purple air block

[Torjan]

Hi!
can we remove the temp and hum from the new purple air block. The values are far too way off.
This is not just me, think it applies to all purple air sensors

Regard Torjan

[Jachym]

I guess yes.

Just a little provoking question - do you think the emission measurements are any more accurate?

[Torjan]

Hi!
Thanks that will be fine.
If it measures right values on air quality I do not know, I do not have anything to compare with that.
I have on temp and hum.

Regard Torjan

[Jachym]

My opinions might be radical given the fact I work with the pro stations, stations worth 20-30 thousand dollars and Ive seen the comparisons to these cheap sensors. Some of them show total nonsense, some of them measure with a 20-50% error at a very specific temperature and humidity range and outside that range it is again just a random number generator - unfortunately - of course I would be more than happy if I could use these, buy 100 of them instead of 1 of these expensive ones and have hundred times more stations and much more accurate models and imission concentration maps. However it just isnt possible to do. The technology is different and not accurate.

[giordolo]

I just observe the right correlation with the pollution map provided by Windy (monoxide and not PM) and the situation measured by Purple Air. This usually works fine.
I'm not a tech but I agree with you: for sure it is not a professional sensor but I hope it is good enough for a personal and hobby usage (like the weather stations we are using, I mean)
cheers
giordolo

[i_fiorentino]

Hi,
i've just installed purple air block, thanks Jachym!

I've two id (any idea?) but i've seen that purple air block shows only pm2.5 and not pm10 value?
Is it normal?

How to show both values?
Regards,


Alessandro

[Jachym]

The API Ive seen only had PM2.5, so Im not sure if it even measures it and if yes, what the output looks like

[ozgreg]

The API Ive seen only had PM2.5, so Im not sure if it even measures it and if yes, what the output looks like

The bright boys over in WxForum figured out how to get the PM10 but it is not straight forward

https://www.wxforum.net/index.php?topic=33482.0

I second removing the temperature etc as the sensors are inside the unit and thus are wildly incorrect, I am not sure why they added them as having them inside a unit with a heat source (laser) is going to make them inaccurate and they are..

[Jachym]

I see, although let me say again that the air quality sensors are probably even less accurate than the meteorological ones

[ozgreg]

I see, although let me say again that the air quality sensors are probably even less accurate than the meteorological ones

I think that goes without saying The AQI certified sensors have a string of 0's after them..

The AQMD did review the current (PA-II) device against one of the $20,000 sensor and it performed very favourably but was proven to have a 30% error margin against the $50,000 unit (When exposed to heavy particles)..

I just removed the php code for the temperature display out of the block , maybe if your in the mood adding a setting to turn off displaying temperature and humidity would be a great outcome..

What is very interesting is the local json page for the device which now exposes the 0.3, 0.5, 1.0, 2.5, 5.0 and 10um values..

Again thanks for coding up the block

[Jachym]

Its definitely not 30% +/- - this is what they tell you, the problem is that this only holds for a specific Temp/Hum combination. Otherwise it is completely off. What is also true is that all these comparisons are done after the device is manufactured and given there is no calibration (which normally occurs regularly on pro stations), the device drifts more and more. I wont go into it, but I have some data as well, which I cannot publish here, but which analyzed similar devices or even PAII

[ozgreg]

From what I understand the designer put the temperature sensor chip on the same board as the twin laser samplers.. I am not an engineer but it is pretty clear to me what impact a pair of lasers are going to have on the temperature sensor

It is a cute little device, what is strange is we hear very little about AQI readings and from what I been seeing once it gets dark we still got a number of houses in this area that have fires and the effect is very noticeable if it is still outside...

[Jachym]

There are times, especially at the beginning of its lifetime, when the device will be partially accurate and spot the trends, however it is totally unpredictable and thats the problem, you never know if just now it is accurate or not.

With regards to the temp sensor - that really is a very dumb idea

[tpweather]

I'm a bit confused which is easily done. However, on the purple air map the temperature matches my weather station. But on the purple air block it's way out. Any ideas?

[livejerseyweather]

I have got the same issue with mine too as the temp is out as is the barometer too. Seems to just be on the block and not on the purple air site.