Reef Discussion
This is a view of the whole system. Silver perch in top tanks, koi and minnows in bottom right, yabbies in upper left down the slope and the one underneath is the sump. The sump will soon have glass shrimp which I will be growing as feed. The other shallow tanks and the coiled ag pipe are the grow beds.
Not sure why the image is upside down, it looks fine on my tablet. Also not sure why image did not upload in previous posts.
The pdf attached shows the circuit plan for my interface board. I have had to compromise on numbers of pins for various purposes. I chose 6 switch inputs and 7 open collector outputs. The outputs are not being used yet, but will eventually will control things like pumps, automatic feeders, valves etc. They interface through a ULN2003 chip which has 7 channels. The 8 analog inputs go through an MCP3008 chip and use SPI.
Will post python code soon.
The pdf attached shows the circuit plan for my interface board. I have had to compromise on numbers of pins for various purposes. I chose 6 switch inputs and 7 open collector outputs. The outputs are not being used yet, but will eventually will control things like pumps, automatic feeders, valves etc. They interface through a ULN2003 chip which has 7 channels. The 8 analog inputs go through an MCP3008 chip and use SPI.
Will post python code soon.
The code I am using is in 2 parts. The first reads any change in the siphon state, ie from filling to draining and vice versa. To do this I am using "interrupts". This means the program can run in the background, but does not sit in a loop stopping other things from happening. The program starts up when the Pi is booted and stays running (hopefully).
I am a messy programmer, so please forgive the style. I tend to be an "if it works, don't change it" sort of person, but if anyone has any suggestions about improving it feel free to comment.
Here is the code for this:
#!/usr/bin/env python2.7
# script adapted from Alex Eames http://RasPi.tv/
# http://raspi.tv/2013/how-to-use-interrupts-with-python-on-the-raspberry-pi-and-rpi-gpio
import RPIO
import RPi.GPIO as GPIO, datetime, json
import httplib
#print GPIO.RPI_REVISION
GPIO.setmode(GPIO.BCM)
pins = {14, 15, 18, 23, 24, 25} # Pins I hav used on interface board.
SENSE_API_KEY = "your_key" # Key to connect to open.sen.se website.
Air_Temperature = “an_id” # Each data stream has an id.
Fish_Tank_Temperature = “an_id”
Grow_Bed_Temperature = “an_id”
Light_Level = “an_id”
GB1_Siphon_State = “an_id”
GB2_Siphon_State = “an_id”
GB3_Siphon_State = “an_id”
GB4_Siphon_State = “an_id”
GB5_Siphon_State = “an_id”
GB6_Siphon_State = “an_id”
de_bounce = 5000 # 5 seconds
def senseit(gpio_id, val):
#print "\t=> Sending to OpenSense: %s" % sensedata
#value = GPIO.input(gpio_id)
# prepare data
if gpio_id == 14:
datalist = [{"feed_id" : GB1_Siphon_State, "value" : val}]
if gpio_id == 15:
datalist = [{"feed_id" : GB2_Siphon_State, "value" : val}]
if gpio_id == 18:
datalist = [{"feed_id" : GB3_Siphon_State, "value" : val}]
if gpio_id == 23:
datalist = [{"feed_id" : GB4_Siphon_State, "value" : val}]
if gpio_id == 24:
datalist = [{"feed_id" : GB5_Siphon_State, "value" : val}]
if gpio_id == 25:
datalist = [{"feed_id" : GB6_Siphon_State, "value" : val}]
headers = {"sense_key": SENSE_API_KEY,"content-type": "application/json"}
conn = httplib.HTTPConnection("api.sen.se")
# format a POST request with JSON content
conn.request("POST", "/events/", json.dumps(datalist), headers)
response = conn.getresponse()
conn.close()
def gpio_callback(gpio_id, val):
#pin = gpio_id
#print("gpio %s: %s" % (gpio_id, val))
#value = GPIO.input(pin)
#print("gpio %s: %s" % (gpio_id, value))
senseit(gpio_id, val)
# GPIO interrupt callback with debounce timeout of de_bounce)
GPIO.cleanup()
for pin in pins:
#print pin
#GPIO.setup(pin, GPIO.IN)
#GPIO.remove_event_detect(pin)
RPIO.add_interrupt_callback(pin, gpio_callback, edge='both', debounce_timeout_ms = de_bounce)
#print pin
# Starts waiting for interrupts (exit with Ctrl+C)
RPIO.wait_for_interrupts()
I am a messy programmer, so please forgive the style. I tend to be an "if it works, don't change it" sort of person, but if anyone has any suggestions about improving it feel free to comment.
