DIY aquarium controller

[felch]

Hi, marines !
Encouraged bymany splendid projects displayed in this subforum i thought that my modest controller prototype could be also useful.
It is designed mainly with tropical tank in mind (since i have only such) but features could be added to suit it to marine aquariums, i guess.
All materials are here: http://parsek.yf.ttu.ee/~felc/index_files/Page625.htm
Sorry, it is all in estonian but if needed, i will translate. Schematics and code are on mentioned page in file: http://parsek.yf.ttu.ee/~felc/Aqua-4P.zip
Manual (in estonian yet but it shows menus so one could get an idea what's up): http://parsek.yf.ttu.ee/~felc/kontroller.pdf
There are not much menus with foreign language but if you think it's useful, i'll compile an international version of code as well.
Some features:
controller drives 4 outputs (220V) ON/OFF style. Used for main lights, another light source (used when maintaining the tank for example), heater and filter.Actually, itäs only a question of devices names displayed on sceen (borrowed from old _Nokia 3310), one could use mentioned outputs for things needed.
Each of them could be controlled by internal timer (4 different ON/OFF times)
There are also 4 outputs rated for +12V (internal supply of 10W). Two of those are designed for connecting LED n(night-) lights, one for vents under tank cover and one for automatic feeding system. First 3 feature user defined power and switchingtime (example: one can set BLUE LED to turn on at predefined time, power in ON state 60%, such power achieved in 30s). Like 220V loads, all 12V loads have also similar timer control.
Thereäs a built-in clock naturally with summer-winter time correction (if enabled).
There's an analogue water level sensor and sensor indicating state of tank cover (raised or not).
Alarms (buzzer)will be generated (no SMS's in this version, memory full already, will be in next release) in case of over/under temp, low water level, temp. sensor error or clock error/clock not set.
Heater will be disabled in case of temp. sensor error and also in low water.
All devices could be set to follow temperature (OFF/MAX PWR/dont care).
May-be i forgot some features...
Input for designing a more versatile controller that could be useful in salt-water aquariums is most welcome! Bugs report, if anybody cares to repeat this construction is even more welcome
Have fun!
/Felch[/b]

 

[Anonymous]

I will definitely look over it when I have a decent amount of time. Sound very interesting!

 

[felch]

Hi!
I am developing a new version of the controller with improved features.
- user calibrated capacitive water level sensor
- output for 2 dimmable ballasts (0..10V)
- output for 220V vent
- may-be a Ph meter
- SMS reporting of alarms (in PDU format, so really old phones could be used)
- english language menus
- etc.
So, may i suggest that those,interested in constructing the controller will wait for some time until i finish new code. Old version DOES work but i don't like it any more and support is limited to fixing only severe errors (luckily none found yet).
Here's a glimpse to the improved circuit: http://parsek.yf.ttu.ee/~felc/aqua_4_1circ.zip . Please note that circuit is not complete yet!
Any ideas or wishes to the code?

(Felch

 

[_Adrian_]

well theres an array of things that can be added

PH, Salinity, Calcium, ORP, Nitrate, Thermometer with Temperature Control ( heat/cool )...
possibilities are endless...now if a bigger screen can be grafted to be able to display the data and parameters then were talking a lot of interested ppl

I would personaly start from scratch and use a strong processor a good display that is easy to see and read ( at least a minimum of 3"x5" )
and a fast slave processor that can comunicate and power the headunit through a single RJ45 cable which would make it very portable easily mountable where it can be seen and kept on eye on and the slave unit can be housed in a larger case they will be handle all the input and outputs.
Also this allows for more "intricate" control/monitoring strategies through the use of USB / LAN and software updates.

Also the processor must be stable and fast and have lots of overhead to allow for future expansion and upgrades, have a built in RTC that will allow multiple timers and events that can be run simultanously without interfeering with other functions, and this also means that it should have a fast bus where data can be transfered from master to slave so CAN Bus or equivalent would be advisable.

this is my point of view as an automotive tech, where i have been tuning and built a lot of cars though out the 10 years that i had my licence for.

 

[felch]

Hi!
Can You point me links for electronic devices (or principles how such should work) for measuring all mentioned water parameters? I have read articles about colorimetric devices but...would they be reliable enough? I mean - dust and possible carbonates (hard water) and algae could spoil readings of optical devices.
About screen. Is there a really handy, cost-effective (wishfully easy to control - electronically) one out there? I was thinking of a controller with smaller screen (from cell phone). One could use it for changing most important parameters of system. All others could be controlled via web page generated by the controller (not in this version but surely in the next one, i'm experimenting with the NIC chip already).
Additionally i thought of adding water speed sensors. I have found my Eheim stuck for 3 times already. It could be handy to have an alarm for such things.
Are there some de facto industry standard feeding mechanisms? Mine is self built (solenoid opens "door" of food bunker for time defined by controller and as many times as the user wishes).
What else?

