Thank you very much Anton! I'll try all of these ideas :)

Also, I just talked to a friend who works for a company that calibrates
medical instruments. They don't have a calibrated 'UV emitter', but they do
have a calibrated UVA and UVB receivers (for babys that just born and need
UV light). So my idea is to use any abstract source of UV (even sunlight),
measure with a calibrated sensor and with mine.....and change my sensor
coefficient calibration until I get the same result as the calibrated
device. I hope this works!

About protection, you gave me a great idea..... I'll try to use one of
these PIR sensors.
[image: H9a8d776db26f4d1294d2be5175ecad93f[1].jpg]


Em seg., 8 de fev. de 2021 às 13:45, Anton vanNwnhzn@GMail <
[email protected]> escreveu:

> Jonis,
>
> My first suggestions are for 2 very pragmatic solutions.
> Calculation based on reported values for UVA and UVB incl. effects from
> reported cloudcover is a solution in completely different league, for which
> perhaps other group-members may help.
>
> 1) Perhaps *for testing of feasability* *start simple with "correction by
> calculation"*
>
> *Be aware that openuv-values are always for 'clear-sky = without any
> attenuation from cloudcover!*
>
> Based on your reported rough orders of magnitude
>
>    - sensor-output reads UVI=19
>    - openuv_value reads UVI=14 for 'clearsky', with OWM-value at same
>    level
>    - TV tells UVI=11
>
> the order is TV < openuv_value < sensor_output
>
> With that knowledge you might apply a *very simple & very rude* *
> scaling-function* like
>
> *Step 1*, to check if sensor_output already below openuv_value (because
> if seeming 'practical', why rescale?)
> if sensor_output < openuv_value then
> real_UVI1 = sensor_output
>
> *Step2*, checking & reducing the value to get below openuv and fitting
> 'reality'
> if sensor_output *>* openuv_value then
> real_UVI0 = sensor_output * 14 /19
> real_UVI1 = real_UVI0 * 11 / 14
>
> Step1 just checks that sensor-output is plausible,
> while Step2 is the equivalent of  taking 50% of the sensor_output
> Because you report that at night UVI from the sensor already is 0, no need
> to check on negative UVI, nor check on range of values.
>
> *Result:* the calculation produces a UVI which *resembles* the value from
> TV
> ;-) Absolutely not scientifically accurate, but a practical solution
> .......
>
> 2) *Empirical attenuation by cap*
>
> For a similar problem with measuring light with a LDR/photodiode, I used a
> deodorant-rollercap as attenuation cover over the sensor setup.
> A very empirical solution, but very easy to try, and ;-) no cost: see the
> pictures.
> Negative aspect is that the colour of the cap *might* affect the reading
> for UVI.
> With cap fitted, compare the reported UVI with the value reported by TV.
> If the reported UVI is lower than the value from TV, you might upscale the
> reported UVI by ratio.
> ;-) Also not very scientific, but a practical solution.....
>
>
> Op 8-2-2021 om 16:06 schreef Jonis Maurin Ceará:
>
> Ok, thank you guys!
>
> Here, in Ribeirão Preto, I'm getting almost everyday warnings on TV about
> extreme UV levels, so I'm pretty sure it's more than 11 almost all day. But
> in my opinion, 20 it's too high.
> I have tried both openuv.io and ogimet....
> openuv.io shows me:
> "uv": 0.041,
>         "uv_time": "2021-02-08T14:54:53.953Z",
>         "uv_max": 13.664,
>
> while my sensor is reading 19. Ok, it's a little cloudy now and maybe this
> measure from openui is during a shade.....still, max UV is a way low than
> my real reading.
> ogimet I could not find any station here in Ribeirao Preto with UV value.
> but looking at openweather map, I could find a value of 14 for UV. It's
> almost the same of openuv.
>
> Now is the complicated part.....
> 1) What do you recommend to use as protection for the sensor? I read about
> teflon diffuser (with 0.4mm thickness), but It's hard to find this. I was
> thinking about printing a case and use a mobile phone glass
> creen protection (very thick) as protection. What do you think?
> 2) I'm getting values from UVA and UVB from sensor in mW/cm². if I'm
> right, weewx expect data in W/m², right? If I convert my current UVA value
> (8,809mW/cm²), I'll get 8,8W/cm².....and converting to meters, it's
> 0,00088W/m². Is this right? I'm thinking its too low.
> 3) Finally, I have sensor coefficients to play and calibrate values. These
> coefficients are 4 values, a,b,c, and d. B and D are for UV light itself,
> if I remember the tech docs that I read. My question is: since UVI uses UVA
> and UVB to be calculated, I'll have to play with both. Do you guys have any
> idea or type how to start calibrating this without any expensive equipment?
> Of course I don't expect perfect values, since this is just a hobby. But I
> would like to get as close as possible to real value, using accessible
> things to calibrate this sensor.
>
> I know that this is a hard and complicated task to do, especially without
> adequate equipment. But I would like to have some ideas from you, more
> experienced users.
>
>
>
