We are seeing more and more wildfires along the west coast. Even if the fire is not near you the smoke that blows in can be debilitating. Though there are already air quality reports available they are too broad. The air quality differs from area to area depending on proximity to mountains, wind patterns and the like and can vary wildly within only a few miles. Knowing what the air quality is in your immediate area would be a huge benefit.
I have built a sensor from https://sensor.community/ and is attached to the pole holding the Tempest
2 Likes
I have one as well! works just fine.
1 Like
Iām located in Portland and agree itād be awesome to have an air quality sensor built-in or some integration with purple air or airnow.gov
You could of course to do your own integration into a third party dashboard such as influxdb+grafana or a more complete solution like weewx, but Iām guessing youāre hoping that WF comes out with an AQI sensor and integrates it with your weather data in ātheirā dashboard ?
FWIW, my purple air integrates quite nicely into my weewx setup.
I use PurpleAir which works great and my Tempest inputting into Home Assistant.
I also currently get alerts from the National Weather Service (USA) through their API and also with the Red Cross Emergency App on my phone.
As for AirNow, the maps and forecasts for my area are always under reporting air quality levels. The purple air map is a good place to start until you get your own air quaility monitor.
I actually like having both my PurpleAir and the Tempest which have some common sensors that allow me to compare how they are doing.
I personally think that when you cram too many components into something then failure rates raise exponentially. Edit: False an independent(ish) fact checker has determined this claim to be false
Is that true?
You could have 10 different units each with some sensor and each with their own power supply (for example) or you could have one unit with 10 sensors and one power supply. For sure the first setup will give you a higher rate of something failing. The second setup probably is less likely to fail, but when it fails you loose all the sensors. But still less failure rate.
Or see it like thisā¦ less components, so less failure.
So less components in your full setup means less failure. Now having more components in one of the devices will make that device fail more often, but your total setup is less likely to fail.
You comments are generally true, however with more components in use in a localized area there are more direct interactions that are not foreseen there is always some EMI from components that do effect others even if not to a noticeable effect. From personal experience, the power supply is what is affected the most or is the conduit between the components for unexpected EMI.
Usually consumers (and industry) want smaller things, which means smaller power supplies as in just big enough (think Lean, 5S, Sigma Six, right size, etc etc) to handle the power requirements of the components given an ideal temperature. With more components crammed in, typically this means less air space therefore less cooling capacity for the power supply therefore more prone to overheating. This is not always the case if you have a quality build but it is prevalent in productions.
These are āgeneralā statements and just a personal opinion (although based on experience).
As I mentioned putting more stuff (technical term) into one device, makes that device more likely to fail. (putting more stuff into the same enclosure, without enlarging the enclosure for better airflow makes it even more likely to fail). And that is what your experience is build on. But that is only true if you look at that single device. If you divide the electronics over multiple devices (probably in smaller enclosure), you need more electronics to do the same, so you are more likely to have something fail. Also having different devices, from probably different brands, makes more likely that there is some software problem when combining them into one interface.
(btw I have yet to come across a device that failed due to internal Electro-Magnetic Interference)
The best example (most common) I have of this, that has bit me the most is manufacturers using un-shielded sensor wire or un-shielded Ethernet cable (or using shielded cable and grounding both sides and causing a current loop on the shield). I will say that I work in a unique environment and have some equipment that is power hungry but the general components and ideas are the same. I have also had pressure and temperature transducers have faults (internal EMI effecting output); there is a reason 0-10A current outputs are used frequently due to less prone to signal interference, however, not every manufacture colors within the lines when they make (assemble) their equipment.
Servos (and drives) can be very prone to EMI.
I do agree with you that statistically more components spread over more power supplies should be more prone to failure (ie more likely to have at least one unit fail than a combined unit).
I like the tempest due to no moving parts. āSomeā air quality sensors have fans that aid the passage of atmospheric contaminates across the laser counter.
So maybe me adding āexponentiallyā was a bit over the top. There is always more stuff that people will want added (heck, some of it I would like myself), but I donāt like to fix something until it is broken.
