How is it testing the sensor?

Everything I've ever seen and heard backs up what Dan said. The test button on detectors (smoke and CO) simply tests that the unit has power, and the alarm speaker functions. Actual alarm testing requires real smoke or CO to be present.

The behavior you see when there is a delay is intentional, caused by a timed relay. This is to allow time to press the button and move away, to avoid being right next to the alarm when it sounds. Not all alarms have this feature though.

You seem to know a lot about what's inside the smoke detectors here. Except there's no relay, no click, and the test ends as soon as the button is released. And it really does test the sensor.

Some detectors "test buttons" work by inserting a barrier between the emitter (radioactive or otherwise) and sensor, thereby simulating smoke.

The ones here don't do that. Instead, pushing the button activates some circuitry under the sensor, which emits something (particles, smoke?) similar to what the sensor is looking for, causing it to flag an alarm. I cannot see exactly what that is, because it's inside the same assembly as the sensor itself, and I don't want to destroy by opening it up.

However, if I provide airflow across the sensor, then pushing the test button no longer triggers it, indicating that the particles/whatever it generates, are being dispersed before they can trigger the sensor. This "works" whether the air is room temperature, or from a heat gun.

Oh, and the ultimate test: "burnt toast detection" also still works fine.

Cheers

So.. googling for the model number (WNW777DC-A), I found a link which suggests it is a rebranded "Family Guard FG888D". The user manual for that device, says it works using "ionization technolgy." So I imagine the test circuit generates some of the appropriate "ions" for it to detect.

On a related note: anybody know how to de-Googlify a Google search result? The link above is shrouded in googlisms, and I really just wanted to post the true link to the maker's manual. How to do that?

Edit: found a Firefox plug-in for it.

Thanks.

And here's a page which explains some of the "ionization technology" used in smoke detectors, including in the $4.89 Walmart version of the one I have here:

http://www.angelfire.com/electronic/cwillis/rad/smoke.html

Wikipedia: http://en.wikipedia.org/wiki/Smoke_detector#Ionization

W says "A self-test circuit simulates an imbalance in the ionization chamber and verifies the function of power supply, electronics, and alarm device." .. but doesn't elaborate on how it "simulates an imbalance."

You didn't notice me taking one down to look at during the visit last year?

I have an extended family member who works in the industry, so some of my information here may be more specific to his brand of detectors. They also mostly specialized in CO detectors which did have pretty terrible useful lifespans early on. 7 years now is a big jump over the initial 2 years for CO sensors.

I did find this to explain what the test button does on an ionizing detector such as yours. http://diy.stackexchange.com/questions/24552/what-happens-when-you-test-a-smoke-detector There is also the lawyered up description in the patent for one method of testing these types of detectors. http://www.google.com/patents/US4456907

I believe your test results using moving air over the sensor may be due to this note later in the Wikipedia article comparing ionizing to phtotelectric smoke detectors:


Clearly the burnt toast test is proving something still works for yours though.

That's a good link, but not quite right for the detector here. On this one, pressing the switch closes a contact, rather than opening one. So perhaps the circuit has a reverse bias to that on the link you found.

The nice thing about this kind of detector, is that it is all solid state, apart from the radioactive emitter bit. So nothing to gum up or fail. If the radioactive emitter itself fails, the alarm should go off. But with a half life of hundreds of years, I think this detector has quite a bit of life in it yet.

Cheers

Hundreds of years? I'm not doubting it, but that would seem an odd choice bearing in mind the waste issue. Don't suppose you know what nuclide it is?

Peter

Thoughts:

- $129 is pretty steep, especially for newer homes like Dan's where there are a lot of detectors. However...

- I really like that it's a combination sensor. To combine heat, light, and motion into one sensor alone is pretty impressive. I know the price seems pretty steep, but I haven't seen a sensor that does those three things alone for less than $75, and this is also a smoke/CO sensor, which doesn't exist for any other HA systems for some reason.

- Isn't the current advice to use smoke detectors that use photoelectric AND ionization sensors?

- I love that it will hush with the wave of a hand. The oven in our old place made that a very desirable feature.

- These smoke alarms mostly eliminate one of my primary issues: that many people have thermostats in locations that they never walk by, so the motion sensor feature was useless up until now.

- I still don't think the "learning" feature would be worth anything for me, as my schedule is completely unpredictable.


I do think this is an exciting development, but I don't like that it's still very proprietary. We just don't know enough about where they're going with this whole thing yet, and that makes me a little cautious about going with their products yet. Like I keep saying, I have no idea who these Control4 people are, or how they fit into what Nest is doing. That's a big deal for me, because the most important part of HA for me is lighting control, and I don't know how Nest plans to work that into their system.

Nest is a really cool product, but I tend to like having a lot more control over my HA, and their focus is on making everything happen on its own. I suppose that fits more with the definition of "automation," but it's not really what I'm looking for.

The Wikipedia article suggests that americium-241 is commonly used, with a half-life of 432 years. Apparently banned for smoke detectors in parts of Europe, and for that very good reason.

Clearly, they're doing a lot of their engineering with an eye toward the future. The original Nest thermostat had an 802.15.4 antenna, and it was unused. Now, it's used to integrate with these smoke detectors.


By itself, I agree. Combined with a truly viable auto-away system, however, and it starts getting interesting. If these smoke detectors have a good idea that you're gone, then the thermostat can respond.