Here is the code for this:
#!/usr/bin/env python2.7
# script adapted from Alex Eames http://RasPi.tv/
# http://raspi.tv/2013/how-to-use-interrupts-with-python-on-the-raspberry-pi-and-rpi-gpio
import RPIO
import RPi.GPIO as GPIO, datetime, json
import httplib
#print GPIO.RPI_REVISION
GPIO.setmode(GPIO.BCM)
pins = {14, 15, 18, 23, 24, 25} # Pins I hav used on interface board.
SENSE_API_KEY = "your_key" # Key to connect to open.sen.se website.
Air_Temperature = “an_id” # Each data stream has an id.
Fish_Tank_Temperature = “an_id”
Grow_Bed_Temperature = “an_id”
Light_Level = “an_id”
GB1_Siphon_State = “an_id”
GB2_Siphon_State = “an_id”
GB3_Siphon_State = “an_id”
GB4_Siphon_State = “an_id”
GB5_Siphon_State = “an_id”
GB6_Siphon_State = “an_id”
de_bounce = 5000 # 5 seconds
def senseit(gpio_id, val):
#print "\t=> Sending to OpenSense: %s" % sensedata
#value = GPIO.input(gpio_id)
# prepare data
if gpio_id == 14:
datalist = [{"feed_id" : GB1_Siphon_State, "value" : val}]
if gpio_id == 15:
datalist = [{"feed_id" : GB2_Siphon_State, "value" : val}]
if gpio_id == 18:
datalist = [{"feed_id" : GB3_Siphon_State, "value" : val}]
if gpio_id == 23:
datalist = [{"feed_id" : GB4_Siphon_State, "value" : val}]
if gpio_id == 24:
datalist = [{"feed_id" : GB5_Siphon_State, "value" : val}]
if gpio_id == 25:
datalist = [{"feed_id" : GB6_Siphon_State, "value" : val}]
headers = {"sense_key": SENSE_API_KEY,"content-type": "application/json"}
conn = httplib.HTTPConnection("api.sen.se")
# format a POST request with JSON content
conn.request("POST", "/events/", json.dumps(datalist), headers)
response = conn.getresponse()
conn.close()
def gpio_callback(gpio_id, val):
#pin = gpio_id
#print("gpio %s: %s" % (gpio_id, val))
#value = GPIO.input(pin)
#print("gpio %s: %s" % (gpio_id, value))
senseit(gpio_id, val)
# GPIO interrupt callback with debounce timeout of de_bounce)
GPIO.cleanup()
for pin in pins:
#print pin
#GPIO.setup(pin, GPIO.IN)
#GPIO.remove_event_detect(pin)
RPIO.add_interrupt_callback(pin, gpio_callback, edge='both', debounce_timeout_ms = de_bounce)
#print pin
# Starts waiting for interrupts (exit with Ctrl+C)
RPIO.wait_for_interrupts()
The next program is triggered periodically by "cron". I have it set to read every 10 minutes.
This one is really messy, I will clean it up soon, but it does work. The calibration for the thermisters is done by loading data from a calibration table.
