Yes, unfortunately that’s a “bug” ? in livolo switches, or it has to do with the fact that they have their extra synchornization with their slave switches in two/three way mode.
What you could try though is use a single remote ID and have all of the switches that you want learn first a toogle button (let’s say button 10), and then have them learn the On Only Scene button. This way you can turn them on with the On Only scene button, and turn them off with the remote’s ALL OFF button.
To be more specific, when you have a switch learn a remote’s TOGGLE BUTTON, then the switch will also respond to the remote’s ALL OFF button (no need to teach it the all off).
I have try several ways. The best is toggle on/off and with different remote id for each button.
For scene mode. I learn only the all off command using same remote id, to turn all of them off.
For scene mode, there is some delay and sometime turn off only 2 lights instead of 3 lights / as 3 lights come from 2 switches. 1 switch is normal switch and another 2 gang switch is 2 way switch
Keep in mind that since the Livolo switches do not provide feedback to the remote (i.e. tell the remote that OK, I am now ON), the remote (or RMPro) will only send the command, and not check whether the switch has turned on or off. So if the command fails (and this can be due to distance, RF interference, or poor reception by the switch) the remote will not re-send the code.
Livolo’s fault tolerence with RF signals is rather limited, and for that particular reason what the remote does is send the command 100 times in quick succession so that it somehow ensures that one of those repeats will be received by the switch. When the switch receives the command, then it enters a 2 second timeout where it will not accept any other command, to prevent turning on and off again (or off and on again).
This also explains the delay when you have multiple switches bound to the same command, as when the command is sent 100 times, the first switch might capture the command on the first repeat, while the second one might capture it on the 50th and the third switch on the 100th (and perhaps the 4th switch will not capture it at all).
I have resolved this by using Alexa’s device grouping function.
So when I have taught a command to a switch (or to a group of switches), I add that command to HA two times as two separate devices, let’s say “living_room_1” and “living_room_2”.
Then in Alexa’s Group devices, I create a group called “Living Room Lights” and put both living_room_1 and living_room_2 in there, so when I ask Alexa to turn on Living Room Lights, she will send the command loop twice, increasing the odds that the command will be received by all switches.
I got my 2 gang and 3 gang switches running 4 scenes from the same remote id.
Another info is that I allways assigned a scene to all buttons from the same Switch. Pressing each button at least once before sending the scene code.
I have assigned 4 scenes on my living room, all from the same remote id.
I have also two three gang switches connected 2 way, but only one of them has RF feature.
They also are part of all 4 scenes I assigned.
I have no problem to create more than 1scene for normal 1,2,3 gang switch. But for 2 ways and 3 ways switches. I can create only 1 scene either all on or all off to group them together, I do using same remote id. My is US switch.
My 2 ways and 3 ways switches can learn only 1 scene. using same remote id. There is some fault when trying to learn the scene with different remote id.
Great work guys I managed to setup on/off only codes very easily which I consider a very good improvement. Still having reliability issues with RM Pro which cannot match remote control yet. I will try to send the command multiple times to see if it will help.
The 2 second wait time (might not be exactly 2 seconds, it’s just an approximation) applies to all livolo switches.
When a switch succesfully receives a command (like On, Off or toggle), the same will not accept any new command for a period of 2 seconds, so that it doesn’t switch on and then back to off immediately in the case of toggle commands. After the 2 seconds are out, it will be able to receive commands normally again.
It’s not a design flaw, it’s a feature of the switch.
Plugging the RMPro to an adapter that outputs bigger amperage might help with the reliability issues, though I’ve found that the improvement is minimal. At some point I just gave up trying to learn the codes from Livolo’s remote as the end results were pretty inconsistent and the reliability of the command actually being performed was sketchy and depended largely on the switch and RF interference in the environment.
Using the pre-constructed codes by tyjtyj’s application on the RMPro instead of the remote, improved the reliability by orders of mangitude. I haven’t measured the results accurately, but the perceived failure rate has dropped from about 30-40% to just under 5%.
Using tyjtyj’s codes also eliminated another issue I had faced, which was that the transmission of a code intended for a switch, sometimes (though rarely) also triggered another switch, which was scary. I haven’t seen this happen at all now.
What is still apparent though is that the most stubborn switches are the 2/3 way ones. They are the ones that portray the smallest reliability with the RMPro.
Finally, sending the code multiple times does help, as long as you opt for ON only or OFF only commands. Toggle commands might go beyond the switch’s 2 second inactivity buffer and cause it to turn on and back off again. And of course, sending the same code multiple times will “clog” the RMPro’s buffer and will cause you to have delayes in response times of the switches if you are performing combos. So for combos I’d use separate scenes.
