View Single Post
Old 05/10/2011, 04:18 PM   #5
Registered Member
Join Date: Apr 2011
Location: concord
Posts: 1,280
Originally Posted by kenargo View Post
IMO; I am very concerned at the use of floats to manage a water exchange; I use and have had the best luck with dosing pumps but if you want to go this route I would suggest some minor changes.

1) Turn off ALL pumps in the area of the floats; you want very calm water so that the floats don't bobble - Makes sense. If i give this a try, i will definitely do this.2) Cond is a measure to protect; using it to control absolute is problemmatic since salinity is affected by temerature so you can get into trouble relying on cond too much - I suppose i could do some checks on temp as well. This would prove problematic (as stated)3) You need to start with a known reference; that is to say you must have things in such a state that the operation is repeatable. To reach this point you must have the 1st part of the operation to be replace evaporated water; this sets the start - I had my original code written this way with just water changes. I actually changed it this way, so that after a the water removal, it would check conductivity levels and then choose to refill with salt or fresh, depending on if cond was low or high. Essentially forming a minor correction to be later read. I definitely have not done the math on how much of a correction that would make. I mainly did it this way as food for thought. If i could not get good stable corrections, i would definitely roll back.
3a) Now empty to a low float - I was a little concerned about emptying to a low float. I guess that if my tank is always topped to the same exact level, that would make perfect sense though. I was trying to compensate for unknown levels and drain rates. I guess ill just empty for say 45m or till float say's its empty, etc
3b) Last add new salt back to the level of step 3

Cond should be used to override the operation; that is to say if the tank isn't in expected range then don't perform the exchange. Trust me, you really want you to be involved at this point. Cond could be high/low because of temp, air bubbles on the probe, etc so best to use it as an override than a control (IMO). Definitely the safe bet. No argument there! Im just trying to figure out how to thread the needle a little bit on this one

Hope this helps!
comments in red. thanks a lot for your feedback!

Everything else aside, im getting the impression that cond is definitely too unreliable. That sucks, considering the price of the module/probe. IF i could get better temperature/variable adjusted results, I definitely believe that i could make this happen.

I dont mind checking salinity every few days. I just didnt want to have to manually adjust the levels. Even know it seems like a risk, I may just give it a try (with suggested adjustments, etc) and watch my levels daily. More for the sake of doing it, than needing it.

insomniac2k2 is offline   Reply With Quote