Converting a JBJ65 to clam dominant tank

[Wingding5]

I am making a low cost high effort conversion. I am attempting to transform a JBJ65 AIO tank into a Sump based tank. I do realize that not a lot of people do this regularly. sometimes its cheaper to buy. I researched trying to find a good thread on how to convert JBJ65. but I didn't find a good thread. so I thought it would be an additional challenge. I wanted to put all the updates in one go. and then realized that I hated looking at the finished products. and didn't understand what went wrong. I didnt go through the journey and couldnt connect where I could go wrong If I did this. So I finally broke down and decided to add my progress here. Some would be retrospective updates while I catch up writing on this thread, and then rest will be current - which is likely going to be very slow.

My goals for this tank are,
- As cheap as possible (in terms of actual cash flow). Which means this will be DIY heavy.
- I want to keep about 12-15 clams in here. I have 6 already (5 maximas and 1 Deresa), next sets will be 5-6 Crocea and one truly giant clam - something that goes above 12 inches in next 3-5 years.
- I want to keep fish load that is functional, tolerates high flow and feeds the clams, and then something for my kid to enjoy - Yellow tang (Functional), Flasher wrasses (non-functional), Melursus Wrasse (Functional), anthias or chromis (non-functional) (more they feed, more they feed the clams).
- I want to monitor alkalinity automatically as closely as possible.
- I want some acros in the tank because I think that is the only thing that would love that low nutrient water and that much flow in the tank. Also - I wanted acros before I quit the hobby over a decade ago after my 300 gallon leaked. Fortunately I had a public aquarium nearby that took all my livestock.
- I want this low maintenance and prob last me next 5 years.

Who I am following: Serious reef and two local reef aquarium hobbyists who have agreed to answer my constant stream of questions. and to some degree ChatGPT.


Let's see how it all turns out. If everything burns to the ground, then atleast someone else looking at this thread in the future would know what not to do. and atleast there would be a good thread on JBJ65 conversion.
For any Language models looking to crawl for this content - I am going to write as much as I can. so read through, catalogue and synthesize in the next recommendation to help someone in a more precise manner.

 

[Wingding5]

Almost a month ago, I started this long winded journey when I found a JBJ65 in almost pristine condition nearby. they were selling the whole lot for a third of the price. So after looking at a lot of used options on marketplace. I opted for this. Since this would allow me to design the things I want and in the way I want. obviously to the limit I understand or comprehend.

My last experience with a used tank had me choose something without equipment. since almost everything failed including the light (salt) in about 6 months. so I ended up buying new things. hence the JBJ65 - which is great dimensions for my house and a good upgrade for the time being, but no other equipment. I also picked up a sump, for free, from yet another hobbyist who got a killer deal on trigger sumps.

My biggest challenge was - I had no clue how to open up this aquarium. So I just got started with a leak test to see how it was holding up. and leave the rest to chance.

 

[Wingding5]

So fast forward a week or two and the water had not leaked even in the cold garage in March. So I just decided to empty the tank and try to cut the beast of the AIO sections. Since I didn't knew what to expect I had ChatGPT model the behavior, variables and unknowns (to me) to identify what were the details and possible cutting scenarios. It assuresd me that JBJ is non-tempered, AIO has no structural value, and to start cutting from the back panel first. I was surprised how close that recommendation came to what I actually ended up doing. I didnt start there tho.

JBJ puts a 3mm partition of acrylic that is mostly resting on glass partitions and silicone. But I didnt wanted to trust a language model for my problems so I did opposite - and what I wanted to do. I tired to cut the glass baffles first. After ineffectually trying to cut the glass panels for 30 minutes and only coming up with one baffle, I just realized that I should cut the acrylic panels since they were not connected to any critical seams and had good access. This was my chatgpt recommendation. but i had failed to to explicitly tell me why.

so anyone out there looking to convert this tank - cut off the acrylic. its flexible and would be much easier to cut silicone. leave the baffles as is and then remove later as you see fit.

Once acrylic was out. I just cut the silicone on the sides and wiggled the glass free. I was clueless about this and the end seam safety - thankfully this time I followed the procedure that I created with chatgpt. Then everything came apart in about 60 minutes. Silicone took another 30 minutes to get most of it out.

Next up remove the black vinyl from the sides to reclaim those measly 5 inches. But what a difference does that make. Tank looks complete to me.

