IPv6 Experiment - bridge problem

Hello everyone, I encountered an issue yesterday. The network structure consists of two bridges: Bridge A and a Station Bridge (Bridge B), which are configured for one-to-one bridging. Bridge A is connected to the operator's PTN (Pseudowire Technology Network) via an optical fiber, while Bridge B is connected to the operator's BBU (Base Band Unit) through another optical fiber. This entire network is used for transmitting telephone and Internet services. It has been confirmed that the two bridges are identical in terms of device version and model.

What makes this case special is that the wireless network and optical fibers are functioning normally. However, as soon as an external antenna is connected to Bridge B, the network cable loses connectivity. You can see that the local area connection on the computer keeps attempting to restore the connection, but whether it can succeed depends on luck—sometimes it works, but 90% of the time it fails.

I have already used the elimination method to rule out issues with the devices, network cables, power adapters, POE (Power over Ethernet), and power supplies. Currently, the only component I cannot eliminate is the antenna, as I don’t have a spare one on hand. Although the wireless link and optical fibers at this station can be used normally, the operator has reported intermittent outages in the wireless service. Does anyone have any solutions to this fault?

With "bridge", do you mean what is named a bridge in MikroTik terminology, or do you use this as a name for a wireless dish?

It refers to selecting the bridge or station bridge mode for connection in the device's working modes. I have roughly identified the problem: when the device is operating, if an optical fiber network is plugged into the SFP1 optical port, the Ethernet1 electrical port will encounter such an issue. Simply unplugging the optical fiber from SFP1 will immediately restore the Ethernet1 electrical port to normal. Reconnecting the optical fiber to SFP1 afterward will keep both ports functioning properly. Finally, I used the VLAN feature to separate the SFP1 and Ethernet1 ports into two VLANs (VLAN10 and VLAN20), and the previous problem no longer occurred.

Maybe something with Spanning Tree.

Impossible to say from remote without specs of the devices, their config, etc.