Hmmm, that is too bad. It looks like it could still be a settings thing, but if you did a factory reset it doesn’t seem like that would be the case. Can you post the values of the “State Variables” that are on the device page? I just want to look them over to see if something stands out. First do a refresh though:
Looks like the mosfet for the negative cycle is short circuit. The positive cycle is dimming to 50% properly.
Even when the switch is turned off the negative cycle is still at 100%
Trailing edge would probably increase the reliability of the switch and fix the buzzing problem with some bulbs. You say it’s not possible because the switch won’t be able to auto detect an aux switch vs a dumb aux switch? I’m pretty sure the constant current spikes caused by leading edge dimming are killing the switch.
Thanks for the info, I’ve verified the settings you have and they should be fine. Is there anything interesting about the load attached to the switch? I have some similar circuits to (3-way, neutral, etc.) and have not seen this issue.
Regarding changing to trailing edge, one of the reasons is that it would cause problems with the circuit that detects when a dumb switch is toggled or an aux switch button is pressed. So best case would be that we could get it to work in single pole mode, but the engineer said that is likely not possible via firmware either. I’ll PM you with our support rep so we can get the device replaced.
Edit: One quick thing that I am wondering if you can check is if you adjust the “max level” down to 95, 90, 85, 80 to see if you notice any difference in behavior.
I wouldn’t say there is anything that interesting. The load is 6 13 Watt dimmable bulbs. 3 of them are ecosmart the other 3 are feit electric. It has been working normally since early November. The only thing I can think of is I leave them at 30-50% brightness most of the time.
I don’t see how trailing edge vs leading edge would confuse the switch. Doesn’t it just sense when the load/traveler is pulling current? Maybe the Inovelli white aux switch is using the leading edge to send current pulses to the main switch? I think with a 3-way dumb switch trailing edge should be doable since it doesn’t need to send any signals to the main switch.
Does the switch just need a little dead time to test the traveler wire to make sure its not an aux switch?
I can try changing the max level and see if that changes anything.
Ok, so I understand needing a little dead time for the switch to get power without a neutral wire. Couldn’t this still happen by having the deadtime as a trailing edge? So it never quite hits 100%.
I also understand that it might need a bit of time for zero crossing.
Why can’t it either
1.) keep a small leading edge but accomplish bulb dimming by varying the trailing edge?
2.) have a smooth leading edge, keep a small trailing edge for power and switch the relay shortly after the mosfet is turned off just before zero crossing?
With a dumb 3-way switch the relay shouldn’t ever have to interrupt the current. The 3-way switch breaks the circuit and the relay only needs to be used to turn the light on.
I can confirm that the 40N20 N-channel mosfet Q7 failed short circuit.
My load is 750ma RMS 913ma Peak.
Edit: Now that I think about it, I did set the ramp on time to instant. If you set it to instant and the brightness is 50% does the switch turn on at peak voltage? It should probably spend at least 2 or 3 full cycles at 100% to charge up the LED drivers before going to 50% brightness. With leading edge 50% brightness during a cold start is way worse than starting out at 100% brightness.
Alright so after doing quite a bit of testing I’m a bit confused. The switch is doing this routine where it pulses the power then outputs a 50V square wave when off. I assume this has something to do with the neutral power or zero crossing detection.
What’s the point of switching the relay at zero crossing if you are just going to wait and turn your mosfet on at the absolute worst time? (Not trying to be mean here but I think there is room to improve the startup behavior.)
For now I think you should tell your customers not to set the Off to On Ramp rate to instant.
QuickStart is not [yet] officially supported in the Dimmer firmware. Its something that the Fan firmware has and inspired me to try and EMULATE it in the Hubitat dimmer driver until it gets added to the Dimmer firmware later. The intention is for LEDs that need some minimum brightness to turn on, but can be dimmed lower after they are on.
The driver code should be turning on the light at the “Quick Start Level” setting for “Quick Start Duration” milliseconds and then go back to the previous brightness level. But now that you mention it, I think I see a bug in how it handles an “On” command versus a “Set Level” command. I’ll fix that.
Not really sure how you think it should work. The opposite would be starting at the previous level and ending at the Quick Start level which would not make sense.
With that said, I see I need to check if the Quick Start Level is less than the current level then just turn on at the current level without doing a Quick Start.
I originally had it that way, but it caused a very annoying bright flash of the light when Quick Start was enabled. I changed the DEFAULT startup level to 50 to make that initial flash less dramatic (especially at night). Its a settable parameter and you can change it back to 100 or any other value you choose. Ideally, that value should be the minimum value needed to turn on the light before dimming down lower. But that varies with different light fixtures so 50% seemed like the best default value.
I’ve been running my Blues with instant Off to On Ramp rate for almost a year and haven’t had it cause any failures. I think you just happened to get a bad one.
