Powerball: 13
There, you guys can have that for free.
5 Likes
So this worked for me but your milage will vary… I had a switch that I could not get connected at all… I pulled a plug from a different part of my house, plugged it real close to my non working switch… and the hub found it and connected it for the first time!
I then moved the plug back to its original location and the switch is still working as of now. Could be a way for some of you to get the already installed (but not working) switches onto your network and see if they will stay connected.
EDIT: ok it fell off the network… so I had to move the plug back… and it worked again… so as long as you leave a plug close it can work until Eric finds the real fix.
1 Like
In case anyone here wants to quickly look through all their switches (without taking them out of the box to inspect the sticker inside) and has an iOS device, I made this shortcut that scans all the QR codes and writes them into a CSV-formatted icloud Notes document: Shortcuts It’ll do max 100 at a time, but you can re-run it as often as you like.
Just scanned all 40 of mine, and welp, I’m 39/40 for afflicted IEEE address ranges.
8 Likes
I think I’ve managed to confuse myself…
Thus far, we suspect 94 and 38 are the sketch ones WRT signal/routing, but 04 seems to be OK – do I have that correct? Thanks!
Installed my first one today and could not get it to pair… even with an ikea router on an extension cord within inches. Eventually found this thread, looks like I’m 30 for 30 94:34:69 and 38:5B:44. Think I’ll hold off installing anymore for now, maybe buy one of those lotto tickets too. 
Looking forward to hearing more from @Eric_Inovelli, certainly disappointing for all, but I suspect you guys even more so.
2 Likes
38:5b:44 and 94:34:69 to be a bit more specific, one of those had an ok range in the 2nd octet, so good to use those 3 to confirm if the switch is in an affected ‘batch’
2 Likes
I see a lot of people reporting signal levels from HA… is there a hubitat equivalent measurement or a way to see if the switches around routing though each other?
I have 40 switches to install. I was going to get started this weekend, but at this point I will wait until there is a resolution to this. I don’t want to install/uninstall/install again.
I believe it’s possible but I don’t have the steps and the ‘easy’ value to get from Hubitat is the last-hop LQI, which isn’t what you’re looking for or an accurate representation. I think the recommendation I would go off currently is to hold off on installing any switches in those batches until we know more. (I’ve got one from that batch installed but holding others back for now, was really to validate behavior and see if I could get it to join)
Thanks for confirming - now that I finally went back and documented those numbers for my 9 switches, I was curious to cross-check those against the odd-ball unsolicited binding issues I had… But no direct correlation.
For now, mine are all still in the wall, but all reset/unpaired so just acting as dumb switches/dimmers.
The binding issue should be a separate one and I know they’re working to get the fix tested/shared.
1 Like
Thanks,
My Zigbee logging is completely blank. I only have two Zigbee devices installed. Both Inovelli switches. Is there somewhere I need to enable logging or should that be on by default?
40/40 Switches are 94:34:69 or 38:5b:44 with date codes 2208 or 2209.
I’m gonna wait until we hear from Eric on what we should do.
1 Like
I don’t think I have any debug logging on. Mines not showing every single one either; however there’s more today than yesterday.
1 Like
Zigbee info on Hubitat (the page @harjms shows above) is not very easy to interpret, and it doesn’t necessarily show info for every ZB device in your mesh.
I scanned my bookmarked posts and found this pretty good summary of that page from one of the Hubitat community members who’s smart on it.
I use that page more to check for any major buffoonery, but I focus more primarily on making sure I have plenty of good ZB repeaters located well to ensure a strong ZB mesh.
I had mostly 94:x Blues in my order, and I had no issues pairing them or with performance (in terms of responsiveness etc). Unfortunately, I had to decommission all my Blues due to some unsolicited binding issues, so I never got a chance to really take a hard look at that ZB performance page…
But I remember looking at it once at some point and not being impressed with the Blues’ numbers, but since everything was working fine and I knew I had a good mesh backbone minus the Blues, my plan was to just let it ride… I’d do the same thing in the future.
Wanted to give everyone an update. I likely won’t be able to respond as quickly as I’d like as I actually came down with a pretty nasty cold and am typing from my bed bc I know you guys are anxious for an answer and you deserve one.
I will explain further either tonight or tomorrow when I am by my laptop and can type out an action plan.
Edit: nm, I got carried away in typing that I just basically laid out everything. But more info will come in regards to next steps early next week.
–
The manufacturer was able to identify the reason for the poor signal issues and it boils down to a hardware issue from one of their suppliers.
Those affected are the IEEE addresses mentioned above (94:xx and 38:xx)
The reason why it passed QC was bc they only tested pairing at 0.5m (wut…?) and didn’t notice any anomolies.
We had a good conversation with them yesterday about how to solve this and they’ve agreed to replace all units that are defective. Not only that, but they are putting together a much more comprehensive testing procedure so that this doesn’t happen again. Especially with production starting next week for Batch 2.
That said, I at least want to be appreciative of them for owning this and for not hiding behind some excuse to try to screw us over. This has happened in the past with another manufacturer and is sadly common in China when dealing with them as a small company.
We spoke directly with the owner and he immediately apologized, and said this was not the way he wanted to start this new relationship off and he offered, without hesitation, to replace these and compensate us for any costs associated with this.
–
On one hand, I’m relieved to know that the product we all tested and knew was an amazing switch still holds true. When you put so much work into these and know what you have is good (if not the best) it’s extremely frustrating to see all these issues pop up. Especially issues that were not there in beta.
On the other hand, I’m frustrated that this wasn’t caught at the factory and better procedures weren’t in place as this seems like a very simple fix (to test at the appropriate distance for Zigbee - not 0.5m). And now all you guys have to spend your time, energy and patience in getting and installing new products.
–
Timelines
We have a batch that starts production next week and that should be finished by the end of the month. This batch, “luckily” can be used for both replacements and for any backorders that were placed.
I would give my best guess that these will be in the US in early December.
Two weeks after that, another batch of units will come in and following that the remainder with everything being made up to us by the beginning of January.
To be clear - early/mid December is my best guess as to when replacements will be sent out (hinges on customs).
–
The Process
This is what I’m still discussing and was going to discuss today but unfortunately my operations guy is out with COVID and I’m bedridden with something so we’ll have to construct the logistics next week.
But at the very least, we will send out replacements before we ask for the old ones back (if we do - honestly idk if it’s worth it).
I’m trying to figure out a way to not find the defective ones on eBay or something.
We’re also going to give out some sort of compensation for your time wasted - likely in the form of a coupon off future orders or some rewards points, I’m not sure yet.
–
I want to sincerely apologize for all the frustration and at any time if you’d just like to say forget it, I’m out and would like a refund, I completely understand.
This is definitely not the way we wanted to kick off a brand new lineup.
38 Likes
There’s no changing the facts of the situation, no matter how disappointing/frustrating/inconvenient it may be for everyone. The most useful thing to do is to resolve things in the best way possible for customers and that seems to be the plan. Thanks for the transparency and focus, Eric.
8 Likes
On the plus side, those who have working switches will help iron out the remaining firmware bugs. Ppl like me whose entire batch is compromised may have to wait, but when the replacements arrive I expect there’ll be better firmware available, and some great know-how posted here on how to configure things.
2 Likes
I appreciate the update and the transparency. Disappointed that I’ll have to replace the majority of my switches again, but I’m glad that this is being taken seriously and you and the manufacturer are doing the right thing.
1 Like
OK to order now?
I’ve been following along before making a purchase. Seeing that we are backordered and need to replace some number of items in the field… what does it realistically look like for deliveries for any new orders placed now? Lets say I want 60-80 switches.