DIY Hydros 3D printed roller filter and lessons learned

[SlowAndStupid]

After numerous hours of designing, 3D printing, further tinkering, and fine-tuning I have been successfully running my own custom-made roller filter in my DIY plywood sump for the past 2 and a half weeks. What follows is hopefully a well-documented post on how I went about designing this roller, what issues I have run into, and how I have adjusted for them. I'm sure it's not perfect but as I said above everything has been working for several weeks now and I am exporting much more waste than I was previously.

Bear with me, it’s a long post. There's enough pictures and rambling that I have to post this as several sections.

About the tank: This is a custom 1000 gallon aquarium measuring 8 feet long by 4 feet deep and 5 feet high. It’s the first aquarium I ever took care of as it came with the house I purchased. Talk about jumping into the deep end! Feel free to check out my tank thread for further information on the whole thing (https://www.reef2reef.com/threads/1000-gallon-tank-for-a-beginner.1031993/ ).

How it began: Space is tight in the small maintenance room behind the tank (it was very poorly thought out and designed), and I have had to really plan out how to make the limited space work. I currently run two plywood sumps built by myself that are stacked on top of each other.



The water from the tank enters into the top and down into the bottom sump and is then pumped back up the tank from there. I have always wanted to have a roller filter for the tank so I tried to design the sumps with that in mind. But my priority was to get the tank up and running and a roller filter was not a necessity at the start.

Instead, I designed and 3D printed my own filter socks and holder for these socks. All 3d printed pieces were printed from PETG (high flow from Bambu and rapid PETG from Elegoo based on availability). All pieces appear to be holding up well.




The 3D printed filter sock holders are 2 pieces that I then place 5 gallon paint strainer bags in between to act as my filter socks. I purchase these (link) in bulk and clean them a few times and then toss them. The holders are roughly 7 inches wide and they would get clogged about roughly every 48 hours requiring either cleaning or replacement.

Here it is in action:


I finally got to a point in my build were I could turn my attention back to the filter roller in an effort to cut down on the frequency of necessary maintenance. I sat down in fusion 360 and began designing this monstrosity.



This was printed on my Bambu Lab P1S. I initially designed my piece and then broke them up into smaller sections within fusion 360 so that they would fit the build plate of the printer. I consider the filter to be made of four main parts:

The Hanger:


As I mentioned previously space is limited and the drain pipes from the tank obstructed my ability to simply mount a large roll of filter paper directly above the sump itself. Instead, I designed a separate piece that filter roll will sit in, modeled after a paper tower holder. This was printed in 3 separate parts and superglued together.

The filter roll sits on a dowel made of PVC pipe press fit onto printed ends.



R20-2RS bearings (link) fit over the end pieces and slide into slots into the hanger which allows the paper to easily roll. Lastly 12mm diameter acrylic rods are used in this design (link) and one slots into the back of the hanger during assembly and the paper is passed behind it directing it down towards the sump.



The hanger fits just above the return pipes and in its initial print would fit a 12-inch wide 100 yard roll of coolant filter paper purchased from ebay from the user “preconfilter” (link). While this was great I dreamt bigger. I designed spacers to place into the original hanger giving me more room to work with.



The hanger will now support what I jokingly refer to as the “vacationator,” a 250 yard roll of filter paper. At 50 microns it filters well and this roll will last quite a while.

The Starter Section:


Again separated into 3 pieces for printing, this section guides the filter paper from the hanger down into the water. Two more acrylic rods are placed during assembly and help to guide the paper down without flopping too much one way or another. This piece sits over the back of the sump and the top of it is sloped back towards the sump so that any condensation that might form on it gets directed back into the tank rather than out and behind it.



The filter paper passes behind the top rod, in front of the bottom one, and down behind the next rod as it enters the water.

The Filter Grate:


Printed as 12 main pieces and then glued together along with 4 acrylic rods to guide the filter roll along the grates. It’s a tight fit to pass the paper through but not too difficult to accomplish. At the top right of the picture above you will see a mount for my float valve. My aquarium is connected to Hydros controller and I purchased a Hydros float valve to allow for management and programming through it. More on that further down.

The grates sit within the sump with the starting and ending guides sitting over the front and back portions of my top sump like so:

 

[SlowAndStupid]

That brings me to the fourth portion of the build:

The End Section:


This portion was printed as 11 main pieces (in white) along with 3 more acrylic rods to continue guiding the roll through the filter. Within the end piece is a bucket I designed that can be lifted out to carry the dirty roll if I wish to do so or to at least be removable for cleaning. It is assembled from 4 printed parts.






