Growing stuff and doing things

dstroy0

Zeroes and Ones
I've been working fairly consistently on software for the new hardware over the past few days.

1641492206803.png

Those are the main meat and potatoes (for now!) of the controller. Some stuff needs to go, some stuff needs to be added. I'm up to 70% program space on an ESP32-wroom 4mb with spiffs.

Default behavior is to send that over RF24, if for some reason the RF24 radio isn't available it will try and reach the base station over wifi. The controllers are going to be on battery backup in case the power goes out they'll go into deep sleep and wake up to feed and then return to normal operation when input power returns.

The base station could be anything that has access to the rf24 network and wifi, I wrote a logging script in C++ that takes these payloads and turns them into a json.
 

dstroy0

Zeroes and Ones
1641579605216.png

1641579650593.png


mqtt is working, I need to configure auth and then figure out how to scrape it and stream that to a database connector, and then write up a monitor that uses the database to set flags, and something to push configurations to the controllers.

before I do all that I have a shit ton of other work to do on the controller
 

dstroy0

Zeroes and Ones
Sending whole objects on mqtt is really easy, this is how I'm doing it:

C++:
//send an ENV_CONTROLLER_CONFIGURATION_PAYLOAD object over wifi
void send_controller_configuration_mqtt_message(ENV_CONTROLLER_CONFIGURATION_PAYLOAD& temp)
{
  //a place to store things
  uint8_t out[(sizeof(ENV_CONTROLLER_CONFIGURATION_PAYLOAD) + 4)] = {0};
  //copy temp into the place to store things
  copy_class_to_data_array(temp, out);
  //calculate a crc and add it to the payload
  bool ok = crc_out(out, sizeof(out));
  if (ok)
  {
    mqttClient.beginMessage(mqtt_status.topic);   
    //iterate over the place to store things one byte at a time
    for (uint32_t i = 0; i < sizeof(out); i++)
    {
      //add bytes to the message, you can change the print format for different data types ie print(out[i], HEX) will print bytes in hex
      mqttClient.print(out[i]);
    }
    //send the mqtt message
    mqttClient.endMessage();
  }
}

If I add a byte separator in there and go to crccalc I get the known constant for the string that's spit out on the broker, everything seems to be working...
 

dstroy0

Zeroes and Ones
"Sending whole objects on mqtt is really easy"

It's not in English so I'll just stare?
Frickin awesome. I do a lot of things diy but electronics are not one of them.

Since I'm moving permanently from wired to wireless I wanted it to be resilient, all that means is that it has two radios in it, one for wifi and one for rf24, and I can use one or the other or both to send/receive messages. The way it works over wifi is by using a lightweight messaging protocol called mqtt, and I can send a whole chunk of data at a time easily with the wrapper in this library I'm using.
 

dstroy0

Zeroes and Ones
I got the serial monitor working over HTTP, each sensor has it's hostname generated in a configuration macro at compile time, but is able to be changed by the user to whatever. Makes it easier to keep track of a device with a name vs using an ip. This is working alongside OTA, I'm up to 80% program space with nearly everything implemented. Plenty of ram left though. There's definitely room for optimization because I'm not using all the member functions of all the libraries and some of them are not getting optimized out at compile time.

1641763506388.png
 

Caddis

Zinger
Next run I’ll start adding the mkp just a day or two after flipping to 12/12, I think that will help. Everything seems to be working pretty good. Thoughts?
First, thanks for posting something I understand!
I’ve been waiting a week or so to switch to Jack’s Bloom, curious why your considering at flip?
I’m on the curb, so very interested.
If any computer skills are required, disregard! :stoned:
 

dstroy0

Zeroes and Ones
First, thanks for posting something I understand!
I’ve been waiting a week or so to switch to Jack’s Bloom, curious why your considering at flip?
I’m on the curb, so very interested.
If any computer skills are required, disregard! :stoned:

To me it looked like I had some blue tinted leaves that went away when I added the mkp, which could just be specific to the strain, or my feed schedule, I'm not really sure, but it looked like it helped, and I think it wouldn't have happened at all if I had started supplementing earlier.
How's the jacks bloom? I'm really liking the 5-12-26.
 

Caddis

Zinger
To me it looked like I had some blue tinted leaves that went away when I added the mkp, which could just be specific to the strain, or my feed schedule, I'm not really sure, but it looked like it helped, and I think it wouldn't have happened at all if I had started supplementing earlier.
How's the jacks bloom? I'm really liking the 5-12-26.
The 321 is too easy! :stoned:
I was skeptical to be honest about another mix, but have tried the Bloom a couple of times now.
You sorta confirmed my experience so far. It seemed like it cured a deficience that was showing in veg for me. I was actually thinking about feeding Bloom from flip just to see if it will straighten them up a bit.
I’m a hack, so my results?
I was in 1 gal. fabric, coco/perlite dtw, running 321 almost full blast.

I’ve tried the Jack’s Finish, hard to tell yet. Hadn’t hurt anything so far.
I like the one part mix, I did add 100 ppm of Epsom with the Bloom.
 

dstroy0

Zeroes and Ones
The diablo looks shorter, leafier and tighter nodes than the cobs. Is that true?
Yes, there is quite a difference, and I’ve been giving them similar DLI according to the umol readings from the ePAR meter. Environmental settings are the same across both spaces, and the measured environmentals match, feed is the same, feed settings are the same. So the spectrum has a real influence on the type of growth that is expressed because of the efficiency of certain chemical reactions and the mechanism is the amount of certain wavelengths of light according Dr Bruce Bugbee, which has some real interesting implications for space management through spectrum manipulation.
 

dstroy0

Zeroes and Ones
The 321 is too easy! :stoned:
I was skeptical to be honest about another mix, but have tried the Bloom a couple of times now.
You sorta confirmed my experience so far. It seemed like it cured a deficience that was showing in veg for me. I was actually thinking about feeding Bloom from flip just to see if it will straighten them up a bit.
I’m a hack, so my results?
I was in 1 gal. fabric, coco/perlite dtw, running 321 almost full blast.

I’ve tried the Jack’s Finish, hard to tell yet. Hadn’t hurt anything so far.
I like the one part mix, I did add 100 ppm of Epsom with the Bloom.

Do you use an aerocloner to clone? I'm looking for a good 1 part that isn't clonex.
 

Caddis

Zinger
Do you use an aerocloner to clone? I'm looking for a good 1 part that isn't clonex.
I have a Turbo Cloner, use Aloe Vera to start. I’ve never used Clonex, just light Jack’s, but I don’t clone much.
I got lazy recently and just stuck some Lifesaver clones in rapid rooters thrown in the middle of some solos. Expectations are low, they have no future anyway. )

I know you are considerably more precise, so my seat of the pants approach may not offer much insight! :stoned:
 

DopeDaniel

Taste The Spectrum
IPM Forum Moderator
the spectrum has a real influence on the type of growth that is expressed because of the efficiency of certain chemical reactions and the mechanism is the amount of certain wavelengths of light
Personally by the looks of things I'd prefer the cob side.

Blue photons are higher energy than red. Photons are released in the photosynthesis reaction. Those photons posess less energy than the original photon. It takes more than one photon to complete the photosynthesis reaction but I can't find anything that says it can't be the same original photon. I feel like this is what creates the different morphology. In a sense the cell doesn't have to wait for another photon to come along, its already there. So it does what its programmed to do reproduce more cells like it so the incoming energy can be processed. Funny the system breaks under monochromatic light though.
 
Top Bottom