I'm asking for your data logs of your growbox/growroom controllers

pahval

Well-Known Member
Hello there,

currently I'm working on my own custom growroom and growbox controller that I plan to make open source and available to anyone with proper scalability (from simple temperature control to web interface, VPD calibrated temp/hum/co2 control and light timers for normal, uv and ir lights), and I'm in need of data of your temperature, humidity, nutrient solution (ph, ec/ppm), and your relay activation data logs (environment control devices, light timers etc). I can work with any data type, and any kind of time length of collection (the more the merrier). Also, if you want, you can write to me what do you expect from a controller, what features you would like to have, how did your solution work for you (diy or commercial), what price did you pay for it, and in overall how did it improve/degrade your growing experience.

Thank you in advance,
Pahval
 

solakani

Well-Known Member
Interesting. Please describe your development hardware, software environment, timeline of project plan, your expertise in growing. If you are able to shape any data type then why not just generate data for all your test cases to produce the expected results.
 

pahval

Well-Known Member
Well, im interested in how temp and hum change over time, what is the overshoot of hitting setpoints with certain devices (example: when you stop dehumidifying it still goes under a lil bit more), what happens when you turn off lights (ppl say hum spikes) etc... i need this to see if i need to implement PID control or not... Im designing a specific system with 4 flower grow tents, 1 mother, 1 veg, but basically each grow tent is one master which has its own temp, hum, co2 setpoints, light schedule and nutrient regime... i made 3 successfull grows and 2 failed ones (30 grams and 0 grams), for medicine for my friend (she passed away)... currently im at my parents house (guess why) so growbox is no no for me, all the tests i do is with led diodes and hardcoded random *inputs setpoint values that user would change as it wants*...

setup (per growbox):

master - ESP32
8 channel relay board (humidifier, dehumidifier, intake and outtake hydroxyl generator (my own design), CO2 solenoid (vents inside are 24/7 so prob no relay)
DS3231 RTC
RS485 communication module

sensor node - Arduino Nano
SHT35 for temp and hum
SCD30 for CO2
MLX90614 for IR measurement of leaf temp
DPS310 for ambient pressure calibration of SCD30
RS485 communication module

light node - Arduino Nano
4 mosfets for white, red, infrared leds and UV dimmable ballast with needed voltage converters
RS485 communication module

nutrient node - to do

and then each growbox would be connected to openHab via rs485 connection (PJON protocol) that has phone app, web interface and data logging...

so far environment control is prototyped and tested, i am succesfully achieving rh and temp calibration inside vpd range, making uv schedule calculation so that process starts in the middle of growth, and starts in a way that user picks cycle number, minutes on, and minutes off, so that lets say in cycle of 12 hour schedule, choosing 3 x 10 minutes on, 3 x 10 minutes off uv lights get on at 05:30 of light schedule, and finish at 06:30 of light schedule, and ir lights after main lights go on for 10 minutes or whatever user chooses (can be during or after for emmerson effect, and also user can choose all of its own variables, or just low, medium and high reccomended values), and all that works in anytime of day and any duration (noob proof)...

next thing in project is nutrient doser, which has to filter with ro, dose each nutrient, measure ph and ec, and circulate until last 10 or 5% and then to drain, then wash plants and all tanks with h202 while new batch is cooking, then continue... for this part i need a prototype machine with solenoids and pumps, (cant do with dummy devices), i have final sketch (which ill prob change 3 times) but since corona my budget went into alaskan night temperatures range, so im messing with openHab for now, but also, waiting for 3.0 version and UI redesign to make the thing i want... that will give me web acces, UI interface and database logging...

point of machine is to have perpetual production of medical plants for medical users without too many job around its parts (im trying to achieve rez self-cleaning using only chem nutes and h202 but again, until i put this to test with legal plants i cannot say ill be succesfull at that), basically plug and play ready system of devices, so that users dont have to worry, and just pluck a leaf or two... i need those logs just to try to predict what happens inside growtents and grow rooms (when i grew i just had fans 24/7, soil hand feed so no data there)...
 

pahval

Well-Known Member
timeline? huh... cant say... im expecting about 2000 usd for one growtent testing setup, im earning about 750 usd per month, which leaves me with 350 usd to save, so i guess 1 year until its done...
 

solakani

Well-Known Member
Here is my log for one day from a dht22 sensor. An entry is made when variance is greater than 1. It seems like this data needs to be corelated with other devices to be useful. Will you show me how you proceed from here?

