Apex not responding to IO

musaabi · Oct 11, 2025

My apex was working fine then had a power loss. When I turned it back on, responses from the breakout box wasn’t working and it registered it under a new address. Renamed everything and verified the apex is registering the switches when they’re closed, tiles for the switch show open -> closed. The problem is the programming is not responding to the switches. For example my Refugim light is not turning on via the switch it’s programmed too. This is true for every switch that’s programmed such as the maintenance switch or feed switch. Any ideas?

smitten with ocean life · Oct 13, 2025

have you reached out to apex? sometimes for me, when it acts wonky, i just have to give it some time and it starts working again. might need to shut it down and restart

Reefer Matt · Oct 13, 2025

Try this in the task menu. Sometimes a module gets lost during a power outage, etc. and gets reassigned an address. If the module isn’t listed in the task, you may have to delete the whole module from Fusion and reprogram it.

musaabi · Oct 13, 2025

smitten with ocean life said:
have you reached out to apex? sometimes for me, when it acts wonky, i just have to give it some time and it starts working again. might need to shut it down and restart

Not yet. But I did restart it from Power as well as thru software and forced update firmware just in case. Next step is reaching out to support.

musaabi · Oct 31, 2025

Had a tech support call. The issue is with ghost! No really. Basically, the naming convention is still stored and not erased. This means when I renamed the switches to the exact same names to keep programming the same, Apex is referencing the old address and name. So the inputs are changing, however, the name it’s referencing in the programming (the ghost) is not.

The solution is one of these steps:
1) Factory reset the Apex.
Or
2) Rename the switches. For example, my ref_sw that was to turn my refugium on, I changed to SW_ref and changed the programming to reflect. This worked!

Apparently, it’s a minor issue that doesn’t happen often, hence no support article about it (until this post).