The worst startup current I was able to reproduce with this setup was 6.4 amps per bulb. So the startup peak current is 38.4 amps.
The feit bulbs seem to be quite a bit worse than the ecosmart bulbs.
The worst case seems to be with the load on the load terminal, using a 3 way dumb switch, ramp off to on instant and brightness 50%.
I have read this post in its entirety and greatly appreciate all of the technical detail users have provided, as well as the transparency of the founders and leaders of the company itself.
That said, this thread has reinforced a long-time tenant I have held surrounding smart home accessories, which is to stick with companies that have a long history creating the underlying product before the idea of smart was even a thing.
Founders: You need to get the basics right before adding the smart features. I look forward to future revisions and have faith that you can get this right. In the interim, some of the conditions that can exist in the current revisions surrounding dimming will damage certain bulbs, fixtures, and the switch itself in the long term.
Yeah, I’m new to home automation, and the Blue series are basically the first devices I’ve bought (I did get one power outlet while waiting for the Blues). From what I saw, the company culture is great, and I really liked how y’all work closely with the community. That, along with the switches’ feature list is why I decided to buy the switches. After receiving the switches, I still think the smart features are great (e.g., all the parameters that can be controlled, the notification LEDs), but the core switch/dimming functionality is kinda meh.
As someone else mentioned, while y’all intended for this to be a 2-in-1 on/off and dimmer, I don’t think it can be considered an on/off switch. While almost all of my lights are dimmable, I still have two that are not, and while I’d like to have a smart switch control them, I know they’re not going to handle the choppy waveform well.
I remember reading that y’all don’t have an in-house electronics engineer, but perhaps it’d be worth hiring one as a consultant to validate future designs to make sure it does the things you want it to do, and to point out any potential issues.
I apologize for not having the time to properly respond here, but it seems this thread has caught some attention, so I’ll take the time to respond as best as I can.
From what I am reading, it appears there are two things we’re talking through, correct?
If there’s something more, please let me know.
__
On/Off Discussion
I will take the heat on this one, as I was the one who named the switch a 2-1, which I personally still believe is true.
I understand the argument that an on/off switch should be one that completely shuts off the load and there should be a pure sine wave. If I’m being honest, I didn’t really understand it until everyone brought out their oscilloscopes and explained what it should look like.
To me, an On/Off switch is one that turns lights on and off and in most cases, it does just that.
There are, of course, exceptions such as fluorescent lights. This is common in the industry and I think the folks at Zooz summed it up well:
There is also an argument that the switch should be able to turn an outlet on/off and that is currently possible, but we do not advertise it to do so because of a safety concern (people plugging in their vacuum cleaners, space heaters, etc) and the switch cannot handle a strong inductive load.
Finally, I believe it comes down to how this switch is advertised. Yes, initially we thought this would be able to handle ceiling fans and I stated that it would. But later in the same thread, I also stated that it couldn’t due to UL and the actual product page has had this stated there as well.
On the official product page, I feel we’ve been clear about what this product can and cannot do. But I’m always open to making it better.
I do understand the arguments and they are valid arguments, which leads me to my next point…
We have been testing a pure sine wave parameter that should be available to you guys shortly. I want to run some final tests on it, but from what I’m seeing from the oscilloscope readings our beta testers have done, they look promising.
Leading vs Trailing Edge
This topic has been debated about for a while and somewhere in the forum I outlined the reasoning why we had to go with Leading Edge. It boils down to allowing us to have the capability to use a neutral and non-neutral or an aux and dumb switch. Everyone wants to have a choice in their installation and unfortunately that comes with tradeoffs. Trailing Edge being one of them.
Now, I will say we are working on ways to optimize the bulb flickering and just received a file today that addresses certain scenarios. I want to fully test this firmware file before throwing it to the wolves.
–
Here’s what I want to get across though. I do read through most of these comments and I do take these suggestions to heart. These switches, as dumb as it may sound, are a reflection of us and we are extremely proud of them.
So to hear things like, “these switches are meh” or, “we need to get the basics right”, etc kind of irks me as it just seems like we’re all taking a dump on a product we’re proud of. I know it’s your personal opinions and that’s fine, just disappointing.
I know sometimes I may come off as defensive and I’m trying to work on that (I feel like I do a good job at tempering what I want to say lol) and I try to go into these conversations with an open mind. For the most part, I agree with you guys – without you, there would be no Inovelli. You are what makes us who we are and make our products what they are.
I believe in my heart that most of you have the best intentions in mind when you’re commenting on our switches and I hold onto that.
At the end of the day, we want to continue to have the best switches in the market and if that means taking a long hard look at the technology behind it, I’m all for it. I can’t promise you it will change as it’s impossible to please everyone, but I can promise you that we will look into it and have an open discussion as I try to do.