I'll be curious to see if other HA vendors start supporting the Nest APIs. I'll also be curious to see what Nest has planned next. The three obvious choices are door locks, light switches, and home alarm systems. All get away from their one-size-fits-all product strategy. There are tons of different lock designs, and there are a staggering diversity of different light switches, never mind that any attempt to really rethink light switches should also include rethinking the lights themselves, now that some lights can change color. Home alarm systems are probably the most likely next step. If you've got enough motion sensors around the house, then you don't really need window open / breakage sensors. All you need is a cellular outbound data path as well as some clever new way to figure out when to arm and disarm the alarm system. (Maybe integration with fancy new door locks?)

Actually, I'm not sure it is. I believe they use the wifi for that, based on talking with the nest guys in the pub last week. Seems that the zigbee chip they asked me to design into the original nest is useless

(I was ribbing them about their prehistoric wifi setup method, and asking why the nest protects couldn't get the wifi setup information over zigbee from the thermostat)

Interesting. At least my WiFi will continue to connect everything together.

That's a big mistake. From personal experience, WiFi coverage in people's homes isn't always very good. Is ZigBee a mesh protocol like zwave is? That's the strongest part of the zwave system, that all the devices can talk to each other. I have many issues with zwave, but that's the best aspect.

Zigbee does mesh, yes, but if your devices are battery powered, they won't improve the mesh - they'll just be leaves. The power consumption to be a zigbee router node (or whatever it's called) is not compatible with being battery powered.

Nest protect is also not really Zigbee, or anything else standard, from what I hear. It uses an 802.15.4 radio, but so do lots of things.

In any modestly recent construction, fire alarms should have AC power behind them, allowing the Nest Protects to set up a proper mesh network amongst themselves, whether or not they also mesh with the Nest Thermostat. In my house, I've got good WiFi coverage everywhere, but in houses without that but with enough fire alarm density, the mesh network should do the job.

Curious thought: these Nest Protects have all the necessary hardware to become WiFi APs, possibly spreading service throughout a house using the mesh network as a backhaul.

Indeed a curious thought, have you seen the type of bandwidth you get on 802.15.4?

Yep, I'm sure the powered ones do some sort of meshing, and the unpowered ones may also get in on the action in case of an actual alert situation - might as well use that battery up before it's toast, eh?

I'm not sure if you guys saw Revolv on The Verge, so I'll post it here. Looks like a good solution for bridging the void between these various home automation products (which has been my main complaint about all of them so far).

But it's irrelevant to me because they don't support Android. Ugh.

*edit*
And when I say various, I mean Nest and Hue and all those new siloed automation products, as well as ZWave, Zigbee, and Insteon. That's the part that got me listening...until I saw they don't care about me...

Just spit ballin' here:
Given that the Nest Protects are hardwired (yes, I'm ignoring the pure battery ones) could they run ethernet over powerline in addition to the smoke alarm to smoke alarm communication. Then the Nest Protect with the strongest wireless signal would bridge with that network and the other Protects would be come wireless access points.

Having said all that, it would be a shift away from the core product offering, and likely increase the number of support calls, for minimal gain in revenue.

Double or nothing: whole-home audio support! Build in a modest class-D stereo amp chip and you could rig the Nest Protects to support in-ceiling speakers. Support Apple Airplay and/or whatever other standards are out there.

Google to Acquire Nest

$3.2B in cash? Nest must have a lot in the pipeline beyond the current thermostat, smoke alarm, and coming alarm system integration...

Odd. The first time my thermostat or smoke detector tries to share stuff on Google+, I'm done with them.

I hope Hugo got paid

You are OK with Google knowing how you live, what your life patterns are inside your own home?

With occupancy sensors in multiple rooms, including the sleeping areas and kitchen, this feels something like an in-home panopticon.

I'm not really okay with that either. But I'm not paranoid about it. I appreciate the benefits Nest has brought me in the past. Assuming those continue, I simply see the Google buyout as one enabling Nest to continue on.

Google's ad business and push into G+ bothers me more, simply because I've detached most of the other Google hooks into my life. They've been minimal anyhow, as I never really trusted them with my e-mail, documents or much else.

My two times of trying to adopt Android creeped me out too much to ever want to go back. Mostly because their engineering focus has them blind to the privacy issues at times, not out of malice to spy on everything I do. (well, outside spying to provide me "relevant ads").

Which reminds me. If my future Nest ever pops up an ad for an air conditioner on it's screen, that would also be a breaking point.

Yeah, that would be very weird.





I don't see what they, Nest or Google, would have to gain from such sharing.

That was my first thought when I saw the number, too.

I'm not going to argue with the Google privacy fear, which has been the main reaction to this news. I know your minds are set on that issue.

But I'd appreciate if anyone could take a break from that usual tune to discuss the potential positives of this acquisition. It would be refreshing. I'll start:

As a home automation enthusiast, I never once considered getting a Nest thermostat or smoke detector. I thought they were great products, but it didn't fit into my vision of home automation. I saw Nest as part of the obnoxious trend of recent home control gadgets that were siloed off from each other and had no way of speaking to one another. That isn't home automation to me.

I admit, I have no way of knowing whether this acquisition will change any of this, but I have more confidence that Google might make home automation products more accessible to the end consumer than any other company before them. So far the market has been a mess.

I guess the question is if google/nest will follow and adopt the emerging industry standards or go their own way.

Matt, what is your vision of home automation?