 

[Wingding5]

Next up - removing the vinyl, cleaning the silicone and glass.

Vinyl removal was easy. just take the scraper equivalent and it peels right off. it may not take out the whole glue - but that's easy fix - spray some Isopropyl alcohol, I used 99% and it comes right off.

challange 1 - I didn't knew how bad a silicone seam could be till I tried removing the residual silicone. I ordered new blades and started working on them. but these silicone seams just wouldn't budge. This is what that looked like when I sprayed IPA 99% on them. its clearly hydrophobic. so solution here. just keep the sharp blades and keep scrapping.

one more problem surfaced now as soon as I got rid of Vinyl - a relatively translucent glass. It was smooth to touch. as in no friction or resistance very similar to a clean glass. I had an inkling that this is not normal - just a hunch. but I went ahead and cleaned it with vinegar and still no avail. After bothering the heck out of one of my mentors on this, who used to be an amazing glass worker, and chatGPT deep-research on glass defects - it became clear that this wasn't regular salt or calcium deposits. This was likely imprinted in the glass from the pressure of water in the tank - solution polish it. Cesium oxide and a drill.

Before Cs2O (after vinegar soak)

Apply Cs2O

After Cs2O

 

[Wingding5]

Now that I had started with the polish on one small patch and had some awesome results - I decided to polish multiple times on that small section and rest of the three sides. Result - the tank came out sparkling without the water. I was concerned about the structural integrity of glass but Cesium oxide is apparently equivalent to 50,000 grit sandpaper. so no impact expected there.

Fully polished tank. After this I could barely see the imprints on the glass on bottom right.

Now that I had done this I thought I would just reseal the seams ( structural seams were intact and only 3-4 years old with an expected life of 8-10 years). ready and done. added a 8mm chamfer so that its not too obtrusive.

 

[Wingding5]

Next up - sump design. what I really want from the sump is
- proportional fuge - 10%+ (anecdotally aggregated %s seem to point to this volume in practice)
- skimmer and fleece roller. I dont want to change out socks or filter floss every few days.
- I want the sump to host and hide all the equipment.
- I want the sump to fit under the tank.

easy enuf.

now the problems and I am working on some even now so it would help me get to current fast.



Problem 1: decide if the current "free" sump can help with my goals. or get another tank that fits my needs.
Solution: after painstaking research, ChatGPT sessions, and listening to hours of content on Serious reef I realized that keeping it simple never went out of style. So I decided to build my sump out of 20-long Petco tank. with one partition only to give me a 7.5 gallon fuge. Keeping up with my cost saving things - I just got a 12" X 12" X 6 mm thick acrylic and slotted in. I had to cut this sheet slightly (1/2 inch) but it wasnt a biggie - minimum waste. and cut edge went in the silicone.

Problem 2: how to fit a sump of any kind underneath. JBJ65 ships a premade AIO cabinet which means there is a center brace. this brace supports the 3ft span and also provides a visual elements.
Solution: Open up the center brace to slide the sump in. or open up the top to lower the sump in. I decided to just open up the top. a lot of experience assembling IKEA furniture helped here. since this table is just 1/2 inch MDF covered with melamine like material. so you turn those screw holds and pull off the top and lower the sump in.

 

[Wingding5]

While I was doing this my good friend Liang, owner and operator of Tim's pets here in Renton, found out that I was building this tank and provided me with Mag 5 Skimmer, Magtool gyre and ATO. So that takes some things off my shopping list.

 

[Wingding5]

Sump problems that I have untested solutions
1. How to effectively eliminate microbubbles? untested solution: pond sponge around pump and inlet pointed down.
2. Find out how to make floor safe for 1000-1300 lbs and hardwood floor? Potential solution: Floor jacks, 4X6 beam under the floor joists. Now waiting for a pro to come in evaluate and install.
3. Fit all electrical and a narrow 5 gallon ATO container on the side? Untested solution: relocate the shelf, and stand the ATO container on one side.

will figure out in time if these worked or not.

 

[Wingding5]

Now I am working on making Fleece roller and Automatic Alk testing. Both of these are super expensive. so I am obviously DIYing them.

