Probably also have to deal with general home construction. How drafty is the house? Is it an interior or exterior wall? How well insulated is it?
It may be extremely hard to even get this down to anything useful if the engineers themselves can’t come up with a good baseline algorithm
Yeah, I’m sure we could get that data. @EricM_Inovelli what do you think? I think as long as we don’t market it on the product page, it should be fine to have as something that is off by default, but could be turned on as an, “experimental” option or something?
Yeah that’s true – good point.
Quick update – beta units are in the wild and are currently being tested. It’s been submitted to ETL and FCC so we can get a head start on that, while we finalize the beta firmware.
Here’s a screenshot from SmartThings (note: I haven’t verified the humidity reading, but I can see that our bedroom humidifier was reading 54%, so it could be close – close enough for a, “Medium” value).
As it looks right now, we are about a week behind schedule on firmware testing, but a week ahead of schedule on FCC/ETL.
Target production start date on this is Jan 5 to be completed on Jan 20 (does not account for the, “Inovelli Offset” lol).
Will humidity % still be exposed? @SViel mentioned, “medium” would not work well for comparison to other rooms.
Dude, 239.5 kWh? How long has that bathroom fan been running??
So, applying Pi to that estimate, 
I’ll take the week of 3/14 (Pi Day) for these to start shipping…
A couple of thoughts.
Does the switch use a relay? If so, there should be little to no heat generated with the load on, correct?
With the common use-case being control of a bathroom exhaust fan, perhaps the actual value isn’t that important. For example, the rate of change could be used to turn on the fan and a delay or return to the previous value could be used to turn it off.
Awesome to see a production date! Can we get this thing up for presale on the site so I can ask for switches for Christmas? - Also all I want for Christmas is to be a beta tester on this one 
I am also glad to see a date. Been waiting for a switch with power handling that I have hoped the 2-1 switch was going to have in on/off mode. Hope your luck at being a Beta tester is better than mine.
Looks like another fine addition to the growing family of Inovelli products.
A couple of quick questions:
Is this based on the MC24 chip? Will Inovelli be supporting a direct connect to update the protocol to Matter. As a nerd, Zigbee is just fine, but when the time comes and I sell my house, I would convert all the Inovelli products (I have a bunch and will be installing a great many more), to Matter as that would be supported by larger companies that a prospective buyes would be more comfortable with. Personally, I like the Zigbee/Home Assistant privacy, but it appears that most people don’t care and just want it to work easily.
Is the 300 Watt limitation for LED lights accurate or is it a copy and paste error. It would seem that if the switch can handle a 15 Amp inductive load, that this would not be an issue. I have a garage that has a LED light circuit with a 330 Watt load and would like to use this switch, but I will not exceed the official spec limitation.
With regards to generating compensation curves for the temperature and humidity, have you considered the Power Factor (PF - the angular difference in the voltage and current vectors). I noticed when I set up a test jig using LED lights with a KillaWatt meter, the PF for the 2:1 dimmer switch was much lower than I expected. That enegy has to be dissipated somewhere and may the cause of your inaccuracy. My postulation that if the PF is included in your compensation curve, you’ll get a more accurate temperature and humidity reading.
You’ll have to wait to see what the final specs are as that spec is subject change during testing/certification. The lower LED rating is there to account for LED inrush current.
What is the current design spec for LED inrush current?
Lol, I forgot to mention, we use commercial grade space heaters and pray it doesn’t get too humid jk.
I’m honestly not sure why it’s so high. It’s saying I’m drawing 200W on the exhaust fan too and while the fan was here when we moved in, I think the master bathroom was redone fairly recently, so I can’t imagine they put in some super old fan. I thought maybe it was the driver I was using, but one of the other beta testers is using the same one and his is reading normal, so now you have me curious on what fan is up there. Unfortunately, the only way to access it is climbing in the attic bc they put tile around it so I can’t pull it down. At least it’s winter in Michigan so it’s not 100+ degrees!
Let’s see if we can beat that estimate lol!
It does use a relay – let me check with the engineer, I’m not entirely sure on the mechanics of why there’s heat interference. Great question.
Yeah, this is kind of where I’m at with it too. Here’s an example of how I am using it in SmartThings.
If the humidity detection is equal to or above 45%, then the fan automatically turns on and when it is equal to or below 44%, it will turn off.
I plan on finding a baseline humidity for when the shower is on and then adjusting the percentages from there. For example, if I turn the shower on and see the readings are consistently at 50% and above when the shower is on, but normally around 40%, then I can set the automation to run when it’s around 45% or something.
So, scratch my earlier comment around low, medium and high – we will give the percentage reading. Just know it’s not going to be 100% accurate depending on the heat generated from the switch. However, I think it should be consistent so you should be able to dial it in.
