Salinity Drops Every Time ATO Adds RODI — Normal or Something Wrong?

[Alexr54]

OK a stupid question and idea.

Is your salinity tester directly next to where the RODI is being filled during the ATO operation?

If so, that particular area may be temporarily less saline due to the concentration of RODI in the zone.

Maybe move the probe to another spot in the tank? Even just to another area of the sump.

The salinity probe is in the compartment before the return pump and ATO fill hose. So the fresh water does not directly affect the salinity probe without first cycling through the tank.

 

[Vivid Creative Aquatics Verified Sponsor]

I have a zone valve (like used for irrigation) directly powered by the Apex controller (24v) which opens/closes water to my RODI based off this code in Apex for the Zone Valve power. So when my ATO runs, its not a huge stream, its just as fast as the RODI can produce the cleaned water. Its directly from my RODI to my refugium/sump in the same compartment with the return pump.
The salinity probe is in the compartment before the return pump and ATO fill hose. So the fresh water does not directly affect the salinity probe without first cycling through the tank.

Fallback OFF
Set OFF
If WtrLev < 6.5 Then ON
If WtrLev > 7.1 Then OFF
When On > 010:00 Then OFF
Defer 005:00 Then ON
If FeedA 006 Then OFF

With this code i have already attributed for any possible water level sensing lag. And based off the graph trends it is stable.

What you see in the water level sensor graph is pure water evaporation.

But the code doesn't (and cant) account for any possible inaccurate readings from the water level sensor itself.

the Salinity probe seems stable enough from your graph, and is very clearly reacting to an abrupt change in salinity at each 12hr mark, and then holding there with the new reading until the next event.

At this point, based on all evidence you have provided, the Water level probe seem to be the possible weak link, since that one sensor dictates when and how much fresh waster to add.

I say this, because I run 4 nano tanks here at the office; a 5-gallon, a 10-galloon , 15-gallon and a 12 gallon shallow tank, that each have their own ATO controller, but share the same fresh water reservoir. Its incurably hard to get the salinity on these to move even a point. I have had the ATO run dry over night and walked into the office with them all blowing bubble - check salinity, its maybe deviated less than 0.025 from where it should be.

So when my ATO runs, its not a huge stream, its just as fast as the RODI can produce the cleaned water. Its directly from my RODI to my refugium/sump in the same compartment with the return pump.
The salinity probe is in the compartment before the return pump and ATO fill hose.

This here may also be hiding this potential issue. if you are making the RODI water as its being added to the tank, you have no way of knowing exactly how much as been added. If for instance you stored your RODI in a reservoir, you could easily see how much water was added in any given period

 

[Alexr54]

But the code doesn't (and cant) account for any possible inaccurate readings from the water level sensor itself.

the Salinity probe seems stable enough from your graph, and is very clearly reacting to an abrupt change in salinity at each 12hr mark, and then holding there with the new reading until the next event.

At this point, based on all evidence you have provided, the Water level probe seem to be the possible weak link, since that one sensor dictates when and how much fresh waster to add.

I say this, because I run 4 nano tanks here at the office; a 5-gallon, a 10-galloon , 15-gallon and a 12 gallon shallow tank, that each have their own ATO controller, but share the same fresh water reservoir. Its incurably hard to get the salinity on these to move even a point. I have had the ATO run dry over night and walked into the office with them all blowing bubble - check salinity, its maybe deviated less than 0.025 from where it should be.



This here may also be hiding this potential issue. if you are making the RODI water as its being added to the tank, you have no way of knowing exactly how much as been added. If for instance you stored your RODI in a reservoir, you could easily see how much water was added in any given period

Thanks for all your input.
At this point i think the best thing for me to do is wait a few days and see if this behavior continues after i recalibrated the DOS for automatic water changes.
Then if this behavior continues i will try using the ATK with digital sensors to trigger the zone valve.