Duke link said:
What sort of failsafe's do you have built into this thing incase it messes up, trusting your tank to a freeware home built controller seems risky doesn't it? Do you have an estimated total cost of this endeavour?
You are right - it is risky. It is a prototype.
- For start I will keep it for monitoring purposes for a period of let's say six months.
- Lights can be connected right away. If they fail to start and stop? I don't think is a big deal with huge consequences to the tank. I should be able to catch them in time before is too late. Anyway they will not exceed the maximum light level set in LED drivers.
- Looking at the main inputs/outputs, most of them are for monitoring and generate alarms (water level, leak detection, etc). No direct risk to the tank.
- pH is going to be monitoring, no risk if calibrated/verified periodically.
- Power heads for water flow are somehow medium risk. If they fail the tank is going to be without internal circulation for .... until I catch them. One day, two?
- Skimmer and return pump have high risk. I will keep them in testing for longer time. I will connect dummy bulbs and monitor their functionality for long time before I put them in service.
- Heaters in my opinion are very high risk. If the logic fails you can have the tank "boiling" or "freezing". I plan to have a sort of redundant system in place between this controller and the current temperature controllers that I have right now. At least for a while.
- If the controller is freezing, there is an auto-reset function (thanks KBennett).
- I will make provisions for heater, skimmer, ret pump, etc, if the board or connection fail they will stay ON (thanks Adam and Phil).
In terms of money, the cost would probably be 50% of a commercial controller. In terms of time spent .... is not worth, unless you like doing it.