Backyard Aquaponics
A place to discuss aquaponics

That would be annoying. One of the major points of sensors like those is that you can get continuous readings.

It is only a test Stuart

I don't understand.

You can take continuous readings. Just need some isolation to cut out the noise from pumps and lights etc.

I have a couple opto-isolators that will work for the serial data. Just need to find some 1:1 transformers to isolate the power.

Johny5, did atlas say this will happen every time the ph and DO probes are used together. Should I plan on isolating the circuits even before I get my DO kit in the mail?

Atlas said it was the DO probe picking up the stray noise. But is you run either probe without the other then they work as expected. I have isolated the serial and power lines to the DO probe but still get a erroneous readings if both are running.

I do not see it as a magor problem as I currently pol each probe once every 5 minutes and am not looking for a continuous reading to be displayed. If I did I would by a more expensive product. All I will do is change to reading one probe every 10 minutes in sequence whilst the other probe is shut down, that will make it 20 minutes between a probes readings.

I have only isolated the one probe so maybe isolating the other probe as well might work. I should give that a go first.

Here's an update on my Arduino Aquaponic Controller.

I know we all like pics so I'll jump right in...



Here is the assembled box with 40x4 LCD and matrix keypad on the front. To the left is the AC cord and industrial GFCI adapter. There is also a 12V backup battery which is just being used for testing. In the final system, it will be replaced with a deep cycle marine battery and heavier cables. To the right is the sensors and to the bottom is the AC/DC inverter for the battery backup system.


The sensors from left to right. Dissolved Oxygen. pH. Flow sensor. LDR light sensor. A couple of temp and humidity DHT22 sensors. A float sensor for the sump. Ultrasonic range HC-SR04 sensor for tank depth. A bunch of DS18B20 water proof temp sensors. All connected with RJ11 junction boxes.


The right side of the unit has the sensor connections and the AC outlets connected to the relay board.


The left side of the unit has an ethernet jack, banana plugs for the 12V backup battery connection, a fuse, and an EMI filtered AC C14 input jack.


Here is the inside of the unit. On the lid is the 40x4 LCD display and 4x4 matrix keypad. Top left is the ethernet shield. Top middle is the Arduino mega with custom interface shield. Top right is the pH and dissolved oxygen modules. Bottom left is the power distribution, 5V switching regulator and backup power relays. Bottom right is the relay module.


The power board follows the schematic in my first post except I used one of the extra relay contacts to switch the AC power sense going to the arduino. I figured that connecting the 5VDC (iPhone charger) powering the relays directly to the power sense pin on the arduino was not a good idea due to back EMF when the coils switch on and off. When AC power is lost, the relays switch over to DC 12V power. The water pump and air pump stay powered through the inverter and the arduino stays powered as well through the switching buck regulator on the right side of the board. I still need to add a suitably large storage capacitor to the regulator output to prevent the Arduino from resetting when the AC power is lost and the relay breaks from the AC circuit and connects to the DC backup circuit. It's a fraction of a second but enough to brown out the Arduino sometimes.


The relay module switches all the AC Hot side connections. Everything is soldered and heat-shrinked for safety. When I can find a suitable chunk of plexi, I will cover the entire AC section to prevent any accidental contact.


The pH and dissolved oxygen modules are mounted to a perf board. The instructions specifically say not to do this but if you are careful with your soldering and clean all traces of flux, it is fine. I have no issues with noise affecting the results on the test bench but that may change once they are in a tank with pumps. I left enough room on the board for power/signal isolation if required and I have an EMI filter for the water pump as well. I used shielded cables for the signal lines going back to the Arduino.


Here is a shot of the Arduino Mega 2560 R3 with custom shield on top. The shield connects all the signal and power to the sensors terminal blocks. The two ribbon cables connect to the LCD display and Relay module. The two black shielded cables are for the pH and DO sensors. The real time clock module is mounted on the right. The empty terminal blocks are for the demand feeder components and the vibration sensors for the siphon detection. These are still WIP.


Here is a shot of the back of the lid. 40x4 LCD display on top and connection for 4x4 matrix keypad below.


I created a basic menu system to allow changing the system variables. It is nothing fancy but it does the job.


This is the main menu. The setting option takes you to a list of variables that can be changed to alter the operation of the unit. The reset function restores all settings to default. The calibrate function leads you through the process of calibrating the pH and DO probes.


This is the settings categories screen.


