We have a site that has been serviced by a 433ah with an XR2 for years. This had a 30ft run of LMR400 up to the 11dbi omni. It services about 60 Clients, and has worked great.



We saw some interference or what could be water in the coax. We have been elminating as much coax as possible over the past 2 years. So we took this opportunity to replace this with a Metal 2.4Ghz Groove.

However, this did not resolve the problem. Signals went from a range -50 t0 -80 and increased to a -15 to - 55. This is kinda crazy. I am not sure why there would have been such a huge jump in the -Rx signals of the groove. I am afraid something else is causing these weird signals or the groove is not reporting the sigs right.

This is a very old ap location, and we used to use Demarc brand cpe's. So about 85% of the CPEs are demarc with the other 15% mikrotik. I used the access list to only allow the mikrotik cpes to connect. Problem still existed of ping times from the ap to the cpes 700ms+. Even with no traffic. I switched from 802.11 to nv2, and then had PERFECT times, 1-12ms consistently with good traffic running. Once its switched back to 802.11b/g or b, the problem returns.

We have had cpes go bunk in the past and would go into compression and spew out huge amounts of garbage RF. So this is kinda where I am leaning.

Now, when all of the clients are allowed and 802.11b/g is active, the problem still exists. At random times, the problem completely disappears in just 1-2secs, and ping times go <10ms for all the clients, and the interface starts passing around 3Mbps. This will last for less than a minute, and then the problem returns.

I have my methods of identifying the solution, but I would love to hear what others would do. It's been a long, cold, and windy day on towers and am burnt out. But i can't relax with this problem out there.

Please give me some feedback, Thanks!

Adam

maybe try finding a better channel using spectral scan.
are all your CPEs 802.11g capable?
try fixing on b only or g only see if problem returns

Steven

maybe lower your TX power.

Thanks for the replies, but there is definitely something wrong with the groove. I have indeed tried both of those solutions.

Turning down the power on the groove will not affect the rx signals, it doesnt make sense that one interface would read them at a 40db difference. However, I did try turning it down, im trying all different things to find out where the bug is.

Note to Mikrotik: NV2 solves the high ping/no traffic problem.

You mean you are seeing something like this.. (see attached) In the space of 1 minute the signal as seen from the client end fluctuates up and down by 13dB.

Before I installed it at the weekend ( now unable to even mac-tel or ssh into it) everything was fine with a RB433ah + XR2

Yes, similar. And the clients are disconnecting in groups simultaneously. Mikrtotik clients stay connected the longest.

I am climbing here in an hour or so to replace it (I downgraded from 6.0rc11 to 5.1 and it never came back).

That almost sounds like a bit of the shielding on the coax is *just* touching the copper or something in the Groove is failing. Not ruling out water either...

HI

Its directly connected to the antenna, and it isn't raining.

I missed the lack of coax, skimmed it in a hurry.

That really doesn't leave a lot of points of failure. Moisture/condensation in the antenna, in the connector, in the Groove or the Groove is failing.

Sounds more like the Groove.