If designed well and quality components are used then almost anything will work; however with the current landscape of manufacturing this is not the common approach. There are still good manufactures out there but most people want something that is just good enough and is cheap (both senses of the word).
Does my edit to my above post help?
I agree that unshielded cabling is a bad idea when shielding is needed, but most problems occur with those by external equipment. Which actually would be an argument to integrate as much equipment as possible :-). But I donāt need help, I just was struck by your remark that integrating more is causing a higher failure rate, which is only true for that single device, but not true if you are going to buy 10 different devices to cover the capabilities of that single device.
You canāt necessarily assume that a 10-sensor 1-power-supply unit is inherently more reliable than ten 1-sensor 1-power-supply units. It depends on the devices themselves and how good each component is. There is no simple one size fits all answer.
Iāve seen 1-sensor 1-power-supply units that work forever (example - my model-B pi with a ds18b20 temperature sensor).
Iāve also seen 1-sensor 1-power-supply units that are just not good at all, or that reflect manufacturing quality issues.
How many SPA have failed by now ? How many Air/Sky/Tempests have required replacement ? Is the failure rate going down as manufacturing or design issues are being worked ? Are they failing at a rate that is normal or excessive for consumer devices ?
There are also competing optimizations here.
Many (most?) consumers want one device that scratches their itch. A good example is the old Apple AirPort Extreme. Trivial to set up once and then walk away from it for literally years.
Many more professional techies want a set of federated systems that are each optimized for what each device does well. I take this approach, with separate cable modem, router/firewall, small switches, and access point. As technology changes (or something eventually fails) itās easy to just drop in a replacement for that one piece of the puzzle.
Regardless - the original āadd a AQI sensorā thing is more of a design question. Most such devices ādoā have a fan to get the air moving across the sensors. That theoretically should eventually cause a failure, but itās impossible to predict if the expected lifetime of the fan exceeds the expected lifespan of the other components in that unit. I know my Purple Air has a fan. I donāt know if the little arduino board in there should be expected to die before the fan. For $200 Iāll take the risk I guess.
1 Like
Would opt for an āadd-onā AirQuality device.
Arguments:
- full/physical integration is not required for most users.
Even stronger: a good siting of such device probably better different from Tempest-head. - still much development in that area (especially lokking into direction of measuring gasses), and a āstand-aloneā device enables independent upgrade.
Thanks to Amazonās impending entry into the flooded air quality monitor market space, the target price for devices is now under $70:
https://www.amazon.com/Introducing-Amazon-Smart-Quality-Monitor/dp/B08W8KS8D3/
Much like a dedicated LCD weather station console, it is pointless for WF-Tempest to develop a consumer-focused product when they canāt come close to Amazonās pricing thanks to their enormous purchasing power in large quantitiesā¦
oh my - thatās crazy cheap. If they had an API that would be a cool thing to add as an extension to go with somebodyās weewx driver 
2 Likes
Thanks for the link. That would be a fun project.
Butā¦the cost for parts is higher than amazonās delivered price (assuming you buy into the Alexa ecosystem from a security standpoint). Theyāre just steamrolling everybody.
(very happy with my much more expensive Purple Air outside, FWIW)
2 Likes
Thatās exactly my point on both devices that are popular with feature requests here. I can buy a Fire 7" tablet from Amazon right now for $34.99, but a raw 7" Raspberry Pi official touchscreen with no case sells for $73.00. Does it make sense for WF-Tempest to tell someone to buy a $35 tablet and run software on it, or for them to build a dedicated PWS console that will likely need to be priced in the ~$200 range at much smaller quantities? Now itās ditto for AQM sensors, with $70 being the price target that nobody without Amazonās huge purchasing quantities can ever hope to match.
FYI, I gave Alexa the heave-ho here after less than a week. I set the wake-word to ācomputer,ā which coincidentally is also the wake-word used on Star Trek. Made watching reruns painful, especially when Alexa started its self-destruct sequence or babbling pages of help information when triggered by the TVā¦
if you want cheap, go to aliexpress. Various types of air quality sensors for less than $25