Yeah, I wanted to make it to the first round of beta testing before putting it up for pre-order. Now that we have it in testing, I can put it up on the site. The only thing I’m still a bit nervous about is if ETL approves the increased LED wattage we asked for (600W), but I suppose I could put a call-out on the page.
As for the beta tester question, this project was more of a, “assemble the current beta testers and let’s knock this out since it’s a simple project”, but I’m happy to add you to a new project if you’d like?
I can add you as a beta tester for a future project if you’d like? We’ll have plenty, I promise!
Great question – this one is based on the MG21 chip for a couple of reasons. The first was to keep the costs down as the MG24 is starting to become more expensive as more companies are expanding into Matter. I know that doesn’t seem logical as the more chips that are made, the more cost efficient it should become, but when we first created the project, our manufacturer told us it was getting hard to source the MG24 so the costs were higher.
I’m sure it will level out, but at this time, we’re going to just move forward with the MG21.
At the same time, there’s been a lot of confusion around whether or not the MG21/24 can update to Matter and I am under the impression that it’s not an easy process to do so. You can go from Matter to Zigbee, but from Zigbee to Matter requires a special license. This was the latest response I received when I asked about Zigbee to Matter:
"Zigbee updates to Matter require additional DAC certificate fees. We provide the matter tool, which includes a small number of DAC certificates, only for inovelli to test for free. Matter updates to zigbee at no additional cost.
In any case, these update methods are not easy to do, and for now they are only available to us and inovelli, not to users"
–
So, I think what the current plan is to just stick with Zigbee and then create a separate Matter SKU until it becomes more stable of a process. Maybe with the MG26 it will be easier and there won’t be any additional fees, but so far Matter has quite the pricing racket.
@Bry answered this correctly in that we kept it conservative prior to knowing what ETL would approve. We have since requested a 600W LED value to match other companies (in this case Leviton) and we’ll see what they come back with.
Here is what we’ve submitted to ETL:
Let me check.
Just so I’m clear when I ask the engineer, what would this metric look like? Or, I guess better question (sorry, I’m a marketing guy, not an engineer!), if I asked this specific question, would the engineer know what to provide?
Yeah, we can figure something out. We will make it available for “power users” in some way.
The peak inrush current in Amperes (Amps).
This is a more satifying answer. With a LED design goal of 600W, it is likely my situtation requiring an additional 10% in Wattage over present specification will be met.
I think the decision to move forward with the MG21 may be short sighted.
Maybe I’m off base, but I’m thinking that Inovelli is shooting for the same switch functionality across Zwave, Zigbee, and Matter. This is a great goal of having dimmers, on/off, fan controllers, fan modules, and the button controller functionality for each use case.
I’m am also thinking that the difference between the Zigbee and Matter switches could be reduced to the firmware loaded into the switch and maybe the color of the box and inserted instructions. I understand the initial cost considerations, but I would think that the goal is bring all devices under a single hardware chip (i.e. MG24 for the foreseeable future).
One of Inovelli’s goals should be to reduce touch labor for commercial and integrators. Currently, Matter is not ready for prime time, but that day is coming. I have been doing my own switch upgrades, but with today’s labor cost, the cost for an Electrician to change out a switch is more than the switch itself.
The best of all worlds would be to support an OTA upgrade to Matter from Zigbee, but if I had 400 switches to upgrade in a building, it would be less labor (and expense) to procure (rent or buy) N number of serial interfaces to upgrade N units in parallel (where N is based on the time to update a switch to Matter), rather than replacing each switch. If the switch has to be replaced, then the added labor to replace the switch, as well as the purchase of a new switch, the end expense gets at least doubled.
As to covering the DAC license, Inovelli could charge it as an option at purchase time of a Zigbee MG24 switch or purchase them as an add-on based on the number of purchased switches.
For me personally, I’ve been holding off purchasing more dimmer switches awaiting the upgrade to the MG24. I like Inovelli, I really do, but the ongoing Inovelli Zigbee/Matter turmoil has convinced me to stop any additional Inovelli switch installation or purchases until things settle down.
Maybe I’m a small minority of the Inovelli user base, but until the Matter integration in Home Assistant is out of beta and in use for at least a year, I will not purchase anything Matter related. With my projected number of devices, I’m not willing to risk the headaches.
phone I am with you. I went with Zigbee right now as Matter still has its issues. I though of going with Z-Wave LR as ZOOZ has switches readily available with almost same functionality, more chooses, and lower price. But as Zigbee has the potential of upgrading to Matter (my Zigbee coordinator can) I when this way. If Inovelli Zigbee has no upgrade path then I feel this is a major down fall.
Did I miss the opportunity to get these on pre-order?