#hexplanter

A lot of things have happened in my life, but I am finally getting back to a baseline normal again. Although as aftershocks of some of my life changes continue to hit, my uploading / work in general may continue to hit strange and unexpected delays.

But, for the moment, I have come back to the Hex Planter again, finally. I’m looking to have a working and tested version by December.

A couple months ago I completely redesigned it. The last version posted, a two section unit, had some big problems. I moved up to three modular units. But I quickly scrapped that and the next designs because they weren’t actually modular; whenever I thought about repositioning parts I ran into huge problems. So I completely scrapped everything and started over with the base concepts. Hexagon shaped, modular, automated, small.

This is what I came up with: A Power & Control Module, a Water Reservoir Module, and the Planter Module. There is an access port on every panel, with different kinds of spacers depending on the use; but more on that in another post.

In total it will support a single control unit and up to two water pumps, reservoirs, and planters. The control unit also houses 6 LEDs on an LDR circuit to act as ambient light when it’s dark.

The power usage is relatively low, one PCM, WRM, and PM attached with everything in use has a power draw of approximately 700mA or 8.4Watts. With 1 PCM, two WRMs and two PMs, puts the draw at approximately 1.3A or 15.8 Watts.

If I can come up with the space, I could have one controller run up to 3 PMs and WRMs with pumps. Problem is the pumps take a lot of space. Future may hold a small add-on tertiary pump house for this reason.

I am also planning to add a power pass through so that upgrading to a 3A power supply could add in a second PCM, which would easily allow 4 units. But that’s for the future after I work out any bugs.

Third print finished, also at just about 34hrs. I ran an analysis on my gcode and discovered something kind of funny. I had designed all 3 units to be approximately the same. They are within 0.6 grams of each other. Lowest being 206.18g and the highest being 206.76g.

Anyway, the water reservoir printed about the same quality, minor defect on one side, not very noticeable, and cosmetic only. There was a lot of fuzzing but considerably less burring that time. Coupling panel was just as much of a pain though and I was unable to get one between the planter and water reservoir. Redesign is in progress.

Here is the completed code for the hex planter. It’s nice and simple, easy to separate, test, and analyze should a problem arise. Only major change, other than adding the second set of sensors and controllers is I took out the pules segment for the pump indicator LEDs; it was an unneeded bit of flourish and was a bit temperamental.

#include <Wire.h> #include "RTClib.h"

const int bLED = 11; const int bLED2 = 10; const int wetLED = 9; const int dryLED = 8; const int gLEDs = 7; const int wPump = 6; const int wPump2 = 5; const int dryLED2 = 4; const int wetLED2 = 3;

const int mSense = A0; const int wSense = A1; const int mSense = A2; const int wSense = A3;

int wValue; int mValue; int wValue2; int mValue2;

int wState; int wState2; // 2 = no water 1 = water

RTC_DS1307 RTC;

void setup() {  Serial.begin(9600);  pinMode(bLED, OUTPUT);  pinMode(bLED2, OUTPUT);  pinMode(wetLED, OUTPUT);  pinMode(dryLED, OUTPUT);  pinMode(wetLED2, OUTPUT);  pinMode(dryLED2, OUTPUT);  pinMode(wPump, OUTPUT);  pinMode(wPump2, OUTPUT);  pinMode(wSense, INPUT);  pinMode(mSense, INPUT);  pinMode(wSense2, INPUT);  pinMode(mSense2, INPUT);  pinMode(gLEDs, OUTPUT);  Wire.begin();  RTC.begin();  if (! RTC.isrunning()) {    Serial.println("RTC is NOT running!");    RTC.adjust(DateTime(__DATE__, __TIME__));  } }

void loop() {  Serial.println();  DateTime now = RTC.now();  showTime();  if (now.hour() >= 7 && now.hour() <= 21) {    gLEDs_On();  } else {    gLEDs_Off();  }  delay(200);  checkWater();  checkSoil();  checkWater2();  checkSoil2();  } }