#!/usr/bin/python
import httplib, urllib
# download from <a href="http://code.google.com/p/psutil/" title="http://code.google.com/p/psutil/">http://code.google.com/p/psutil/</a>
import RPi.GPIO as GPIO, spidev, time, math, os, json, datetime
# use P1 header pin numbering convention
#GPIO.setmode(GPIO.BCM)
#Calibration info
DEBUG = 0
Voltage = 0.0
Slope = -534.692
Divider0 = 10000.0
data = json.load(open('./Documents/lookup.json'))
reading = 511
# Open.sen.se info
SENSE_API_KEY = "your_key"
FEED_ID1 = “an_id”
FEED_ID2 = “an_id”
FEED_ID3 = “an_id”
FEED_ID4 = “an_id”
FEED_ID5 = “an_id”
Air_Temperature = “an_id”
Fish_Tank_Temperature = “an_id”
Grow_Bed_Temperature = “an_id”
Light_Level = “an_id”
GB1_Siphon_State = “an_id”
GB2_Siphon_State = “an_id”
GB3_Siphon_State = “an_id”
Sump_Tank_Level = “an_id”
Yabbie_Tank_Level = “an_id”
Water_Level = “an_id”
delay = 0
samples = 5
def get_adc(channel): # Read by SPI
spi = spidev.SpiDev()
spi.open(0,0)
spi.max_speed_hz = 4000
# read SPI data from MCP3008 chip
# Only 8 channels 0 to 7 else return -1
if ((channel > 7) or (channel < 0)):
return -1
#time.sleep(delay)
sum = 0
maximum = 0
minimum = 1023
for value in range(samples):
r = spi.xfer2([1,(8+channel)<<4,0])
ret = ((r[1]&3) << 8) + r[2]
#print minimum, maximum
if ret < minimum:
minimum = ret
if ret > maximum:
maximum = ret
#print ret,
sum = sum + ret
#print "Average = ", ret, " Max = ", maximum, " Min = ", minimum
#Average = (sum/samples)
ret = (sum-maximum-minimum)/(samples-2)
#print "Return = ", ret, " Average = ", Average, " Max = ", maximum, " Min = ", minimum
#time.sleep(1)
# Finished reading so turn off channel
spi.close()
#GPIO.cleanup()
return ret
def calibrate(reading, parameter): # Different calibration for different sensors
#reading = 1023 - reading
#print "Resistance = ", Resistance
if parameter == "Light":
Resistance = ((1023.0*Divider0)/(1023.0-reading))-Divider0
Light = (Resistance)
return Light/100
if parameter == "Temp":
#print reading
# Lookup Temperature from imported data table
reading = int(reading)
#print reading
TempC = data.get(str(reading))
#print TempC
return TempC
if parameter == "Level":
return reading/10.0
def senseit(sensedata):
#print "\t=> Sending to OpenSense: %s" % sensedata
# prepare data
datalist = [{"feed_id" : Air_Temperature, "value" :sensedata['temp1']},
{"feed_id" : FEED_ID2, "value" :sensedata['temp2']},
{"feed_id" : FEED_ID3, "value" :sensedata['temp3']},
{"feed_id" : FEED_ID4, "value" :sensedata['light']},
]
headers = {"sense_key": SENSE_API_KEY,"content-type": "application/json"}
conn = httplib.HTTPConnection("api.sen.se")
# format a POST request with JSON content
conn.request("POST", "/events/", json.dumps(datalist), headers)
response = conn.getresponse()
# you may get interesting information here in case it fails
#print response.status, response.reason
#print response.read()
conn.close()
def doit():
field1 = 0.010
#field2 = 100.0
#field3 = 100.0
#field4 = 100.0
for channel in range(4):
Voltage = get_adc(channel)
#print Voltage
#print channel, Voltage
if Voltage <> 0:
if channel == 0:
para = "Light"
field1 = round(calibrate(Voltage, para),1)
elif channel == 4:
para = "Level"
field5 = calibrate(Voltage, para)
else:
para = "Temp"
if channel == 1:
field2 = calibrate(Voltage, para)
elif channel == 2:
field3 = calibrate(Voltage, para)
elif channel == 3:
field4 = calibrate(Voltage, para)
#time.sleep(1)
#print channel, " ", Voltage, para, round(calibrate(Voltage, para),1)
sensedata = { 'temp1' : field2, 'temp2' : field3, 'temp3' : field4, 'light' : field1}
#print sensedata
senseit(sensedata)
# Run Main
if __name__ == "__main__":
doit()
This one is really messy, I will clean it up soon, but it does work. The calibration for the thermisters is done by loading data from a calibration table.
