IT’s possible the low end dimming range of LEDs might be better with trailing edge.
Was this hooked up with a Neutral wire? I was thinking 100% would not even have that step. I wonder if you put the switch in “On/Off” mode + neutral if that step remains?
Yes, it’s with a neutral wire. I think it’s common that dimmers at full brightness aren’t actually on 100%. Both the Lutron ELV and a Lutron forward phase dimmer were the same.
I haven’t tried hooking a scope up to the switch in on/off mode, but my understanding is that if it’s using the relay, it’s going to be on 100%.
It’s possible I misunderstood this post but options in the bulleted list here (bottom of post) seem to imply it should be using the relay in single pole + dimmer + neutral:
I’m pretty sure the relay switches the output of the MOSFET dimmer circuit between the load and traveler terminals.
If you look at a 3-wire with a dumb switch wiring diagram, the dumb switch connects the light load to either the load terminal or the traveler terminal depending on which direction it is switched. So, in the dimmer device the dimmer circuit output which is normally connected to the load terminal has to be switched to the traveler terminal. Hence, the relay is needed to do this switch.
I believe this means if the device is used in single pole neutral mode and set to use the relay that the relay click simply switches the MOSFET dimmer circuit to be powering the traveler wire when the device is turned off which disconnects the MOSFET dimmer circuit from the load terminal and doesn’t allow it to power the load. The relay isn’t in parallel with the MOSFET dimmer circuit so it doesn’t bypass it.
The same thing can be done in a RED by setting it to 3-wire dumb switch mode even when it’s a 1-pole neutral wired application. The internal relay starts clicking when it’s turned on or off.
This also explains why this device can’t do a fan. The internal relay isn’t bypassing the dimmer so the dimmer would have to power the fan, which just isn’t done.
2 Likes
So according to that, Single Pole, Dimmer using MOSFET + Relay is an option:
I’m not clear on what that would do exactly though. And I don’t know how to configure the switch between the two modes either. I see in ZHA a “Disable relay click in on off mode” parameter, which I assume is 261 Disable Relay, “Click” Sound:
In neutral on/off setups, the default is to have a clicking sound to notify you that the relay is open or closed. You may disable this sound by creating a, “simulated” on/off where the switch only will turn onto 100 or off to 0.
But that says it’s only for on/off setups. I have that setting disabled on my switches (which is the default)–disabled the disabling, so the relay click is enabled. On the switches set up as dimmers, there’s no click. And on the on/off switches, it does click.
I guess it might be cool if at 100% brightness, the relay clicks on to bypass the MOSFET dimming and connects the line directly to load, but I have no idea if the internals are wired so that’s possible.
I have mine set up in smart bulb mode. If it didn’t bypass the dimmer wouldn’t that cause a problem with my zigbee bulbs? I guess it’s possible my bulbs could be compensating for that brief off period but idk.
No, it wouldn’t. It’s on enough to power smart bulbs.
Huh, I just checked the waveform with the switch in On/Off mode, relay enabled and clicking, and to my surprise, it looks like it does when in dimmer mode at 100% brightness–it’s actually still using the dimming circuitry and not sending out a pure sine wave 
From what I understood from the previously-linked thread, this is not behaving as expected: “We were able to add the relay function … so that there’s more of a, “hard on/off” for single-pole settings.” “… we haven’t had any complaints around flickering or safety concerns in on/off mode, at least there will be an override if someone experiences this.”
1 Like
Slightly concerned on this. If the dimmer module cannot be bypassed and it cannot put out 100% sine wave power, will it damage non-dimmable lights?
I’m not personally affected by this – i think my lights are dimmable. But when i see “on/off switch” my assumption would be it outputs unmodified sine wave power
4 Likes
It’s exactly what I expected. The relay is a form C contact switching the MOSFETs between the traveler and load screws. It doesn’t bypass the MOSFETs.
Popping in here to say I’ve noted this thread and am bringing it to the engineers to give an explanation. If this was a, “head-fake” last minute to simply add an audible click, there’s going to be an issue for sure.
Thanks for running these tests @dahanc!
3 Likes
Did they promise trailing-edge dimming, or was the decision made to go with leading edge bc it allows non-neutral setups? The reason I ask is that the one Blue I’ve installed so far replaced a Zooz Zen27, which uses trailing edge. The Feit LED bulbs on that circuit dimmed nicely with the Zen27, but with the Blue the bulbs put out a buzzing noise when not at full brightness. I’d keep the Zen27, except it just wasn’t fast enough wrt automation execution in that location. Zigbee has improved that situation, at least…
Here are my raw notes/conversation from when this was brought up with the PM from the manufacturer. I’m not an engineer by any means so rather than trying to interpret this for you guys, I’ll just share what I have.
