Weather Station Ideas

[uhb1966]

1 hour ago, Gina said:

Reckoned to be about the best weather station is the Davis Advantage Pro 2 and I can see advantages in the wind sensors in this station.  This format is much more compact than my last version.  There is very little interference from the support bracket.

Wow ... a 400+ squid weather station! Oh boy ...🤯 Mine is definitely more frugal...  https://www.froggit.de/product_info.php?info=p316_wh1080-se-profi-funk-wetterstation-solar-touchscreen-usb--neuer-aussenmast-.html

...but i fire my money out for cars  

Regarding the BT/Wlan question: I'd also go for Wlan if possible - you can then beam the data over longer distances.

 

 

Edited June 26, 2020 by uhb1966

[skybadger]

18 hours ago, Gina said:

These? :- IZOKEE 2pcs ESP8266 ESP-01S Serial WIFI Wireless Transceiver Module with 1MB Flash Support LWIP AP+STA for Arduino UNO R3 Mega2560 Nano

I would be very interested in how you made your weather station.

That's the badger! but using the -12 version due to having more pins. My git pages have a bunch of Alpaca drivers on them that use the same general code pattern. The change is adapting for the sensors. I'd have thought you'd be up to building an ultrasonic sensor for speed and direction in one hit rather than remake the delicate propeller parts. 

[Gina]

Do you have a link, please?

[Gina]

My current wind vane design uses magnetic damping with strong magnets and an aluminium disc but with Consensus Averaging this seems unnecessary and indeed may actually be a disadvantage.  I was planning to us Consensus Averaging anyway and have used it before.

[Gina]

I'm now thinking about the angle sensing for the wind vane.

Possibilities :-

1. Magnet and read switches - some magnetic force
2. Magnet and Hall switches
3. Reflective optical sensing with Gray code disc
4. Transmissive optical sensing with Gray coded discs
5. Magnetic encoder chip

Hall switches instead of reeds avoid the magnetic attraction of reed switches providing freer vane movement but experimentation would be needed to see how well this worked and what overlap occurs.  Optical sensing provides analogue output with code needed to clean up and provide Schmitt triggering.  This is no problem with the Arduino.  The Melexis magnetic encoder chip system has proved less accurate than expected and is difficult to set up.  I don't like it!

Option 3. is probably the easiest to implement mechanically but relies on getting a good differential reflectivity from the Gray code disc.  Some experimentation to get the switching levels right would be needed.  Option 4. give better light to dark ratio but is more complicated mechanically.

Whichever system I use seems to need experimentation.

[tekkydave]

I have had one of these inspeed eVane wind vanes for many years now. Claims to use a Hall effect sensor but I'm not sure how they get the analogue output as a proportion of the input voltage. It works great with 1-wire AtoD to give a very accurate reading to less than 1 degree resolution. It's a sealed unit so I'm in the dark about how it works internally.

Edited June 27, 2020 by tekkydave

[tekkydave]

Maybe it uses one of these.

[Gina]

Thanks Dave.  Interesting but a) I don't want to spend that much money and b) prefer to make my own.

[Gina]

Rummaging through my box of chips and other components I've found the transmissive opto sensors but not the reflective or the Hall switches which I know I have somewhere.

[tekkydave]

1 minute ago, Gina said:

Thanks Dave.  Interesting but a) I don't want to spend that much money and b) prefer to make my own.

Appreciate that but I thought those sensor chips look interesting to have a play with 😮

 

[skybadger]

Link to my github for ESP8266 code : https://github.com/skybadger

The maplins weather centre uses this scheme: https://www.argentdata.com/catalog/product_info.php?products_id=145

If I wanted high resolution wind diiretion sensing I'd probably go for a s cheap HP 400ppr HEDS encoder. Actual resolution is 1200 so loads of data but you might suffer from missed edges if it rotates fast. 

THE ESP8266-01 has 4 io pins , of which 1 is TX out for debugging, Rx is used as GPIO and GPIO0/2 are used for i2c to the ADC and BME280.

Which means if I want to count anemometer edges I need an extra pin, same too if I want to count water bucket closures. Which meant I went to the ESP8266-12. I am using the adafruit huzzah ESP for this and currently boxing it up to fit inside the old maplins enclosure, which is where the challenge lies, the coding and parts all worked on the bench....