#!/usr/bin/python
import httplib, urllib
# download from <a href="http://code.google.com/p/psutil/" title="http://code.google.com/p/psutil/">http://code.google.com/p/psutil/</a>
import RPi.GPIO as GPIO, spidev, time, math, os, json, datetime
# use P1 header pin numbering convention
#GPIO.setmode(GPIO.BCM)
#Calibration info
DEBUG = 0
Voltage = 0.0
Slope = -534.692
Divider0 = 10000.0
data = json.load(open('./Documents/lookup.json'))
reading = 511
# Open.sen.se info
SENSE_API_KEY = "your_key"
FEED_ID1 = “an_id”
FEED_ID2 = “an_id”
FEED_ID3 = “an_id”
FEED_ID4 = “an_id”
FEED_ID5 = “an_id”
Air_Temperature = “an_id”
Fish_Tank_Temperature = “an_id”
Grow_Bed_Temperature = “an_id”
Light_Level = “an_id”
GB1_Siphon_State = “an_id”
GB2_Siphon_State = “an_id”
GB3_Siphon_State = “an_id”
Sump_Tank_Level = “an_id”
Yabbie_Tank_Level = “an_id”
Water_Level = “an_id”
delay = 0
samples = 5
def get_adc(channel): # Read by SPI
spi = spidev.SpiDev()
spi.open(0,0)
spi.max_speed_hz = 4000
# read SPI data from MCP3008 chip
# Only 8 channels 0 to 7 else return -1
if ((channel > 7) or (channel < 0)):
return -1
#time.sleep(delay)
sum = 0
maximum = 0
minimum = 1023
for value in range(samples):
r = spi.xfer2([1,(8+channel)<<4,0])
ret = ((r[1]&3) << 8) + r[2]
#print minimum, maximum
if ret < minimum:
minimum = ret
if ret > maximum:
maximum = ret
#print ret,
sum = sum + ret
#print "Average = ", ret, " Max = ", maximum, " Min = ", minimum
#Average = (sum/samples)
ret = (sum-maximum-minimum)/(samples-2)
#print "Return = ", ret, " Average = ", Average, " Max = ", maximum, " Min = ", minimum
#time.sleep(1)
# Finished reading so turn off channel
spi.close()
#GPIO.cleanup()
return ret
def calibrate(reading, parameter): # Different calibration for different sensors
#reading = 1023 - reading
#print "Resistance = ", Resistance
if parameter == "Light":
Resistance = ((1023.0*Divider0)/(1023.0-reading))-Divider0
Light = (Resistance)
return Light/100
if parameter == "Temp":
#print reading
# Lookup Temperature from imported data table
reading = int(reading)
#print reading
TempC = data.get(str(reading))
#print TempC
return TempC
if parameter == "Level":
return reading/10.0
def senseit(sensedata):
#print "\t=> Sending to OpenSense: %s" % sensedata
# prepare data
datalist = [{"feed_id" : Air_Temperature, "value" :sensedata['temp1']},
{"feed_id" : FEED_ID2, "value" :sensedata['temp2']},
{"feed_id" : FEED_ID3, "value" :sensedata['temp3']},
{"feed_id" : FEED_ID4, "value" :sensedata['light']},
]
headers = {"sense_key": SENSE_API_KEY,"content-type": "application/json"}
conn = httplib.HTTPConnection("api.sen.se")
# format a POST request with JSON content
conn.request("POST", "/events/", json.dumps(datalist), headers)
response = conn.getresponse()
# you may get interesting information here in case it fails
#print response.status, response.reason
#print response.read()
conn.close()
def doit():
field1 = 0.010
#field2 = 100.0
#field3 = 100.0
#field4 = 100.0
for channel in range(4):
Voltage = get_adc(channel)
#print Voltage
#print channel, Voltage
if Voltage <> 0:
if channel == 0:
para = "Light"
field1 = round(calibrate(Voltage, para),1)
elif channel == 4:
para = "Level"
field5 = calibrate(Voltage, para)
else:
para = "Temp"
if channel == 1:
field2 = calibrate(Voltage, para)
elif channel == 2:
field3 = calibrate(Voltage, para)
elif channel == 3:
field4 = calibrate(Voltage, para)
#time.sleep(1)
#print channel, " ", Voltage, para, round(calibrate(Voltage, para),1)
sensedata = { 'temp1' : field2, 'temp2' : field3, 'temp3' : field4, 'light' : field1}
#print sensedata
senseit(sensedata)
# Run Main
if __name__ == "__main__":
doit()
The result of all of this can be seen at:
http://open.sen.se/sensemeters/tab/4745/
Hope you like it.
http://open.sen.se/sensemeters/tab/4745/
Hope you like it.