2020-09-01 03:17:39 INFO Temp=25.1*C Humidity=58.8%
2020-09-01 03:19:37 INFO Temp=24.8*C Humidity=59.9%
2020-09-01 03:22:33 INFO Temp=24.5*C Humidity=61.0%
2020-09-01 03:26:30 INFO Temp=24.3*C Humidity=62.1%
2020-09-01 03:26:57 INFO Temp=24.2*C Humidity=62.0%
2020-09-01 03:35:30 INFO Temp=24.0*C Humidity=63.2%
2020-09-01 03:50:22 INFO Temp=23.7*C Humidity=64.3%
2020-09-01 04:22:45 INFO Temp=23.4*C Humidity=65.4%
2020-09-01 05:04:17 INFO Temp=23.1*C Humidity=65.9%
2020-09-01 05:54:42 INFO Temp=23.0*C Humidity=66.5%
2020-09-01 08:18:23 INFO Temp=22.7*C Humidity=67.6%
2020-09-01 09:05:36 INFO Temp=23.0*C Humidity=66.5%
2020-09-01 09:11:16 INFO Temp=23.3*C Humidity=65.4%
2020-09-01 09:12:27 INFO Temp=23.6*C Humidity=64.3%
2020-09-01 09:13:41 INFO Temp=23.8*C Humidity=63.3%
2020-09-01 09:15:28 INFO Temp=24.0*C Humidity=62.2%
2020-09-01 09:16:31 INFO Temp=24.2*C Humidity=61.9%
2020-09-01 09:18:25 INFO Temp=24.2*C Humidity=61.1%
2020-09-01 09:26:19 INFO Temp=24.6*C Humidity=60.0%
2020-09-01 09:44:53 INFO Temp=24.9*C Humidity=58.9%
2020-09-01 10:24:32 INFO Temp=25.3*C Humidity=58.2%
2020-09-01 10:33:10 INFO Temp=25.3*C Humidity=57.8%
2020-09-01 11:19:03 INFO Temp=25.6*C Humidity=56.7%
2020-09-01 11:43:54 INFO Temp=25.5*C Humidity=57.8%
2020-09-01 12:44:31 INFO Temp=25.8*C Humidity=56.7%
2020-09-01 14:36:17 INFO Temp=26.4*C Humidity=56.3%
2020-09-01 15:15:13 INFO Temp=26.4*C Humidity=55.6%
2020-09-01 15:26:44 INFO Temp=26.3*C Humidity=56.7%
2020-09-01 15:36:13 INFO Temp=26.3*C Humidity=76.9%
2020-09-01 15:36:19 INFO Temp=26.3*C Humidity=56.3%
2020-09-01 15:46:57 INFO Temp=26.3*C Humidity=58.0%
2020-09-01 15:47:08 INFO Temp=26.3*C Humidity=56.4%
2020-09-01 15:49:12 INFO Temp=26.3*C Humidity=57.9%
2020-09-01 15:49:17 INFO Temp=26.4*C Humidity=60.9%
2020-09-01 15:49:41 INFO Temp=27.2*C Humidity=58.6%
2020-09-01 15:49:51 INFO Temp=26.3*C Humidity=56.7%
2020-09-01 15:51:49 INFO Temp=26.4*C Humidity=58.2%
2020-09-01 15:51:54 INFO Temp=26.4*C Humidity=59.6%
2020-09-01 15:52:05 INFO Temp=26.3*C Humidity=56.5%
2020-09-01 16:27:57 INFO Temp=26.4*C Humidity=56.9%
2020-09-01 23:56:59 INFO Temp=25.8*C Humidity=46.9%
 

pahval

Well-Known Member
what is your setup? sensors? what devices you control (humidifier, dehumidifier, vents, heater)? do you have timestamps of those also?
 