In the picture of the end piece a couple of images above you can see the bracket for the motor for this whole contraption. I purchased a 12 volt, 5 RPM motor from amazon (link) (and a second as a backup). It has impressive torque and does its job well.

Power supplies always worry me considering they could be a fire hazard. I went with a Mean Well 12 volt 5 amp power adapter purchased from digikey (link). The adapter sits in its own housing which I printed in leftover ASA as I really didn’t know how much heat it would be exposed to with the motor.



I designed spaces within the motor housing and the end the endpiece to house magnets (link). The motor will sit within a bracket but holds firm against the filter as well. The large piece above the motor is for a bracket to hang the housing onto as it has to be removed from the dowel to be able to freely spin the dowel during changes.



The dowel that the filter roll ends at is similar to the starting one but with a small adaptor that connects the roll to the motor.



I find it much easier to remove the adapter along with the motor and it is much easier to reseat the square adapter back into the dowel rather than the D-shaped motor rod.

The end piece is fairly heavy to simply dangle over the side of the sump. I then printed support pieces that attach to the T-slot aluminum stand that the sump sits on and the large end piece rests onto it.



And here it is in all its assembled glory:



And here it is in action:


Issues Discovered:

I ran into a couple of troubles once I had the filter up and running.

1. The filter would oftentimes get pulled off-center both within the sump itself and on the end roller as the waste material was collected.

I printed guides that snap onto the acrylic rods to keep the paper centered, and it seems to be working well.



I have them on both the starter and end pieces.




Next, I designed guides for the end roller as well.




Thankfully I saw that as potential issue and had an opening created into the original design for these to easily slot into.



This keeps the roll well-centered as it rolls up.

2. The turbulent water as it spilled into the grate area would pull the filter paper into the water and then allowed water to pass behind and bypass filtration.

I solved this issue by printing an additional set of guides that prevented any slack in the roll from falling into the water as seen here:




It now stays snug and out of the way.

3. I designed the grate for a 12-inch roll but designed the grates to be almost 12 inches wide. This meant that when the filter roll got pulled barely off-center water was passing through open portions of the grate and bypassing the filtration. Enough water was bypassing the filtration that the water level was not rising enough to activate the float valve so it would filter one time until the paper in the water was clogged and then all the water passed through the open portions of the grate until I went and pulled the roll back into center.

I corrected this by printing pieces to block off the half hexagons on both grate edges giving me more dead space to work with. I also was worried that despite making this correction I could still run into this issue and I did not want the filter to potentially sit there all day not filtering until I came home and discovered the issue.

I sought to correct this by programming in a timer to roll the filter paper at a given time regardless of whether the float valve was being activated. What I wanted was to be able to program it in such a way that the valve could activate the roller and the roller could activate at a specific time. I could not initially figure out how to do this through the Hydros controller. I sought assistance through the Hydros facebook group and reached out to CoralVue as well but had no luck. On my second week of running the filter we went on a 10-day vacation to Disney. I ended up just programming it to run on a timer during that period and it worked great. I have a webcam set up and pointing right at this thing so I checked in on it constantly and it did great. When I returned home I sat down and figured out how to program it how I wanted, more on that below.

4. The turbulent water as it spilled over the initial baffle and into the grate area was turbulent enough to lift the edge of the filter paper and water was again bypassing the intended filtration. I designed a little piece to direct water further into the grate portion of the sump but made it too long and the turbulent water was hitting the far baffle and doing the same thing to the other side of the filter paper. I goldilocks’d it, redesigning it so the water spilled into the middle without adding turbulent flow to either side.




5. I eventually noticed the torque on the roll and subsequently on the final 2 acrylic rods was enough to bend them downwards. I think this was also contributing to pulling the paper roll off-center. I designed a couple of supports to place below the final two rods to prevent this.



And with those issues addressed we have been running very smoothly and consistently.

6. Salinity loss? There is a good bit of water being rolled up and into the end piece. It is entirely possible to see a drop in salinity over time as the ATO refills that loss. I am monitoring this daily with a salinity probe but also check with a Tropic Marin hydrometer as I continue to monitor this. I can report no significant drop the week I was out on vacation so thankfully anything that does happen should do so slowly. I can correct this with occasional top-offs of saline into the tank rather than the typical RODI water with the ATO.

7. Swings in water height. If you pull the entire dirty portion of paper out of the filtration area for a full clean portion of paper this results in large swings in water height at the end of the sump where the ATO activates. As the filter paper clogs and water passage slows the relative height of the water column at the ATO drops and fresh water rolls in. The float valve activates and new paper rolls in. Suddenly more water passes through and the sump seems overfilled at the end of the sump. I now only pull a portion of the paper through each time. This results in more frequent activation of the motor but as of this post it activates for 8 seconds every 20 minutes which I find acceptable.