–
Give me a week or two and I hope you all will be excited about the firmware release we’re working on to address some of these concerns.
Thanks for coming to my Ted Talk.
Long time lurker, first time poster. I really appreciate the candid and honest discussions taking place in these forums. I won’t need any of these products for a year until my house is done, but I will definitely be making use of some of them when I am. I’m hoping to mostly use Zigbee lights, and therefore the limitations of the blue series won’t impact me much, but I would love if we figured out the dimming and switching issues before then so that I can use these switches throughout the entire house even when we’re not using Zigbee lighting.
A lot of people, myself included, who are into smart homes are pretty geeky, and a lot of us have testing gear the average person does not. I think we just find it a little weird how small your company is, and that you don’t have the same resources some of us do. But we are not normal.
If there’s one thing I’ve learned when working with other companies, particularly in another country, is trust, but verify. I do think you guys should get some on premise help, or a nearby firm to help you test the boundaries of what you are receiving, rather than trusting the vendor to do the right thing.
+1 with emphasis. The head of our mechanical engineering department preaches this with his two rules:
Rule #1: all vendors lie.
Rule #2: see rule #1.
Add in the language barrier, and even a mostly-honest vendor can end up delivering something that doesn’t fit expectations. @Eric_Inovelli you should be proud of these switches! On the smart side the number of parameters and the flexibility blows everyone else’s products away. Dissappointment in our reaction to the relay not being parallel to the dimmer and the leading edge dimming is understandable, but from our perspective there’s disappointment, too. These two issues were an unanticipated setback for us, and those always come with frustration. It sounds like our message is resonating, and I’m optimistic and hopeful that future iterations incorporate changes that eliminate the current crop of fundamental performance issues.
Just chiming in to say that I definitely agree with everyone that we need some sort of engineer to verify everything. I think the problem we run into is twofold:
Now one can argue that we can’t afford not to get someone and that’s where I’m at right now. I know we need someone, but where do we find one? And find one that we can also trust?
If I’m being completely open here, COVID and the last couple years has done some incredible damage to cashflow which we’re finally starting to get out of. But, there’s still a ways to go. I won’t bore everyone with the nitty gritty, but the gist of it is that we’re trying to climb out of the back-order/pre-order cycle where we place an order for product, but by the time it gets here, it’s all sold out and the cycle continues.
We do much better cashflow-wise when we’re fully in stock and unfortunately due to us being out of stock, we’re running lean and we’ve had to watch every nickel/dime (I’m exaggerating, but also not really).
–
Just trying to paint a picture right now of where we’re at.
So how do we solve this?
Well, there are a couple of options, but I’m open to any and all feedback:
Thoughts?
Option one is definitely a good place to start and then expand into option two as capital becomes available. I’ve had the privilege of helping beta test the latest set of switches and there are some very knowledgeable community members volunteering their time. Personally i find it enlightening (snicker) and I’ve learned a bunch from the other members.
I really appreciate how candid you have been in your responses here. It’s not my decision to make, but I would argue that you are NOT clear enough in the marketing for what the switch is and is not capable of. You are marketing to the DIY community, who doesn’t always necessarilly know (or care to look at) the difference between the wattage ratings of switches. They will assume that an on/off switch would be capable of supporting anything a normal on/off switch would. And IMO an on/off switch SHOULD. I understand you look at Zooz as your competitor and they have made a market selling products at a cheap price point.
However, I don’t think looking at the lowest common denominator and trying to emulate them is a good path to success. It’s more valuable to look at what the long-time players in the switch industry are doing. Leviton’s on/off switches are designed to support 15A hardwired outlets. In fact, I struggle to think of scenarios where I would want an on/off switch that isn’t capable of doing that. What you have is a switch (actually dimmer) that is ONLY capable of lighting, and if that’s what you are after you need to make it 100% clear that it does not support anything other than lighting.
People will do inevitably things they shouldn’t, despite any warnings. If you aren’t 100% clear of the switches capabilities, when someone’s switch melts (or the house burns down), whose reputation is on the line? Do you think “well the tech specs say it only supports 600W LEDs, and there is a UL number that they should have looked up listed” is a valid defense?
I know I’m one of the few naysayers here, but I’m not trying to be a complete downer. I love what you guys are doing and want to see you succeed.
I will add that I consider myself fairly tech-savvy and I incorrectly assumed that the 2-1 would be capable of numerous on/off duties around the house. I’m sure there are plenty of others that made or will make the same mistake. I was hoping to go all in with your products, and now am trying to figure out what I should do. Having mismatches switches drives me crazy, and you don’t offer a true on/off, so I need to look elsewhere for that and it’s unfortunate.
Yes.