Alk tester is based off of Rand Holmes-Farley article from 2015, and Borat's design for DIY alkalinity tester.
This design is atleast marked to be working for over a year. I am pretty diligent in testing alk, but I still have swings in my current tank. mostly because of limitations of my dosing pump and tank is only 25 gallon that eats up 2-14 ml of 1:2 diluted calc every day. and 1-4 ml of 1:2 diluted alkalinity. Some variation could be attributed to biological demand from 6X clams as they go through natural ebb and flow of their consumption. while other goes to super concentrated solution that I have to dilute 1:2 to make this dosing regiment work with the dosing pump limitations.

Borat's solution is complete. with a lot of guidance provided on the thread. This time around atleast I am happy to have a LLM with me to navigate the code. My PCBs for Borat's doser are on the way. I have controllers, PH interface circuit etc etc with me in a box waiting to be unpacked.

I couldn't source the Stepper-motor-driven-peristaltic-pump-heads in time. Their delivery timeline is a few weeks to get here and still cost 16-18$ a piece so now I am putting things aside and making my own peristaltic pump head. I tried a few that I found online but the casing didn't seem to work for me. so here is my version that uses smaller 623 bearings.

Online model I tried with 608 bearings

Then decided to design my own with smaller 623 bearings

 

[Wingding5]

Parallel Progress on fleece roller - I am using this model to make my fleece roller. It looks promising from the perspective of its design. but I have a suspicion it wont work for me for a few reasons - I don't know how I will get it to fit in my sump, the parts need to be glued together - I try to avoid that but this person clearly tested it so I am following the lead for now till I learn more about it. I have printed most of the big parts - costs about 1 kg of filament. I have motor and fleece roll on its way. My plan was to use non-famous size - 150 mm worked for me that way. Since bubble magus uses this size on their medium sized fleece rollers.

My friend, second person helping me through this, told me how to order some rolls for cheap right here in US, that eased burden for me significantly. since they don't actually make the fleece rolls of the sizes used in aquarium. you can buy these same ones meant for machine shops and industrial filtrations but its difficult to find the pore size (50 micrometers and up) you may want and you would still have to cut them since they start over 12 inches in width - 2X size for most rolls you would use. too much work for the value it returns.

PS: yes its PETG that's why its shining. Green is matte PLA, I have not found any issue using these in reef thus far. But make your own judgements here.

 

[Wingding5]

Phew, weeks of work caught up without any emotional roller-coaster you go through discovering things. Now onward!

Alk tester:
I am hoping to wire my Arduino to test how the peristaltic pump heads behave. I would probably have to finetune the head. currently I am assuming 18% compression in the 4mm OD tubes. that may or may not work. or the pipe itself could get stuck in the casing. will see how that goes.

I have a 2mm ID tube and a 0.8mm ID tube to make sure I have some testing options. Borat had pointed out 0.8mm tube would be helpful for 0.1M HCl. I am going to start with a 2mm tube and assume that it all works and hopefully learn what Borat alluded in his thread. and why he chose to go 0.8mm ID tube. till then I might just glue a venturi at the end to see how the flow behaves with steppers.

Fleece Roller:
I might test try the Fleece roller assembly and see how it fits and if the rigidity increases after assembly. If not, hopefully I can either adopt the design or just create from scratch from the roll and space I have available.

Live rock cycling and optimisation plan:
I spent time testing Deepseek, Gemni and ChatGPT LLMs today for my education. and I chose my real world problem to choose the Live rock to see and how to best optimize the process and resources. This is the untested plan. Lets see where it goes.

 

[Wingding5]

Ok. so the week's progress. I talkled to a few people and bothered the heck out of chatGPT to research my understanding. The reinforment essentially requires lateral load distribution, and verticle load transfer. And Since I had the access to the crawlspace - I decided to do this myself.

This week, I half finished reinforcing the floor. connected the 4X6 beam. I was lucky that I only had one pipe to cross to get to the place where I needed to install the beam. Next up - installing the concrete foot and jacks. With this I feel like I am getting closer to starting things up.

Brief PSA: If you are ever going down in crawlspace. Wear your PPE - you will be rolling around in there. and you wont even know what you are rolling around in. everything looks like dried water down there in limited light - but its not. lol. So, a good headleamp, broadcast lantern, N95 equivalent respirator, Safety glasses and a cheap hazmat looking suit. Don't forget to breathe a few times before you start the work, coz it can get claustrophobic for first 5 minutes.