We initially asked for TRIAC bc it seemed like MOSFET was not as compatible with bulbs (Zooz being the brand often referenced when bulbs wouldn’t work with our switch). Initially it seemed like this was ok (as referenced my my quote where they said TRIAC can both be used with dumb & aux), but I think they realized later that it was not possible.
Here is the original notes from our Red Series manufacturer as to why MOSFET (Leading Edge) was chosen (sorry for the highlights, I was searching MS Outlook for the email and those were the keywords).
Based on the above picture/chart, it’s saying that TRIAC is not compatible with both Dumb and Aux switch, nor will it meet the requirement for EMC (I disagree bc there are examples in the market – for the record so did our current manufacturer, but the bigger point is that TRIAC will not work with both dumb/aux).
–
I think this is the same conclusion our current manufacturer came to as shown in the first few pictures.
The question outstanding though is why leading vs trailing. I honestly don’t know the answer to that and I’m happy to ask once things calm down a bit with the engineers.
The weird thing is that most information I’ve found online says that MOSFET uses trailing edge – is this true?
2 Likes
Triac can only do leading edge, since it uses the next waveform zero crossing as an auto-shutoff signal (that’s my understanding). MOSFETs can do whatever you want, I think…
Edit: does anyone know if a MOSFET trailing edge requires a neutral? Maybe this is why Inovelli ended up with a leading-edge configuration… I would happily forgo the non-neutral option to get trailing edge dimming. Looks like hardware changes are required to swap between the two, though, so with the current switches we’re stuck with leading-edge. Guess I’ll have to do some research on replacement LEDs that will work with this type of dimmer. Disappointing, but I understand not wanting to break out a separate SKU for non-neutral folks.
Lutron has an app note, “Challenges of Dimming LED Loads on ELV and MLV Transformers” which is pretty informative. They don’t state that trailing-edge requires a neutral, but do note:
Most ELV dimmers require a neutral wire for proper operation which helps the dimmer switch on and off properly, regardless of load characteristics.
That same app note claims that forward/reverse in the same dimmer is possible:
There are also “universal” or “adaptive” dimmers, that are capable of creating both forward-phase and reverse-phase waveforms, either through manual configuration or by automatically detecting whether the load is inductive or capacitive. Universal dimmers can be extremely useful, as they can be selected without prior knowledge of the type of transformer that will ultimately be installed.
I’d be curious to know if any devices in the market are actually universal, but it sounds like it’s possible.
Interesting that the Lutron C•L version of the Diva dimmer (model DVWCL-153P) is leading-edge, and doesn’t require a neutral. The trailing-edge version (DVRP-253P) does require a neutral, and is significantly more expensive, so perhaps the hardware differences are significant.
I thought I had the trailing version of the Diva, but it’s the -153P, so is leading-edge. Despite this, the LED compatibility is fantastic, so leaves me feeling a bit better about how good forward-phase can potentially be.
1 Like
We are seriously considering this. The main driver being B2B sales as they are much more price conscious as we’re competing against the Jasco’s of the world who can offer a much cheaper switch and B2B generally doesn’t care about all the bells and whistles.
2 Likes
If it happens, add my vote for a higher-end consumer version that’s neutral-only and does reverse-phase dimming, please!
3 Likes
I have a few Leviton DZ6HD dimmers, which Leviton claims to be universal. What exactly they mean by that, I’m not sure, but there is a setting for load type with incandescent, LED, and CFL as the options.
Unfortunately, I don’t have an oscilloscope to check to see what it’s doing.
1 Like
Following on the on/off mode waveform topic, I have a question: is the MOSFET module capable of output 100% on, unmodified sine-wave power, after some firmware change? Maybe someone with experience in EE field can answer this…
The problem is, currently in on/off mode, the output power is only on for 88% of the time, as @dahanc has observed. (I think) the desired behavior is power-on should be 100% of the time.
I don’t know the answer because I don’t major in EE. But @epow said “MOSFETs can do whatever you want, I think…” and i also heard MOSFET is very versatile. So maybe this is possible?
edit: Blue Series Dimming Tech - #27 by epow - General Discussion - Inovelli Community just saw epow’s comment. So perhaps it’s impossible without hardware change (?)
2 Likes