 

[skybadger]

Links to ultrasonics: 

http://embedded-lab.com/blog/making-an-ultrasonic-anemometer/

https://hackaday.com/2013/08/21/ultrasonic-anemometer-for-an-absurdly-accurate-weather-station/

https://www.dl1glh.de/ultrasonic-anemometer.html

The hardest part for me seems to be the resonant tuned driving circuit but I expect there are examples of them on the net too. 

[Gina]

Looks interesting but...  Does it need a faster processor than an Arduino or RPi?  What are the ultrasonic sensors - cost and availability?  I could manage the hardware.  If there was a design with readily available sensors and using either Arduino or RPi with sketch I would be interested but I suspect it't not that simple or there would be lots of them about and rotary wind sensors would die the death.

[MarkAR]

Wonder if there's any similarity in the components in ultrasonic cleaners.

I've done a couple of repairs on mine in the past and they basically use pairs of ceramic transducers oscillating at high frequency around 20k to 30k Hz i think.

The PCB also has a widget to fine tune the frequency for optimum performance.

[Gina]

I've ordered some more optical reflective sensors and Hall switches to replace the ones I've lost.  Arriving tomorrow.

ARCELI 10 Pcs TCRT5000L Tcrt5000 Reflective Optical Sensor Switch

10pcs A3144E 3144 OH3144E Hall SIP Three-Pin Effect Sensor

[Gina]

Been looking at using the ESP8266 with Arduino Nano.  I see it needs programming with things like SSID and WPA password first, which can be done from an Arduino.  Then I presume the Nano is connected to the ESP8266 to send data.  Haven't quite worked that part out yet.  I don't know if the ESP8266 has EEPROM to hold the SSID etc. or whether this needs sending on powering up in the field.

[tekkydave]

Why do you need nano as well as the ESP8266? Is it just a lack of I/O pins on the ESP?

[pete_l]

31 minutes ago, Gina said:

I see it needs programming with things like SSID and WPA password first,

The simplest way to do that is through a package called esp-link  That provides a web interface through its own (on the ESP8266) Wifi access point. Then all the config stuff is held in the 8266 n/v memory

Generally, the ESP8266 is rather old now. ESP32's have more speed, more pins, more memory (both for code and to store files) and more features, such as bluetooth. As far as cost goes, they are a little more £1-ish more expensive

Edited June 27, 2020 by pete_l

[tekkydave]

The ESP32 seems endowed enough for a weather station:

• 18 Analog-to-Digital Converter (ADC) channels
• 10 Capacitive sensing GPIOs
• 3 UART interfaces
• 3 SPI interfaces
• 2 I2C interfaces
• 16 PWM output channels
• 2 Digital-to-Analog Converters (DAC)
• 2 I2S interfaces

[Gina]

20 minutes ago, tekkydave said:

Why do you need nano as well as the ESP8266? Is it just a lack of I/O pins on the ESP?

Yes.  I thought the ESP8266 was just a WiFi adapter hadn't thought of it as a microprocessor as well.  I will want one digital and 4 analogue inputs or maybe 9 digital inputs depending on how I implement the wind vane sensor.  OTOH that encoder chip only needs 3 data lines (SPI or I2C forget which).

[tekkydave]

3 minutes ago, Gina said:

Yes.  I thought the ESP8266 was just a WiFi adapter hadn't thought of it as a microprocessor as well.  I will want one digital and 4 analogue inputs or maybe 9 digital inputs depending on how I implement the wind vane sensor.  OTOH that encoder chip only needs 3 data lines (SPI or I2C forget which).

ESP8266 only has one analogue input - you need the ESP32 I think

The ESP chips can be programmed via the Arduino IDE as if they were arduinos with the addition of a few extensions.

 

Edited June 27, 2020 by tekkydave

[Gina]

4 minutes ago, tekkydave said:

The ESP32 seems endowed enough for a weather station:

• 18 Analog-to-Digital Converter (ADC) channels
• 10 Capacitive sensing GPIOs
• 3 UART interfaces
• 3 SPI interfaces
• 2 I2C interfaces
• 16 PWM output channels
• 2 Digital-to-Analog Converters (DAC)
• 2 I2S interfaces

WOW!!!  That looks good.  I'll see how much they cost.  I though of using an Arduino and WiFi add-on since I'm very familiar with coding Arduinos.

[Gina]

The ESP8266s are already on their way to me!

[tekkydave]

Programming ESP32 via Arduino IDE

Note: Also works using Arduino IDE in Linux

Edited June 27, 2020 by tekkydave

[Gina]

£7.49 on Amazon.  I'll order one I think...  Just a bit more than 2 of the others.

Edited June 27, 2020 by Gina