software also list CH0 / CH1 and then be surprised when WISPs expect CH0 to mean CH0

OR maybe Cambium can comment on this issue?

Ninedd,
It was kind of to both but more for digitalman2020’s sanity.

For the record, the ports were labled h and v at one time. Not sure why they changed, but they did.

To be honest as long as both ends are the same then it wont matter which is which. Better, more intuitive markings like on antennas would be helpfull though.

Do both ends even have to be the same, other than for troubleshooting as ninedd points out ? I agree that it seems like it should matter but I’m pretty sure I remember Cambium saying it didn’t matter. I asked several years ago because I had been told Ubiquiti antennas were backwards from ePMP and that you needed to cross the pigtails when using ePMP on a Ubiquiti antenna and we were doing a lot of ePMP on Ubiquiti antennas and Elevated Ubiquiti radios at the time (also it seemed you would have to do that on almost any antenna because at the time most/all 3rd party antennas were the same as Ubiquiti as far as I knew) but pretty sure when I asked Cambium about this I was told it didn’t matter , the radios either didn’t care or could figure it out.

The reason for crossing the pigtails on UBNT antennas was due to direction of the rf elements in respect to polarity of the propagated wave. This polarity does matter as you do not want your received wave to be out of phase with the receiving antenna. Most high quality antenna manufacturers follow the same standards which are reversed of WIFI which UBNT follows.

With mimo, the link is created with one radio chain and each end will determine the best radio antenna pair to receive the signal (notice not transmit). Basically the two most in phase with each other. The second chain is on the remaining radio antennas which may not be anywhere near the same phase and still make a low quality link.
Keep in mind that this is completely different from OFDM-MiMo which truly does not care and uses all radios and antennas at the same time on all chains.

Personally I would like to see the epmp line use 5mhz ofdm chains and we select the center frequency and number of chains ( width) with higher modulation schemes that are determined by the 5mhz single chain. More bps/mhz is always good.

1 Like

Thank you to everyone’s comments/advice on this thread. So I have not climbed yet to try the RP pigtails but I did open a ticket with Cambium and explained my situation of a 10db chain mismatch on all of my 6 SMs that are attached to the AP. So they said it could very well be the cables between radio and sector antenna so they suggested that I try the AP in ptp mode and then I can see the ealign on the AP side to the one SM that would connect. The chains for the SM on the ealign on the AP were perfect -63 and -63… put it back to ptmp mode and logged into the same SM and the chains had a 10 db mismatch. Cambium said they are looking into it.

@Douglas_Generous @ninedd @brubble1

So we climbed the tower, disconnected and reconnect the 4 RP cables and issue went away… so @Douglas_Generous was correct! When I swapped the radios on my first climb one or more must not of been seated correctly. I thank all for their help/advice! This a great community!!

3 Likes