 

[SlowAndStupid]

Hydros programming:

As I said above I set out to program the filter to activate the motor when the float valve activated AND at specific time intervals. Doing either is very easy. Doing both is not intuitive. Here’s how I accomplished this programming:

First, I set up the float valve as your typical sense port. I did flip the float valve over to so I could activate when wet rather than dry just as a personal preference.


Next, I made 2 advanced/generic outputs that I named RF Valve and RF Timer. I made them invisible to keep my UI clean. The RF Valve one is pretty straightforward setting it to activate when the float valve is wet. The RF Timer is just a timer set to activate at my desired time frame. It has no inputs or outputs associated with it.






I then created a third output as an advanced/combiner output with the 2 outputs above. I set the combiner mode to OR and pointed it to the XP8 outlet my motor is plugged into with the advanced settings I wanted as well.


And now my roller filter will turn for 8 seconds each time the float valve rises AND once an hour as a backup!

And with that this long post is complete! The amount of time spent sitting at the computer and designing in fusion 360 is likely over 24 hours. According to my Bambu slicer I used just over 20kg of filament with a total print time of 14 days, 21 hours, and 36 minutes. In then end, I have a functional filter that is filtering out more waste than I think my filter sock did and significantly improves on the amount of maintenance I have to do to keep up with my large tank, and, for that, it is worth it. I am pleased.

 

[Gumbies R Us]

Wow, this is impressive! Thank you for sharing!

 

[PPBlimpy]

This is AWESOME. Coming from a 5-axis aerospace cnc guy that does lots of design and manufacturing using Fusion360and a love for 3d printing. I love it! I have been thinking about making my own roller for my system, much smaller though and this give a ton of inspiration.

following along!!!!

 

[PPBlimpy]

Also using an off the shelf coolant paper is genius. I was looking at a reefmat 1200 but would rather spend the time building my own.
I was thinking about making it a stand alone type system, using the float switch to activate a temporary feed timer. Similar to your 8 second of feed.

Got me thinking. Anyways great project!

 

[pwnzorman]

This is great work, thank you for sharing!

 

[reely989]

this is one of the coolest things I’ve seen on here

 

[PPBlimpy]

I have come back multiple times to reference your design and updates. I am slowly progressing on my design.

I really appreciate you posting this. Its a great design!

I think I am going to use a JBJ ATO controller for the brain portion. I have a couple on the shelf and i think the logic will work. Has 2 float switches and can plug a 110 motor into it. Idea is when the float switch is activated because of reduced flow/dirty filter it will advance the roll. I will use a 24v gear reduction DC drive motor with a potentiometer to adjust speed at which it rolls. Have not gotten there yet though.

Anyways just want to say thank you again for posting!

 

[SlowAndStupid]

I have come back multiple times to reference your design and updates. I am slowly progressing on my design.

I really appreciate you posting this. Its a great design!

I think I am going to use a JBJ ATO controller for the brain portion. I have a couple on the shelf and i think the logic will work. Has 2 float switches and can plug a 110 motor into it. Idea is when the float switch is activated because of reduced flow/dirty filter it will advance the roll. I will use a 24v gear reduction DC drive motor with a potentiometer to adjust speed at which it rolls. Have not gotten there yet though.

Anyways just want to say thank you again for posting!

Awesome, best of luck. The float switch alone should work though I like having it hooked up to my controller where I can access it remotely if needed and it alerts me if the motor tries to run for too long (it alerts me and turns it off) and that happens when the roll runs out and I know it needs changing.

 

[PPBlimpy]

Awesome, best of luck. The float switch alone should work though I like having it hooked up to my controller where I can access it remotely if needed and it alerts me if the motor tries to run for too long (it alerts me and turns it off) and that happens when the roll runs out and I know it needs changing.

I am not all that tech savy to figure that stuff out yet with the controller. Just got my first apex about a month ago. Came with a used system. I had been looking at hydros.

Before that I was running reefkeeper lites from circa 2010 on my systems. figured the ATO setup would be an easy think to adapt

 

[mmorrison55]

Very cool filter roller.

 

[ReefingRealm]

Any chance you'd share the stl files for these ?

 

[PPBlimpy]

Any chance you'd share the stl files for these ?

maybe once i get it all printed and tested. project is put on hold for a bit, had a tank crash and trying to get my ducks in a row while i rebuild