void gLEDs_On() {  digitalWrite(gLEDs, HIGH);  delay(200); }

void gLEDs_Off() {  digitalWrite(gLEDs, LOW);  delay(200); }

void checkWater() {  wValue = analogRead(wSense);    Serial.print("W");    Serial.print(wValue);    Serial.println();  delay(200);  if (wValue > 1020) {    digitalWrite(wetLED, HIGH);    digitalWrite(dryLED, LOW);    wState = 1; //there is water  } else {    digitalWrite(wetLED, LOW);    digitalWrite(dryLED, HIGH);    wState = 2; //there is no water  }  delay(200); }

void checkWater2() {  wValue2 = analogRead(wSense2);    Serial.print("W2");    Serial.print(wValue2);    Serial.println();  delay(200);  if (wValue2 > 1020) {    digitalWrite(wetLED2, HIGH);    digitalWrite(dryLED2, LOW);    wState2 = 1; //there is water  } else {    digitalWrite(wetLED2, LOW);    digitalWrite(dryLED2, HIGH);    wState2 = 2; //there is no water  }  delay(200); }

void checkSoil() {  mValue = analogRead(mSense);  Serial.print("M");  Serial.print(mValue);  Serial.println();  delay(200);  if (mValue > 1000 && wState == 1) {    digitalWrite(wPump, HIGH);    Serial.print("Pump On");    Serial.println();    digitalWrite(bLED, HIGH);  } else {    digitalWrite(wPump, LOW);    digitalWrite(bLED, LOW);  }  delay(200); }

void checkSoil2() {  mValue2 = analogRead(mSense2);  Serial.print("M2");  Serial.print(mValue2);  Serial.println();  delay(200);  if (mValue2 > 1000 && wState2 == 1) {    digitalWrite(wPump2, HIGH);    Serial.print("Pump2 On");    Serial.println();    digitalWrite(bLED2, HIGH);  } else {    digitalWrite(wPump2, LOW);    digitalWrite(bLED2, LOW);  }  delay(200); }

Here’s a closeup of the Power & Control Module. It supports up to two pumps, a special panel to house the DC-jack, and a carriage for the control unit, to make it easy to install, change, or update. There are also two independent rows of 3 connectors for the ambient LEDs so you can hook up just the ones you want to use. (To a certain extent, I’m planning on building a resister jumper to take the place of an LED if you don’t want to use all three in one of the rails.)

I kept running into issues with power relay and management, space, and general layout. I went through several iterations of the control system, usually focused around the Arduino Nano, but I tried using smaller units for a more modular multi controller system, or splitting up functions, but what I finally realized is that the Arduino Uno R3 fit pretty perfectly inside, was more powerful, and easier to build the layout around for the control shield. I split the photo sensor and ambient lights off into a single passive LDR circuit and solved all of my control issues. I didn’t expect that using a larger unit would actually save me space, but it did.

Only problem I see at the moment is that the pumps will be above the DC-jack, which can technically be moved to another panel, but if the pumps leak (which they probably will a tiny bit) then a single drop of water could accidentally short out the power supply. I’ll need to either waterproof the DC-jack or build a drip cover over it.

Minor design changes to accommodate the electrical changes, the center inside LED housing had to be made removable, otherwise it would be a royal pain to get the controller carriage in and out. This will make it much easier as the LED mount will just slide out.

I also modified the top plate to be a bit better, it has a lip to make it fit a bit nicer, LED mounts for the pump indicators, which were previously just going to be hot glued onto the side of the pumps the represent. and a bit of a mount lip for the photo resistor so that it is also easier to mount.

Here is the redesign of the PCB, it supports two sets of grow lights, two moisture sensors, two water sensors, two sets of reservoir indicator LEDs, two pump indicator LEDs, and two water pumps. Also better signal isolation for the sensors, so they work and don’t interfere with each other.

I knew I was going to have to make it larger to accommodate the extra hardware and features, but this is larger than I expected. Also the Relays are pretty big requiring me to make a minor design change to the PCM again.

My PCBs arrived today, and as always they look great. Since this is a test set the Hex Planter shield only supports a single planter and pump. The plus side is there is tonnes of space and it was super easy to solder the components on.

I also created one of the 14LED grow arrays, which fits nicely. And wired in everything I have. I still need to hook up the pump, set the clock, upload the program, and I can begin testing.