upvote
I have HA send me a notification to ventilate my office when the air reaches 1000ppm CO2. The IKEA ALPSTUGA is often off by 300ppm even under 1000ppm. If I'd use it, I'd be getting notifications at 700ppm.

It is a thermal conductivity sensor, which is a very indirect way of measuring CO2 and is very sensitive to environment factors. You only get somewhat good readings in lab conditions.

Don't by the ALPSTUGA for anything but very rough trends, there are much better affordable options.

reply
Within 300 ppm is more than good enough. Realistically 1000 ppm is not that bad. The average meeting room is multiples of that.

Also in my experience it’s much more accurate than that.

reply
I notice that thinking becomes less clear when going above 1000ppm, so I let HA send a notification at 1000ppm. With ALPSTUGA it would send already at 700ppm. By the way, above 1000 the divergences become even larger due to the inaccuracy also being 10% of the ambient CO2 concentration (in optimal circumstances, probably larger IRL). So, suppose you want to be notified at 2000 ppm, the IKEA sensor might already do so at 1500 or 1600 ppm and it continuously drifts, so it's not like you can use a particular offset.

Besides that, what's the point? There are much better meters in a similar price class. As an additional benefit, they can last months or up to a year on two AA batteries.

ALPSTUGA is an inferior product.

reply
Can you recommend some?
reply
>With ALPSTUGA it would send already at 700ppm.

"oh no I am getting too much fresh air"

I get your point but come on.

reply
In some situations it means opening a windows, with big temperature drops when it is cold outside.

At any rate, this is really a weird discussion, because you can get far more accurate meters at similar price points. Why waste your money on a much worse meter?

reply
Can it work with Zigbee network or is Matter/Thread required?
reply
I'm using a bunch of IKEA's "smart home" stuff, all via Zigbee+HA, works great. Look for the Zigbee icon on the package, and the pairing for Zigbee vs their own home controller might have slightly different pairing sequence on the device, otherwise it just seems to work.
reply
ALSTUGA does not work with Zigbee.

They recently overhauled their lineup and replaced all Zigbee devices by Thread + Matter. Some of the new devices (mostly those who support TouchLink, e.g. some of the lights) have a secret pairing mode with which you can use them with Zigbee, but it's only a subset of the new products.

reply
> They recently overhauled their lineup and replaced all Zigbee devices by Thread + Matter.

Uuh, seems not keeping up with social media finally backfired. That sounds horrible! So far IKEA been a great experience when it comes to HA+Zigbee stuff, and I started buying stuff relying on they'd keep just keeping up with that, really sad to hear they've changed course.

The "secret pairing mode" stuff sounds the same as currently/before though, but they only do so for a subset is new and hope they again change their mind.

reply
Thread is significantly better. Zigbee relied on proprietary hubs and apps or hacky work arounds. Matter over thread devices don't need a brand specific hub or app. You can literally control the new ikea products direct from a modern iphone which includes a thread radio, no hub, server or app required.

If you already own the ikea hub, they secretly put thread radio in it which was just sitting unused in preparation for this range.

reply
It's complicated. Matter over Thread is indeed nice in that it you only need generic Thread and Matter servers. It also makes it easier to share credentials between ecosystems. Thread itself is also a pretty nice standard technically.

There are also strong downsides though, one is privacy and future cloud lock-in. Zigbee is fully local. Previous Thread standards added the option for NAT64 so that Thread devices can access the internet and there were some Thread + Matter devices that already require internet access for full functionality (IIRC some Nuki smart locks, but I might misremember). However, Thread 1.4 also adds support for Thread devices to get a globally routable IPv6 address. The Thread 1.4 whitepaper is pretty blunt about what this enables:

Simplified Cloud Integration: Thread devices can now seamlessly connect directly to cloud services, enabling remote control, monitoring, and over-the-air firmware updates.

https://www.threadgroup.org/Portals/0/Documents/Thread_1.4_F...

The fact that Thread and Matter are strongly pushed by Google, Apple, etc. should tell you enough.

Now, a TBR may simply allow you to disable NAT64 or globally routable IPv6 addresses (e.g. Home Assistant's addons), but many consumer implementations don't. E.g. the Apple TV is a Thread Border Router and does not allow disabling NAT64, so Thread devices can access the internet, send analytics, and can be cloud-controlled.

Also, the ecosystem is still pretty immature, as a result of which you can encounter issues, typically resulting in unstable device connectivity. E.g. TREL does often does not work well. Apple has some hacks to fix most of the issues, but it only works well between Apple devices. So it's generally the best to avoid combining multiple TBRs into the same network.

reply
> Thread is significantly better.

Better than what already exists and is deployed? I dunnno, hardware already in use always beat "hardware conceptually better but I don't have it", that's why Zigbee is better, for me. Protocols much like everything in the world, isn't correct/incorrect or universally "better", it's all down to use cases.

Personally, as someone who started to rely on IKEA providing Zigbee devices, Thread is obviously worse, because 100% of the devices I have are already Zigbee and not Thread.

reply
Ikea preemptively sorted this out by putting thread radio in their hubs years before rolling this out. There's also thread radios in the latest chromecast, apple tv, and loads of other products. If you have a single thread border router in your house from any brand you'll be able to connect to any thread device from any brand. Phones can also directly control thread devices without needing any network or hub.

It's a vastly better system and the transition period is so smooth because the smart home companies have been deploying the thread hardware for years before anyone started using it.

reply
smart home companies have been deploying the thread hardware for years before anyone started using it

Also worth mentioning that many modern Zigbee radios can also be Thread thread radios using different firmware. There are even multi-PAN radios that can do Zigbee and Thread at the same time. Some smarthome hubs use multi-PAN (e.g. Homey Pro), but it's generally recommended against now because of lower reliability.

The same applies to devices, e.g. some of the new IKEA devices work over Thread or Zigbee (Zigbee pairing is triggered using a non-documented sequence, presumably they added support for TouchLink). Or e.g. the Aqara FP300, which can be flashed with Thread + Matter or Zigbee firmware. It works because the same radio can be used for both protocols.

reply
This wasn't true for zigbee either. I used a zigbee usb stick with home assistant, could use any stick that was supported.
reply
Yeah, I bought a bunch of INSPELNING smart plugs when they were clearing out the inventory. The new GRILLPLATS switches are more compact though, which is nice.
reply
It’s part of the new range which is all matter over thread only. The existing ikea hub can do thread though.
reply
> The thing runs on usb 5v so the power consumption is negligible.

There’s a huge leap from that to the power consumption being low enough to be integrated into a smartphone, as demanded by OP.

reply
I don't think power use is the issue. I have this cheap CO2 sensor: https://www.domadoo.fr/en/devices/5882-heiman-zigbee-air-qua... which draws 0.5W. This includes thermometer and humidity sensor, Zigbee transmission, and acting as a Zigbee router, but it gives us an upper bound. It also measures continuously (picks up someone breathing on it within 10s), which is overkill. A phone could measure CO2 levels once every 10 minutes which would average under 0.01W, so that would work.

However, this assumes the sensor would fit in a smartphone, which is not a given. And these things need air flow. And they also wouldn't work while the phone is in a bag or a pocket.

reply
>A phone could measure CO2 levels once every 10 minutes which would average under 0.01W, so that would work.

Not sure about that, at least NDIR sensors have to be at certain elevated temperature to work and they do some preheating when you turn them on from standby.

So it's not possible to just measure less often as then energy would have to be spent on heating the sensor.

reply