pahval

Well-Known Member
yea, only thing i could extrapolate is that at night when temps are falling (i pressume lights off) humidity rises and reverse, there is a steady curve from Temp=25.1*C Humidity=58.8% to Temp=22.7*C Humidity=67.6% to Temp=25.3*C Humidity=57.8% again... DHT22 is known for not being so accurate, so false readings (i think) i see here are expected... i remember doing an algorithm that would have 2 sensors and if one went crazy without other that false reading would be rejected (controllers are literal and stupid, if you get that wild reading it could trigger some unwanted process)...
 

solakani

Well-Known Member
what is your setup? sensors? what devices you control (humidifier, dehumidifier, vents, heater)? do you have timestamps of those also?
I have multiple dht22 sensors at various locations. I will look for sht35. The light schedule was 18/6 in 2020-09-01. All my devices are manual control. I am just beginning to setup my grow environment. Just looking for a turnkey solution for fast startup. It would be nice to have a sensor to measure moisture and temperature of soil substrate. My focus is to build my grow environment first. Fixing problems with ventilation and irrigation is a priority. Automation is just something to play with. I am willing to invest in more sensors and provide all logs if it will help me grow better.
 

pahval

Well-Known Member
from what i have researched, temperature in soil is not important, and in my experience and from what i have heard you water every 2-3 days, but you have good... i suggest you check out this site:


and see this:


it explains why certain ratio of temperature and humidity is important. If you run soil only, you can have simple automation of ventilation and dehumidifier, and if you put RTC (clock) you can automate your lights also...
 

pahval

Well-Known Member
yes, but i would also like to point that from what ive seen other people do, they just use air temperature sensor (even DHT22 can be useful if you can correct errors in your code) without taking leaf temperature, and they estimate that is about -2C of air temperature... you dont have to calculate VPD, there are charts online you can use from here:


or here:


both have calculators, second one only has leaf temperature difference (that some people just guess its 2 degrees C lower)...

i use VPD in a way that i have temperature set-point that i want, and then i calculate RH from that being in range of VPD that is needed by this formula:

double e = 2.7182818284590452353602875;
rhSet = ((0.61121 * pow(e, ((18.678 - leafTemp / 234.5)* leafTemp) / (leafTemp + 257.14))) - vpdSet) / (0.61121 * pow(e, ((18.678 - airTemp / 234.5 ) * airTemp) / (airTemp + 257.14)));

*courtesy of Northern_Loki from Overgrow.com: https://overgrow.com/t/source-code-c-temperature-conversion-and-vapor-pressure-deficit-classes-tools/13811/20

vpdSet i choose depending on which stage plant is, and i have 2 temperature set-points, one for day (lights on) and one for night (lights off)... i used values as described in pdf i attached here...
 

Attachments

pahval

Well-Known Member
altough, Arduino platform is very slow with float and double math, these calculations i use on ESP32 because i need them for the custom setpoint inputs, you can just set daily temp and hum and night temp and hum based on that chart, and simple sensor error correction with relay control...
 

pahval

Well-Known Member
DHT22 doesnt have anything to do with rs485, its a standard of communication, like i2c is or spi, its just that i use separate chips for that (its industry standard), i will have multiple devices, separated each by 2-6 meters, so i need fast and reliable way to send data... you need reliable library, the one i use it PJON, but there is also this one:

http://www.gammon.com.au/Arduino/RS485_protocol.zip

i forgot to mention, if youll use MLX sensor, i suggest to read this on how to connect bare sensor, without evaluation board:

 

pahval

Well-Known Member

pahval

Well-Known Member
Does dht22 on sensor node meet pre-requisites for rs485 transmission?
oh, if you mean sending data via RS485 on Arduino, then i can say that most libraries need byte data (you get float data from your sensor, 32.85 and you need values 0-255), so you need to convert float to byte to successfully send...
 

pahval

Well-Known Member
it gets temp and hum value, that it then compares with the desired value, and if its not, it starts heating or cooling and humidifying or dehumidifying (my setup will have air conditioning outside of grow tent with temperatures -6C from inside, so that when i pull in air, i pull in colder air, and dehumidifier and ultrasonic humidifier (mist maker)), but mostly master will be used to communicate with light and nutrient feeder, and on other side to communicate with openHab raspberry pi...
 

solakani

Well-Known Member
Master node must perform some action to trigger a send from sensor node. Do you have these coded?
 
Top