Tank settings screen


Grow light control settings screen


Screen for changing setting. Once the setting is changed, it is recorded to EEPROM and retained after power cycling the unit.

Here is a list of all the variables that can be changed in the current revision:
Flow meter sample period in milliseconds
Maximum ultrasonic sensor distance
Maximum tank water level before turning off pump
Minimum light level before the lights will turn on
Maximum light level before the lights will turn off
Time in morning that lights will turn on
Time in evening when lights will turn off
Minimum tank temp must be below for lights to turn on
Maximum tank temp must be above for lights to turn off. Hysteresis
Time in morning when pump cycle switches.
Time in evening when pump cycle switches
Inside temp must be below for pump to turn off at night. Prevents growbeds acting as heatsink.
Inside temp must be above for pump to turn on at night. Hysteresis.
Air temp inside must be below to turn off fans. Hysteresis
Air temp inside must be above to turn on fans
Tank temp inside must be below to turn on heater
Tank temp inside must be above to turn off heater. Hysteresis
Keypad debounce.
TimeZone
Time of no activity on siphon sensor before alert
Time to disable interrupt for siphon sensor to prevent multiple triggers while siphon dumps.
Time that demand feeder motor is on after triggered
Time that the demand feeder can not be triggered after initial trigger


The only thing still to be completed hardware wise is the demand feeder switch/motor and switches for the siphon detection. The vibration sensors (bottom) I thought would work have a really high detection threshold and I'm not sure if the act of flushing the siphon will trigger them. Plan B is to use a long arm limit switch with a paddle on the out tube of the siphon.

The greenhouse construction is under way and the sump hole is started!

Kick starter

Very nice indeed.

I am interested in how you PH and DO go. In your testing did you have both sensors in the same test water together as that is the only time I have issues with them.

I am old school here and like to have the AC and DC sides in separate boxes. Interested to see the durability of those dc relays as I use them to control solid state relays for the AC.

Yes the two probes were in the same bucket for testing. No issues so far.

The relays are rated to 10 amps at 125VAC. The largest load in the system is the 800W heater which draws about 7A @ 110VAC. I have used the same modules in other projects to switch large loads and they are very reliable despite being CCC.

I have protection in the code to prevent both the heater and grow light from coming on at the same time because the GFCI and AC socket EMI filter are rated at 10A max.

Jayendra... Maybe once I have a couple seasons on it and worked out all the bugs

They still are mechanical point relays, I guess I am perhaps over cautions.

I had my probes connected via one of the Atlas multiplex boards so all the soldered joints where theirs. I now have the DO in the tank with PH in a bucket alongside the tank. Both are now reading correctly. Hopefully I will get time in the next few weeks to play with them again.

Two questions I have:

Are there any other suppliers of sensors for the Arduino market?

What is the estimated life span of the Atlas gear?

..
What a great project.. Makes me wonder if I could cope with programming one to do my system..
..
.

Hi Stuart

Not sure about the life span of the probes as time will tell. Mine have been submerged most of the is. I think the DO since January. I calibrate it every few months just to check it.

This Arduino Aquaponic gear is so interesting, bookmarked this thread straight away !



Ill be up for one ! when they come up for grabs.

GTK

It's good to be cautious when playing with AC power. I have personally heard of more issues with solid state relays failing than mechanical ones. Usually it is due to just not driving them properly and blowing out the front end.

Any pH or DO sensor can be used. Typically they output a tiny (couple of mV) signal. You can put that through an OpAmp and feed it to a analog pin on the Arduino. The beauty of the Atlas Scientific stuff is that it does all the processing for you and just spits out a serial string. I have yet to find a good (affordable) ammonia, nitrate, nitrite sensors, though there are rumors that Atlas is working on some for the aquarium crowd.

I've owned many pH probes in the past and the Atlas probes look to be very high quality. They should last as long as anything else on the market as long as you take care of them. There are reconditioning kits for both pH and DO probes but reconditioning should only really need to be done if you let the probe tip dry out.
https://www.atlas-scientific.com/produc ... recon.html?
https://www.atlas-scientific.com/produc ... nance.html?

All the code is linked in the first post. No need to cope with programming! I will be uploading another revision as soon as I add a couple more features.


Side update : I tested the vibration sensors on a mockup siphon and it worked perfectly.These are a very clean and simple solution to detecting siphon activity. I'm very happy that I didn't need plan B. More sump hole digging this weekend, man am I sore.