/Felch
PS. Firmware updates will be in new code for sure. Firstly form MMC-card but as soon as i learn TCP...

 

[_Adrian_]

from what i have read is that most probes out there need to be "calibrated" with a calibration solution before they can be used and they put out an electrical signal on the mV scale>
all we really have to do is define the input, calibrate and away you go.

so we eould mark input 1.... 5 (just an example)
where 1 would be PH - Calibrate, 2 would be salinity - calibrate, 3 would be nitrate - calibrate, and so on.

as far as the temperature would be quite easy...
Actual Temp = 79
Set Temp = 75
if temp +5 then COOL
if temp -5 then HEAT

personaly i think that the most important piece on there will be the the lighting control.
this section must be fully programable, with On times and off times.
at least 4 timered channels with maybe 2 dimmable outputs or 10V control
this would allow actinic in, VHO 2 and 3 dimming in, 4 MH on/off

now also we could work on a soft start for pumps and even better a built in wave maker !
and as far as addon's it would be nice to have some spare I/O for what ever the end user want to configure it to.

when it comes down to it all ..
im good with the board side of things ..
but i really suck at programing things.
As far as the display...
theres a wide aray of LCD's out there that size and if you are really looking for a challenge... there a lot of touch screens as well
here's a link
http://store.earthlcd.com/

PS... more hits on this page
http://www.alibaba.com/productsearch/Lcd_Panel.html
http://www.altadox.com/lcd/products.htm

 

[felch]

Idea of changing temp. scale is good, mine works in centigrades only. Other features, You mentioned, are already implemented in my present prototype: it has 2 dimmable outputs @12V for LEDs and one for vent. Delay for achieving programmed power (user defineable) is also defined by user. Delay =0 means switch at once. Every device has a max. of 4 switching periods a day.
Ph measuring is not yet present. I understand that probes require calibration not only before first use but periodically. Another problem is to find a probe that can be in water for a long period (constantly).
I was more interested in measuring other parameters of water, like salinity that You pointed out, hardness perhaps, NO3, etc. It looks that there is no good electronic means for it.

I guess it's time to translate the "manual" for the prototype.

 

[felch]

Translated "manual": http://parsek.yf.ttu.ee/~felc/kontroller_eng.pdf

 

[Anonymous]

Salinity is done by a conductivity probe. In a simple device, it can be just two stainless steel wire, but there are probe used in oceanography that cost as much as a small car. Basically, you measure the resistance/conductivity between the two wire. The tricky thing here is that you need to drive the electrodes with AC, since if you put DC on it, it just electrolysize the seawater, and give off nasty gas.

For temperature, a simple thermosistor will work. Same here, measure the resistance (DC, unlike the conductivity).

For NO3, you need to make use of ISE (ion selective electrode). It is a real pain to use due to the ions in the seawater, and the technology is not that mature for the hobby yet. Same goes for calcium.

For pH /ORP, you need to measure the voltage generated by the probe. Literature are readily available, so I am not going to say much.

 

[_Adrian_]

been poking around the net for a decent processor ....
i've had some encounters with motorola products ( Megasquirt EFi )
and seen this on their site ....
http://www.freescale.com/files/shared/doc/selector_guide/SG1006.pdf?tid=tcDRsg

was looking at a couple of options that they offer....
http://www.freescale.com/webapp/sps/site/overview.jsp?nodeId=0162468rH3YTLC6951fgqk7SDQ

 

[richard.c]

Hi All,

Hope you do not mind me 'breaking in' but I was interested in your thread on the diy controller - just thought you might be interested in my well proven diy aquarium controller pic micro - temp and cooling the priority but with a few bells and whistles as well.

The original version was published over a year ago at http://ozreef.org/diy_plans/temperature ... oller.html
but unfortunately most of the files are now corrupt.
Also since then the controller has been considerably updated so if anyone would like the latest full details then please mail me at [email protected] and I will send the full set of files - about 1mb.

thanks for your time

Richard

 

[Anonymous]

from what i have read is that most probes out there need to be "calibrated" with a calibration solution before they can be used and they put out an electrical signal on the mV scale>

Exactly right. The calibration comes from using a fluid of known value, then calling that the "zero point" for which means "Value X" and things work from there.

 

[Anonymous]

Actually, it is a linear regression where you use two known value for the slope and zero-intercept.