> Em seg., 8 de fev. de 2021 às 04:58, Anton vanNwnhzn@GMail <
> [email protected]> escreveu:
>
>> For clear-sky reference values you could look at https://www.openuv.io/
>> That site provides info dedicated for the location you ask:
>> info includes much more than UVI only.
>> For automatic extraction of info a script will be needed.
>>
>> If you want automatic info on attenuation by clouds, then
>> https://www.ogimet.com/home.phtml.en is your friend:
>> look for it's octa-values for your location [=WMOID] through the url
>> "http://www.ogimet.com/cgi-bin/getsynop?block='..WMOID..'&begin='..UTC..'"
>> <http://www.ogimet.com/cgi-bin/getsynop?block='..WMOID..'&begin='..UTC..'>
>> Or look for a weathersite in your area providing values for cloud-cover.
>> With that info [= 'clear-sky'+ 'octa'] you are able yourself to calculate
>> the expected practical UVI.
>> Op 8-2-2021 om 1:10 schreef [email protected]:
>>
>> Hi!
>>
>> I would like to know if these values are plausible for UV sensor here, in
>> Brazil.
>> I'm located in Ribeirao Preto, SP, Brazil (-21.1767, -47.8208) and the
>> sun is very 'hard' here. We are almost in tropical line and the sun pass
>> almost 90º degree above us at 12h00 pm......also, we are just in the
>> summer. So my question is: UV Index of 18~19 is a plausible value? peak of
>> 20.
>>
>> I'm asking because i'm testing a new sensor in my station,
>> VEML6075....but I'm first trying it standalone using arduino and this
>> library ( https://github.com/adafruit/Adafruit_VEML6075 ). Acording to
>> this library,m the default values (calibration coefficient) are for the
>> sensor direct in the sunlight, without any protection. That's how I tested
>> today, sensor direct in the sunlight....but I'm thinking that 18~19 are too
>> high, even with our 'brazilian sun'. It's very hot here, about 38~40 in the
>> shade.
>>
>> If I use the sensor indoor at night (just using my LED home lights), I
>> got negative values in sensor (UVA and UVB, not index), but I think that's
>> right, since LEd doesn't emmite any UV light and sensor has calibration
>> values, right?
>>
>>
>>
>> Here are the values that I got (just some samples). The lower value is
>> during the shade of a cloud, but still very bright.
>>
>>
>>
>> [image: 3.jpg]
>> [image: 1.jpg]
>> [image: 2.jpg]
>> [image: 4.jpg]
>> --
>> You received this message because you are subscribed to the Google Groups
>> "weewx-user" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/weewx-user/4f966f35-f76a-4275-8212-f54161fbae41n%40googlegroups.com
>> <https://groups.google.com/d/msgid/weewx-user/4f966f35-f76a-4275-8212-f54161fbae41n%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>>
>> --
>>
>> ===============================================================
>> Contactinfo voor Anton van Nieuwenhuijzen:
>> Email    = [email protected]
>> Fax2Mail = (+31/0)84.8397303 [ook Voice2Mail]
>> ===============================================================
>> Deze E-mail en eventuele aanhangende files zijn
>> alleen bestemd voor de geadresseerde(n).
>> Als je deze E-mail ten onrechte hebt ontvangen,
>> dan aub verwijderen en de afzender informeren.
>>
>> --
>> You received this message because you are subscribed to a topic in the
>> Google Groups "weewx-user" group.
>> To unsubscribe from this topic, visit
>> https://groups.google.com/d/topic/weewx-user/zGNStvJwKKk/unsubscribe.
>> To unsubscribe from this group and all its topics, send an email to
>> [email protected].
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/weewx-user/9d542c1f-7218-7376-e898-b2ac9544fbe0%40gmail.com
>> <https://groups.google.com/d/msgid/weewx-user/9d542c1f-7218-7376-e898-b2ac9544fbe0%40gmail.com?utm_medium=email&utm_source=footer>
>> .
>>
> --
> You received this message because you are subscribed to the Google Groups
> "weewx-user" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/weewx-user/CACK8wihV6QWZu-0FPfesPfqrk%3DcX_hbV1tEn1h-6ssJi%2BZK3Aw%40mail.gmail.com
> <https://groups.google.com/d/msgid/weewx-user/CACK8wihV6QWZu-0FPfesPfqrk%3DcX_hbV1tEn1h-6ssJi%2BZK3Aw%40mail.gmail.com?utm_medium=email&utm_source=footer>
> .
>
> --
>
> ===============================================================
> Contactinfo voor Anton van Nieuwenhuijzen:
> Email    = [email protected]
> Fax2Mail = (+31/0)84.8397303 [ook Voice2Mail]
> ===============================================================
> Deze E-mail en eventuele aanhangende files zijn
> alleen bestemd voor de geadresseerde(n).
> Als je deze E-mail ten onrechte hebt ontvangen,
> dan aub verwijderen en de afzender informeren.
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "weewx-user" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/weewx-user/zGNStvJwKKk/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/weewx-user/fab4f4d1-8aeb-8c28-8dbc-58706f3545df%40gmail.com
> <https://groups.google.com/d/msgid/weewx-user/fab4f4d1-8aeb-8c28-8dbc-58706f3545df%40gmail.com?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/weewx-user/CACK8wij9v7UaGu-DAm%3DJpsNR_iszYp-_TycsnBVn3Rg0qku6-w%40mail.gmail.com.

Reply via email to