Και… ελπίζω όλος ο πόνος, το αίμα, τα δάκρυα και ο ιδρώτας που έχουμε ρίξει ένα χρόνο τώρα με τους Livolo να σε βοήθησε να βρεις άκρη
Amazing work done here.
May be off topic but related with Livolo, is there a way to increase the brightness of the blue led?
1.
The “diffuser” plastic is crap and does not diffuse the blue light enough, maybe replaced with a more “frosty” one but cannot find anywhere such “tape”
Will it interfere with the touch function?
2.
Replace the blue led with another one brighter (no my first option:)) or make the stock one brighter by “code flash” or puling additional voltage via a wire from somewhere on the board .
I’ve seen leds this small incredibly bright.
I have no problem them being bright in the dark, actually I use them to navigate the house, in total darkness ,when the nature calls in the night :)
Any opinion from you are welcomed since I want Livolo for their design and build quality and also wide range of switches.
I’m not an expert on electronics by far, but I can offer a bit of input here:
The switch without the glass panel is very responsive to touches, but the glass panel adds a significant barrier to touch recognition. If the panel is not perfectly aligned you might end up having switches that are not very responsive to the touch. Sometimes you have to touch multiple times until the touch is registered, and sometimes you have to press hard.
As such, using a different panel (especially a more frosty one) might have significant adverse effects to the touch recognition.
If I were to replace the leds, I’d try to find leds that are brighter using the same voltage as the ones already installed (not sure if that is possible). Livolo switches operate under very specific minimal voltages so they are able to both power the leds and the RF receiving mechanism, and not send excessive voltage through the circuit able to cause your lamps to flicker (and this is why livolo says that if you have very low wattage led lamps, they might flcker periodically when the switch is off). If you increase the voltage in the loop, you’ll most probably cause your lights to flicker.
I am talking about the small round “plastic” that is on the pcb, that help diffuse the blue light, not the glass panel.
Apparently does not diffuse the blue light enough… rather is transparent look instead of a frost one to even the blue light.
(btw mine are working the same with or without the glass panel, meaning very responsive :))
Soldering another led is kinda of tedious… for all my 30 buttons
just to clrear out what I’m talking about:
-the Livolo difuser (small transparent , not much good
-other difuser, bigger, frosstier, nice bright even light.
I know there is some tape out there but since I am not native English I don’t know how is called.
Maybe will even out the blue light )
Ah sorry i misunderstood. I would suspect that the diffuser also acts as the touch panel. As such I’d be reluctant to change it.
I’ve seen some people using stencil film to diffuse the led light.
Problem is that it’s not self adhesive so you’d have to find a way to apply it. You can get a small batch (A couple of sheets) so you can test to see if the touch is affected, and if you’re happy with the result go on and get more.
Do All your 2 and 3 way switches are RF enabled?
I have a 2 way circuit using 3 gang switches, but just one switch is RF, the other switch is a simple 2 way 3 gang switch. And I can apply more than two scenes on this circuit teaching the scenes to the RF capable switch of course.
Yes, my 2 and 3 ways switch both are rf enabled. So when I learn on one switch will effect the others switch as well.
example for my 2 ways and 3 ways switch
1st switch 2 gang
2nd switch 3 gang
1 button of 1st and 2nd switch join to same light <== I teach either 1st or 2nd switch for button 1 / will turn on/off the same light
2 button of 1st and 2nd switch join to the same light <== I teach either 1st or 2nd switch for button 2 / will turn on/off the same light
3 button of 2nd switch to another light <== this teach to another light
So your setup is completly different from mine 2 way setup.
I think you are having problem to teaching more than 1 scene because of that.
My instalation is pretty simplier than yours.
I have a 3 gang RF eneabled connected 2 way with another 3 gang NOT RF enable…
My 2 way NOT RF acts like a slave.
When I linked 2 way I assigned just the first button and the others were auto assigned together…
So basicaly I teach scenes (more than 4 I have already got working) to a 3 gang and the two way fact does not matter in my case, because it is like a slave, without interfearing with the RF function.
Have you thought there would be a fight for master controlling the same light? Do you have another 2 way switch without RF to test?
This is what happen to the master of 2 way 3 gang switch. I can learn on/off (toggle) for each button 3 gang switch.
When I try to learn 1 scene for each button 3 gang switch, either the second or third button will forget the toggle on/off.