I was wondering if there was any milage intergrating some the airpi aspects into the aquapi. C02 levels in the room would be interesting to know.
http://airpi.es/participate.php
Sent from my Nexus 10 using Tapatalk 4
http://airpi.es/participate.php
Sent from my Nexus 10 using Tapatalk 4
Guys, this is an awesome project, Rob & Clownfishy you have done so well to get this far.
I am new to Raspberry pi and am currently learning ruby so that I can get stuck into this project to monitor my nano tank. Do you think this will be too much for me?
I wanted to see if you guys thought it would be a good idea to also tie in the project over at theoreticalideations.com so that you could also control lighting and set it to come on with online sunrise/sunset times for your location.
I am running a Fluval Sea M40, so am going to look into seeing if it possible to get the current led lamp to work in this way.
I am new to Raspberry pi and am currently learning ruby so that I can get stuck into this project to monitor my nano tank. Do you think this will be too much for me?
I wanted to see if you guys thought it would be a good idea to also tie in the project over at theoreticalideations.com so that you could also control lighting and set it to come on with online sunrise/sunset times for your location.
I am running a Fluval Sea M40, so am going to look into seeing if it possible to get the current led lamp to work in this way.
@Rob :)
We dd have a reefuge FTP site with a backup which when copied across to your www directory gave you the complete interface and separate read script. Maybe the reefuge FTP site vaporized when the original micro communities were lost. Ill take a look on the weekend
I have my lights controlled via my profilux 3. It provides spectrum plots, overlays etc etc. after setting these up I think the interface although nice is way overkill. A RPi using maybe the adafruit 16 ch pwm to i2c interface and a simple script lie would work well. The RPi has proved very reliable.
Another Australian forum MASA has someone using it for LEDs.and developing an interface which looks great.. Again I think a script controlling pwm ver time would get you running real fast. You don't need a fancy interface For LED , the tank lighting provides instant feedback if you have it correct.
Another Australian forum MASA has someone using it for LEDs.and developing an interface which looks great.. Again I think a script controlling pwm ver time would get you running real fast. You don't need a fancy interface For LED , the tank lighting provides instant feedback if you have it correct.
Thanks for the advice, I will take a look at it as currently I have a Fluval all in one LED light with day and moonlight settings. It would be nice to be able to switch between the two and control this device instead of having to make my own light.
Let me know if you find the FTP site with all the stuff needed to build my own monitor, am really looking forward to giving it a try. Also, I saw that Clownfish did create a "How to" document, but the google docs link is not working. If you could forward everything that would be awesome, just let me know if you need my email address.
thanks Rob
Let me know if you find the FTP site with all the stuff needed to build my own monitor, am really looking forward to giving it a try. Also, I saw that Clownfish did create a "How to" document, but the google docs link is not working. If you could forward everything that would be awesome, just let me know if you need my email address.
thanks Rob
@Moreno5, this is Robs project and it is him who should take the praise. Thanks to him, I have it reading the temp of the tank and room temp then posting it to xively. I did start pulling in all the info and putting into one single doc. More than happy to do this as Rob has most the info on here.
@Rob, what do you think, a single doc linked from the first page of the thread to all the info?
@Rob, what do you think, a single doc linked from the first page of the thread to all the info?
@Clownfishy that's what I thought, you did start talking about being put all in one place and as I was reading through the whole thread I could not find it. It would be great if you could put the full instructions all under one files, maybe we could PDF it along with offering the Code that Rob kindly spent time on and the front end that I think another user helped with.
It really excited me to see people from all over the world interacting to build something non profit for a group of people with the same passion.
Members online
No members online now.