As part of trying to solve this problem I bough an IKEA Zigbee Hub/Gateway. As soon as you connect any IKEA device it will automatically update the latest firmware to both the buttons and any IKEA repeaters you have…
And for me the problem remains.
So far to keep my 7 “cheap” Ikea blinds working I have:
Bought 12 cheap dollar store batteries $3
Bought 8 expensive name brand batteries $25
Bought an IKEA Hub $35
Bought a HUSBZB dongle. (on Home Assistant) $95
Got a second Zigbee dongle (gift from a buddy) Free (on loan)
Bought 12 more dollar store batteries $3
Recharged all of my 7 blinds after one month of use
Moved my 2.4 Wifi to the opposite end of the spectrum from the Zigbee (Ch11 and Ch 11)
I have tried pairing the buttons directly to the Hub, and to specific repeaters neither works.
I have lost track of how many times I have reprogrammed these buttons… in Smartthings and Home Assistant and in the IKEA app… Oh for anyone interested the IKEA app really sucks. And does not make the buttons actions open to trigger other tasks… which is the whole point of home automation.
Smartthings and HA can still open and close the blinds, but the buttons are all useless.
Because of the size of my house (>5000 sqft ) I use primary ZWave since it has better range. But HA reports that my Zigbee Mesh is strong and the 7 repeaters are all sprinkled throughout the house so the Buttons are normally within 4 or 5 feet of a repeater.
According to the Mesh map that Home Assistant provides Buttons will often ignore the repeaters and connect direct to the hub, which is weird.
Many devices will choose the hub if they can connect to it directly, since by definition that’s the route with the lowest number of hops. So that part’s not unusual.
So just to refresh the memory about the Zwave part.
It is not IKEA’s issue. It has been highlighted during the Beta firmware test and it still has been released. It is not only effecting ST, many other platforms faces similar issues. It is somewhere down to polling the sleepy devices, what keeps the device awake and drains the battery. Not only IKEA devices affected, but on some platforms Philips Hue remotes as well and who knows what else. (It is really somewhere down to the Zigbee stack and should be fixed by the manufacturer of the Zigbee module and stack.)
And any protocol being Z-Wave or Zigbee is fully relies on quirks. Have you seen the Zigbee certification compliance documentation? You would be surprised how weak is that. It is more unfortunate that none of the producers has fully available documentation how their Zigbee or Zwave devices are working. I read some issues about Qubino devices and their firmware. It is just crazy how much they changed between versions and the old ones hasn’t even got OTA to update. So everything is about quirks, if not then you can throw your compatibility list into a bin, when you are leaving the building.
(Some platforms realised this, some didn’t.)
Just as an example for quirks, have a look at the SmartThings Arrival Sensors’ DH, there are two of them. The devices look identical. But really, look into the old one’s DH (not the HA one). You will understand what quirky means when you reach the beep part. Zigbee has Indentify option. @JDRoberts might can tell since when. Or just the battery reporting… Yeah, compliance…
I am more and more curious what certification and compliance mess Matter will bring. We will see it in a few years. When old is going to meet the new at that time.
I vehemently disagree with this. Theres zero way it ISNT Ikeas problem.
In tech interop, this happens all the time. Vendor A has a minor problem with thier code but it works ‘fine’ because the spec ignores whatever the true issue is. (issue is there but masked or ignored) Vendor B updates thier code to fix another issue or increase interop or compatibility or compliance and the changes are well within spec but something triggers a problem in Vendor A’s code because thier issue has always been there but up till now it hasnt been a problem…
Who’s issue is it to fix.?
Is it A because they have the problem? Or is it B because their change ‘broke’ A’s code?
Answer: This is Vendor A’s problem. Every. Single. Time.
Doesn’t matter that B was the trigger… Its A’s problem to fix. Lawyers would agree (and yes it matters because this stuff isnt cheap. Developers cost money and theres real costs associated with this stuff…)
In this case because Ikeas’ code is broken and exhibits the issue outside of SmartThings and on multiple platforms is the indicator - theyre out of spec. Its thier issue. At the very least the burden of proof is on them.
As this kind of triage used to be my job… I can almost GUARANTEE without even talking to someone in thier support org this is the stance SmartThings has taken internally. Burden of proof is on Ikea. (im not paying for fixing something that isnt mine to fix) ST will cooperate with Ikea helping them figure out what is happening but don’t expect any code changes from them.
If im wrong i will happily eat my crow - serve it with ketchup… 20+ years in enterprise commercial software support says Im not wrong this time. Call Ikea and bug them. Not SmartThings.
This issue surfaced recently on many platforms and nobody can really have a grip on it.
It used to happen with old firmwares as well years ago, absolutely randomly. You used the devices for months without any issue, then you change battery and it started to eat a few of them, then worked again normally for months. You can search for it. The consensus was poor quality control from IKEA and to return it. But the number of incidents has started to increase a few months ago.
