Autonomous watering DIY timer with Sonoff+esp_MQTT

[Marco Sillano]

I'm gad to inform SonoffUsers about a new project: sonoff_watering (https://github.com/msillano/sonoff_watering) with complete DIY instructions.

Some interesting features:

• Autonomous: contains an MQTT server with the operating logic (script)..

• Compatible: can be used with standard MQTT client and MQTT server, has a complete set of commands and information via MQTT

• Autostart: preserves status and configuration in flash ROM: in case of reset or blackout restores the previous status autonomously.

• NTP-client: requires connection to the domestic WIFI, with access to the Internet, to have the correct time. Commands for summer/winter time.

• Serial console: for installation and debugging, via WIFI (telnet) or COM

• OTA update: the operating logic (script) can be updated via OTA (Over The Air), without moving or opening the timer.

• MQTT client: currently 2 options for monitoring and configuration:

  • Full client on PC (node-red)

  • Reduced client on Android (IOT MQTT Dashboard app)

• Power supply: (90-250) 220V AC, 220 V solenoid valve (attention !!!)

• Stand-by consumption: not measurable (<1 W)

• Consumption ON: 6 W (220 V)

• Low cost; less than € 15 (US$ 18)

• Case: waterproof (P66) and transparent, 3 LEDs: power on, irrigation on, wet soil
  
  

see https://github.com/msillano/sonoff_watering/blob/master/watering-sonoff-en01.pdf

Enjoy.

[Phil]

Nice one !, I have only had a glance over but you have done a grand job there..  having had a little experience with similar h/w tho i would suggest you consider ditching the sensor you have used and look at something that uses capacitance to detect moisture. https://tech.scargill.net/plant-watering-fiasco/  As come winter you may be replacing the soil sensor due to corrosion and a subsiquent loss of reading sensitivity.  Tho tbh a new sensor every 6 months or so isnt the end of the world.

Also with a torch style butane lighter you can 'braize/solder' using plumbing solder and FLUX,  SS capillary tube to the sensor legs to extend both depth they read and life ;)  

if you opt for the £10 xaiomi bluetooth sensor a raspberry Pi-0-W is a more than capable host to run the following project or similar to query the sensor and publish its reading as an mqtt topic/payload

https://github.com/ThomDietrich/miflora-mqtt-daemon  btw they work brilliantly ;)

[Marco Sillano]

On Saturday, July 28, 2018 at 7:26:19 PM UTC+2, Phil wrote:

Nice one !, I have only had a glance over but you have done a grand job there..  having had a little experience with similar h/w tho i would suggest you consider ditching the sensor you have used and look at something that uses capacitance to detect moisture. https://tech.scargill.net/plant-watering-fiasco/  As come winter you may be replacing the soil sensor due to corrosion and a subsiquent loss of reading sensitivity.  Tho tbh a new sensor every 6 months or so isnt the end of the world.

 

Also with a torch style butane lighter you can 'braize/solder' using plumbing solder and FLUX,  SS capillary tube to the sensor legs to extend both depth they read and life ;)  

 

I agree with you.

I maked many reflections about that:

1) in this project is not question of  "reading sensitivity", but only a rude dry/wet check: I modified the circuit, to add strong hysteresis.

 So I'm curiosous about probe life in this conditions.

2) the probe is passive (i,e, it don't have electonics) so it is easy to replace it with a mammut and 2 stainless steel bars.

3) I'm study a new version, where the sensor is powered only short times, when the moisture is checked in polling. That must extend probe life.

if you opt for the £10 xaiomi bluetooth sensor a raspberry Pi-0-W is a more than capable host to run the following project or similar to query the sensor and publish its reading as an mqtt topic/payload https://github.com/ThomDietrich/miflora-mqtt-daemon  btw they work brilliantly ;)

I have and I use 3  "Flower care" Xaiomi +  Raspberry but for a different project. 

You can try to evalute the cost (and size) of: xaiomi + raspberry Pi-0-W + power-supply + relay + relay-driver + waterproof case.....

Software considerations apart: this project uses only a script (4K) OTA updatable.

That I propose here is a simple replacement for mechanical or commercial watering timers. So it must be cheap, simple and autonomous, with low power consumption and low water consumption.... i.e. eco-friend  :).

Best regards

[George Ioakimedes]

I agree with Phil, very nice write up and full of details, thanks for sharing