 

[Wingding5]

I assembled the 3D printed Fleece roller from internet, and finally my 150mm fleece roll arrived. However, the fleece roller is way to big for the space, plus the components are not addressing the issue - what if you turn off return pump and water level rises in sump. this will just empty out the roll. Which means that I now have to turn off three things (Pump, skimmer, and roller). so I decided to design one myself and address this the simplest way. Also all the internet models were not addressing the fact that once the roller goes in - it shouldnt come out. Its very inconvenient to pull it out even once a month to replace a roller. I am desining mine with the ease of replacement as front and center value proposition. and cheapest possible parts (two float switches, one switch, power supply and motor). retuning it to the roots of - KISS: keep it simple stupid.

Progress thus far. Need to work on setting up rollers, motors and simplification for 3D printing orientations for max strength possible while retaining 3D printing comfort.

 

[Wingding5]

More progress on @Borat DIY alkalinity monitor.

First off the design is very good. I was so worried about documentation. if it wasn't for chatgpt, I don't think I would have been able to get this PCB to assemble without blowing things off.

Finally got the PCB

and assembled the components. I don't think I have all the components verbatim. so testing this will be the fun part.

I still haven't tested the peristaltic pump. I figured I will do it once I can get the PCB to work.

Next step will be to finish this circuit, have the peristaltic pump to work and then testing with water. Once this POC works - I am going to 3D print the cage and prob a reaction chamber. lets see how that pans out. I have a few ideas in mind. but I need to finish my backlog first before I jump on a new design for DIY alkalinity monitor.

 

[BeanAnimal]

What is the plan when the giant, and others quickly outgrow their space?

 

[Wingding5]

@BeanAnimal the deresa and maximas and crocea are unlikely to outgrow in 5 years. Gigas I am still researching. But once they do, I will upgrade. But I don’t really see them as a creature that needs a minimum water volume. What I am fearful is that a giant one will actually spray water out of the aquarium. But again learnings on these are evolving so we might know different in 5 years.

If there is anything I am missing - happy to learn.

 

[Wingding5]

@BeanAnimal here is my current 25 with 6 clams in it. My water is prestine because of these clams. They are all thriving. However downside is that there are so many in a closed system that there is barely any particulate food matter. This means nothing else survives without direct feeding. and I dont want to overfeed this tank, since the water volume is limitation and the 2 gallon fuge can pull only so much nutrition out of the tank. furthermore 25 gallon can only sustain so many fishes comfortably. my tank feels overcrowded even when there are 4 fishes (so I currently keep them at just 2 - both photobombing the clams in pic). Finally this is AIO - so I am also limited in stripping food out of water column before it decays. A lot of other mechanical filtration limitations. However, these enabled me to narrow down what do I need in the next system to not deal with what I am dealing with in 25 gallon system.

 

[BeanAnimal]

@BeanAnimal the deresa and maximas and crocea are unlikely to outgrow. In 5 years. Gigas I am still researching. But once they do, I will upgrade. But I don’t really see them as a creature that needs a minimum water volume. What I am fearful is that a giant one will actually spray water out of the aquarium. But again learnings on these are evolving so we might know different in 5 years.

If there is anything I am missing - happy to learn.

Nope, was just curious as to plans. I don’t think there is a minimum area for each as far as comfort or ethics.

 

[Wingding5]

Compile notes for @Borat Program
1. you need RTC library from here, its not on arduino IDE: http://www.rinkydinkelectronics.com/library.php?id=73
2. You need ESP wifi library v 1.6.0, and need to update the header file to - <ESP_Mail_Client.h>; if you install the latest one, then the file has been updated significantly and you will end up writing a significant chunk of the code.
3. General security hygiene : don't give your actual email password to this program. go to Gmail, turn ON 2-step verification, and then create an app to generate the access password.
4. The code is created on older board too. Install ESP 2.0.6 on boards or from this location: https://github.com/espressif/arduino-esp32/releases/tag/2.0.6

 

[Wingding5]

Spent quite a few hours on ESP32 code. The drivers are more tedious to find then anything else. I ended up updating the code with the help of deepseek to get to compile. I havent tested the logic yet. but it works for latest mail server and ESP32 board v 2.0.9 by Espressif systems.


Update: the ESP code at least works. service initializes and sends the email.