MT Base > Ubnt CPE reassociating issue

RouterOS Wireless Networking
1

Hi all

We have a RB433 in AP mode connected to a 16Dbi Sector serving 2 customers. Customer 1 is connected via a RB711 in a stationbox and is very stable. The second customer connects using a Ubnt NanoBridge M5 in station mode and this is where the problem exists

On the AP it is showing the Ubnt unit is reassociating almost every 5 minutes!

The base is configured as follows:
802.11n dual chain 20Mhz.
Protection mode is disabled and WPA2 encryption is used (WPA has also been tested but fault remains)
Signal at base is -68/-70 and CCQ is 90/95
HW retries is set to 15
Link is 1.5KM max
Frequency is 5785Mhz
RB433 is running ROS5.24 firmware 3.02
NanoBridge is running XM 5.5.3 build 14763

Anyone got any suggestions or seen this behaviour before?

Thanks. Jon

2

So this has never happened to anyone else on here?

3

Don’t use TKIP and update UBNT to newest 5.5.4

4

Hi Deac and thanks for the advice

Device is now updated to 5.5.4 but still suffers the same issue

I did notice it was using WPA2 AES so i downgraded it to WPA AES but this has not stopped the drops.

Is there anything else that can be changed to stabilize the link?

Thanks. Jon

5

I haven’t had any issues with 5GHz UBNT CPE connecting, but I’ve had issues with the Airgrid M2. One batch absolutely will not see at least 2 of my MT tower APs.

I had lightning damage at a customer’s location, just replace the feedhorn and keep going, but the new batch will not connect.

While I was troubleshooting the problem I had the AGM2 feed on the dash of my truck running off of an inverter on the way to the tower. At 7 miles it would not connect, could not even see the AP. At one mile from the AP it connected while sitting on the dash going 45mph.

I had another that connected at 20MHz and refused to connect at 5MHz width. Considering I had a large percentage of my CPE money tied up in units that now randomly refused to even see the tower, that put me in a pretty significant bind for a month or so.

I still haven’t figured out the cause and it only appeared to affect the M2. One AP was 802.11n 20MHz, the other was 802.11g 20MHz, then changed to 5MHz. Everything was configured identically. I even made a config backup of a working unit and restored that to the new unit. Every CPE had the same config, I set it up once, made a backup and uploaded it to the new units as I got them.

I have not had time to dig into the problem, though I have a perfect location to try, just no time.

At first I thought it may be signal related, maybe the M2 lost 5-10 dB of sensitivity, however the second time I ran into this issue the customer is about one mile out with absolutely perfect line of sight to the tower.