I do have a similar issue. The EVCS is communicating with the Cerbo as it shows actual power flows and battery status. But the cerbo does not get information from the EVCS. Any Node Red flow is not working (power reduction to save main fuse). See picture.
What is wrong?
And then suddenly communication returns and node red is working?
I had bluetooth switched off, but when when no wifi is available, how to access the unit?
I had the same problem with EV disconnecting, when using WiFi UNIFI. I tried to deploy an old wifi device about 10 years old and since then about 10 days without a dropout.
Ah ok. Had to update a lot of UniFy 7 Pro/Max hardware in the past months. Lots of improvements. Even yesterday. Your issue could be caused by the used channel. Please try some another channel to see if this helps.
I installed a TP-Link EAP110 on the wall, within two or three metres of the EVCS (clear line of sight), and reconfigured the EVCS to use that connection to the network.
It seemed to clear the problem. So I started a ping test at about 9am today. Eleven hours later, Iām consistently seeing a 1% packet loss to the EVCS. (The test will run for about twenty-four hours total.)
For my purposes, this is good enough, but it does suggest that the EVCS - for whatever reason - is not reliably holding a connection to the wifi access point. Iām speculating here: itās possible that this, combined with having multiple visible instances of an SSID in the case of a mesh wifi setup, is causing the problem.
Iām happy to do more investigation to try to help Victron nail down the problem, if it will be of help. For me, itās good enough now, but I dislike leaving a problem - even if itās a relatively small problem - hanging.
Yesterday, when I needed to head out somewhere, I unplugged the car - and noticed that the EVCS was showing the W-104 warning again.
That blows my hypothesis out of the water. I donāt know where the issue lies - but itās definitely not (solely?) due to the use of mesh wifi, as the TP-Link station is not linked in any way with the other base stations I have. (Well, okay, if you want to be pedantic they are on the same network segment, but thatās not the sort of link I mean.)
I grabbed a wifi analyser (WiFi Scanner, by Lizard Systems - hooray for trial periods that cover my needs) to look into that specifically. Itās on channel 6, and was running at a bandwidth of 40 MHz; Iāve reconfigured it to 20 MHz. Channel 6 is relatively uncongested - thereās a few networks from neighbours, as youād expect, but channel 6 is the one showing the least sign of competing traffic.
My EVCS showed the W104 pretty consistently after I made a change to my available Wi-Fi networks - and continued to be unreliable after deleting and rescanning the new Wi-Fi ssid (from some TP-Link power line adaptors).
What worked to restore Wi-Fi connection stability for me was to switch off DCHP in the EVCS settings and type in a fixed IP.
The connection has been rock solid for a few weeks since then. All good
Thank you so much for this - this appears to have resolved my problems as well. Itās only been a few days for me, but I was regularly seeing W104 - havenāt seen it since I switched to a static IP. (No regrets about getting the extra base station for the charging station, though; segregating all the IoT stuff is on the list of things to do, and I need the base station for that purpose anyway.)
Same issue here, dropping wifi a few times per day.
I am using the new EVCS without screen and my Wifi network is a TP Link Deco X75 Mesh network. None of my other devices (45 total wifi clients) have any issue, just the EVCS.
I struggled with my EVCS dropping off my dual Orbi mesh network for years. I tried everything including firmware upgrades, channel tweaking, interference eliminating, spatial changes, a new access point, static IPs, scanning modes, DHCP tweaks, countless resets, reboots and re-flashes it still kept happening. I scoured every forum here and elsewhere. The only thing that finally fixed it was setting up a dedicated 2.4 GHz network using an old netgear Wi-Fi router, with a wired link to my Orbi mesh network. The EVCS has been rock solid ever since. Iāve got numerous ESP32-based devices in my house, most of them are absolutely fine with my dual band Wi-Fi set up, but a handful of them have displayed exactly the same behaviour and only work reliably on a dedicated 2.4 GHz network. I think it must be an inherent limitation of the hardware inside the EVCS, so echo calls above for future implementations to include wired connectivity rather than just Wi-Fi. Such an expensive bit of kit should work flawlessly out of the box rather than relying solely on a Wi-Fi connection which is not 100% robust. Iāve been very happy with 99% of the expensive Victron gear that Iāve bought over the years, but the Mk 1 EVCS, although well integrated in many ways, is not on a par with the rest of it in terms of reliability (or build quality TBH).
Thing is, Iāve done the dedicated 2.4 GHz wifi using a TPlink base station. The EVCS is still dropping off, even on that dedicated network connection.
What Iām finding is that usually, if I wait, itāll reconnect. Which is okay, except when Iām trying to charge the car with excess solar. The moment the EVCS drops off the wifi, it assumes thereās no excess solar and starts the 30 minute countdown to stop charging (and will charge at the minimum rate during that period.)
The only thing that makes this tolerable for me is that Iāve gone with an electricity retailer (Ovo Energy) that gives me three hours of free electricity from the grid (11am to 2pm), and I can set the charger to do the full 32 amps (~7 kW) during that time. But this is not why I bought the Victron charger; if it were the solution I wanted, I could have gone with a cheaper EVCS from another manufacturer and skipped the whole ācharge with excess solarā feature.