There are some other topics which provides different information regarding the issue.
It happens with Sonoff ZB bridges running Tasmota, it has happened with Deconz USB dongles as well, also with other ZB platforms.
I do not want to argue about this, but following your thinking in the previous occasion it was Fibaro’s fault, and all the others who has implemented the device firmwares.
This is a different platform but the issue has happened with Philips Hue remotes as well.
Some comments suggests that it is an NCP firmware issue of the coordinator.
By the way, talking about IKEA bugs, have you ever heard about this one?
It is again a base firmware issue for the devices, but it is coming from the manufacturer of the chipset.
I wouldn’t be surprised if in this case the Silicon Labs chipset’s firmware would be the issue as well. All seems to converge to Silicon Labs chipsets, and the v2 Hub definitely uses one.
Just moved my personal 2.4ghz router to CH11 and moved the ZigBee to CH11. The problem for sure is still present. I do live in an apartment so I’m pretty much bathed in every WIFI channel out there. I used a scanner app to determine quietest, but really its still pretty loud on all the channels here.
I don’t know about anywhere else, but in the UK we tend to have an ISP supplied combined AP/Router/Modem and the more enthusiastic ISPs use remote access to it. My one uses some annoying software that automatically ‘optimises’ your Wi-Fi configuration, which in my case meant moving my carefully chosen and manually configured Wi-Fi channel right on top of my carefully chosen and manually selected Zigbee channels. Fortunately they are perfectly happy to disable it. Just something to watch out for.
As I have said previously, I wasn’t expecting that to make any real difference other than perhaps causing the mesh to form differently, and perhaps mitigate the problem in some way.
Took out from the drawer the ikea hub, and paired one IKEA dimmer switch and outlet. Also, paired directly (without any hub) an ikea switch and outlet. A week later both combos are working with lightning fast response, no issues of any kind.
Changed the ST hub channel to 11, paired the above dimmers again to ST (they received an update once connected to IKEA hub) - the battery problem is still there.
All the other IKEA tradfri devices I have (bulbs, drivers, motion sensor, outlets, repeaters) are working fine with ST.
And yet, the biggest issue I see is the lack of official public information from IKEA and Smartthings (at least I couldn’t find any). We are talking about products that brings millions as profit to those companies, not an open source solution that relies on the community for support and testing. It’s been more than two months since the problem occurred, and I just can’t believe the company responsible for the mess (doesn’t matter which is the one) didn’t realized that and didn’t offer a solution yet!
We have to stop thinking as a fancy gadgets beta testers, and start acting as paying customers.
Not sure about that… My Fyrtur Blinds continue to work with ST But the battery life on these are impacted too. These used to last 6 months on a charge, now they drain in about 2 weeks.
I agree completely with what you’re saying, however it also becomes a ST problems as IKEA products are supposedly meant to work natively with ST (rather than via custom/community groovy handlers), and also presumably if IKEA fix this via device firmware update we’re still relying on ST to deploy it to hubs.
I’m not expecting ST to fix, however I would expect ST to put pressure on IKEA to fix, and deliver to us when it becomes available. I think it’s important anyone experiencing the issue open a ST support ticket.
Anyone opened a support ticket with IKEA?
Just wondering is anyone monitoring IKEA or is there some other website that reports what the current firmware is for the buttons and the blinds? How will we know if IKEA ever fixes this?
New to this forum and just read through the comments on this post since I have the same problem. Sot sure but I might have found something. If someone else can try and verify we can be sure. What I noticed was that my buttons, like the rest of you, just lost connection. Various pairing they only lasted minutes or even hours in some cases. Though consistently they at some point lost the connection. Dustin my last pairing I noticed that the button actually worked when the back lid is off. Weird but as I put the lid on it in some cases just turned the pairing led off and also made pared buttons useless again. Then I just thought of it. The magnet is a neodyne and they are quite strong. Maybe that’s the issue? So I removed the magnet (sits clamped in and some adhesive on the back but removes easily with a flat screwdriver) on two buttons on opposite side of my house. 24h passed and they are still working. Tried a 3rd. Same with the pairing being disrupted and stopped working as soon as I put the lid on(magnet still present). Removed the magnet and voila. Works.
So this might just be luck or something else at play but maybe someone else can try.
I see. Just to clarify and I forgot to mention is that the buttons I tried on are the 2-button ones. I haven’t tried this on the 5-button yet. And it’s still consistent. Every remote (2-button) that has the battery off still works. Put one magnet back and it stopped working. Has anybody had the time to try it? You also really don’t have to remove the magnet but just, for testing purposes, don’t put the back lid on while testing. That way It’s still intact and can be returned for those that want to do that.