I operate two Fibaro Dimmer 2 (FGD212) together with an Aeonlab Gen5+ Controller (ZW090) and the ZWave JS integration.
Since several month, I notice that the lights dim up a bit at full hours. “Mystically”, as nobody has commanded them to. As one can see in the screenshot, the brightness settings remain constant, but the power increases, along with the actual brightness.
Sometimes, it does not dim up, sometimes it also does at half hours.
The dimming takes about a 1 minute and is performed in several steps. If I actually set the brightness during this period (dim down), it proceeds with automatically dimming up.
What I’ve tried so far:
actually disconnected the controller from the HA host system → still the same effect at full hour!
Install a original Fibaro bypass at “Esstisch” - had no other effect than making flickering a bit worse
played with power/energy reporting parameters (interval, threshold) → no effect
completely re-setup my ZWave Network, factory resetting every device and the controller, adding only the dimmers and an additional RGBW Dimmer → no effect
Does anyone also experienced this phenomenon?
What could I try next? Change hardware?
Do these Fibaro’s have a debug log like a Shelly does? Can you enable it for a duration?
Also are these no-neutral fittings, ie only have Live/switched live and the lamp (and the bypass)?
To my knowledge, the Fibaro devices do not have an integrated log system like, e.g. my Shelly 3EM. I will try the HA debug logging for the ZWave JS integration.
The dimmer “Esstisch” has no neutral, but a bypass (also worked without), whereas the dimmer “Luster WZ” has neutral.
I have seen dimmers jump sometimes due to voltage changes on the input, some of the no-neutral ones can do this as they don’t have a neutral reference voltage, but if you have a neutral on one and still have the problem then I’d suggest opening a ticket with Fibaro support and see if they have any input?
Hey Ronnie, very good points. I’ve also suspected voltage changes but a) had no means to measure and b) have not seen any effect on other devices.
But now I have a Shelly 3EM at hand and I was able to record this:
However, the dimmers end to go up before the voltage peak ends. And the remain at that level although the voltage returns to a normal level. I’ve seen spikes, where the dimmers didn’t react.
I have no glue, why we have those spikes. They occur on each of the 3 phases. Sometimes also at half an hour or - rarely - in between. Sometimes not at all at full hours. This could be in line with the mystic dimming. I will observe.
We also have an additional off-peak power tarif for hot water boilers. Maybe this is related?
Ok, very interesting. I had not heard of ripple-control before as we don’t use it here. It looks like it was turned off for all devices except streetlights in NL in 2021?
By the look of the voltage graphs something is turning off (because the voltage is rising) high-power devices (such as water heaters as you suggested) for 3.5mins every hour. This would cause the local voltage to rise and it may be what you are seeing on the hour. I think it suggests that it is a local load within your house (or a close neighbour) that is turning off rather than a grid load?
The light Esstisch has the new dimmer, the light Luster still the old one with FW 3.4. At full our, 22h, the voltage steps up a bit and the old dimmer gets brighter/increases power, wheras the new one stays at its levels.
Moreover, I’ve switched the old dimmer off and on again and the actual brightnes/power are restored to the initial level. That indicates that the effect is some temporarily compensating measure.
