Planted tank for tetras (slow burn)

[eatyourpeas]

Eheim 2217 was temporary. It just had some sponges with filter floss in it.

BTW gluing acrylic is difficult and needs a LOT of practice. Most will be using Weld-On 3 or 4, capillary gluing technique, which is what I did. I used the Tensol 70 to fill the corners to strengthen the welds, insurance policy if you like.

I can make the tower and sump using Tensol 70 alone but it would a whole lot messier and require extensive masking on the gluing pieces, only advantage of it being gap filling is that it is more tolerant of imperfect dimensions since it can be use to fill gaps, capillary gluing require much more precise dimensioning of pieces.

I will keep that in mind. It looks like I will be gluing 25mm thick panels.

 

[eatyourpeas]

Eheim 2217 was temporary. It just had some sponges with filter floss in it.

So, how is the creation running? Are you happy? Still debugging?

 

[Advanced Aquarist]

Why are you gluing 25mm panels? Unless you are doing something really big it's not worth it. Also capillary gluing on 25mm panels is problematic if it's not dead square and flat when the pieces touch, for this reason Weld On 40 is used for gluing together larger pieces. 25mm acrylic is also horrendously heavy as well.

If you want to know what gluing with Weld-On 40/Acrifix 190/Tensol 70 is like here is an example, a lot of work with a lot of machinery needed, not for the faint of heart:

This is an example of them using the gap filling to fill in the "channel". The glue becomes one with the plastic. For 90 degree joints you create a "trench" for the glue to pour in, let it set and then machine it off. All a lot of work.

So, how is the creation running? Are you happy? Still debugging?

Tank is fine. Had moved to computer control and monitoring of temperature. Also need to change ballast to be dimmable, so I can create sunrise and sunset effect so as not to shock the fish, the computer program for that had some bugs which I fixed three weeks back, all I need to do now is to change the T5 ballast and wire in the 0 - 10 Volt connection from the computer controller. And I will NOT be using fixed lighting schedules, I will be using a variable lighting schedule taken from a point on the equator, so I program the sunset and sunrise times in the computer to set the lights according to what day in the year it is.

 

[eatyourpeas]

Why are you gluing 25mm panels? Unless you are doing something really big it's not worth it.

I am trying to do a Puget Sound biotope. Cold water tank so I need to control condensation.

Thank you for the video tutorial. It will be bedtime watching tonight so I'll report back tomorrow.

I was contemplating doing a non-photosynthetic environment, but I have to make sure I do not corner myself into having to pressurize the tank. I am definitely not experienced enough to do that! I like your approach to lighting. If I let my husband get involved in that aspect of the tank he'd probably get an Arduino and program the whole 365-day Pacific Northwest cycle! As he says, "you can always trust an engineer!"

 

[eatyourpeas]

This is an example of them using the gap filling to fill in the "channel". The glue becomes one with the plastic. For 90 degree joints you create a "trench" for the glue to pour in, let it set and then machine it off. All a lot of work.

I finally got around to watching the video. It is a lot of work indeed. I have ordered some material to make test pieces out of. I will be looking into making the files for the CNC to cut the trenches out. The glue will be ordered in about two weeks since I will be gone for several days. This has been a very informative post, so thank you again.

And I will NOT be using fixed lighting schedules, I will be using a variable lighting schedule taken from a point on the equator, so I program the sunset and sunrise times in the computer to set the lights according to what day in the year it is.

If you are at the Equator, your cycle will always be 12/12, unless by day of the year you mean rainy season with more cloudy days. Is this how detailed you'll go?

 

[Advanced Aquarist]

You have to practice on some scrap materials first. Also look at the Acrifix documentations in regards how they want you to use Acrifix 190, there's some videos on that too.

What you do is bevel the edge of one of the pieces to be glued then set the two pieces together (mostly with clamps), and in this bevelled gap you pour in the glue, like in this video:

They're using 192 in this video, that's a one part glue that hardens by UV exposure. Acrifix 190 (Weld On 40) is stronger than Acrifix 192. However the problem with two part mixtures is that you have run into problems in how well you mix. If you did not mix it properly, or not consistent enough, you can get weak bonds AND you won't be able to tell even if your bonds are bubble free. With capillary bonding using Weld On 3 or 4 if the bond is bubble free you KNOW its good. With two part gap filling glue you don't, so you have to be very thorough when you mix the glue - and it has a very strong smell BTW!!

The last comment might sound trivial to you, but even the large commercial tank makers get it wrong sometimes. Here is a very famous tank that had that problem:

Coral Tank from Canada (1350gal Display Tank) - split - Page 35 - Reef Central Online Community

Page 35- Coral Tank from Canada (1350gal Display Tank) - split Large Reef Tanks

reefcentral.com

They suspected that the Weld On 40 application was bad at the crack area. This was a very expensive system, I believe the crack took 4 years to develop, use of strong flow pumps for a reef system didn't help either.

How big is this tank you want to make?

 

[Advanced Aquarist]

If you are at the Equator, your cycle will always be 12/12, unless by day of the year you mean rainy season with more cloudy days. Is this how detailed you'll go?

I don't factor in rainy/cloudy/thundery days, they are random occurrences. I just take the sunrise and sunset times from the locality in question and program that into my lighting schedule, so for example here is my times for this month:

3,1,7:17,19:19
3,2,7:16,19:19
3,3,7:16,19:19
3,4,7:16,19:19
3,5,7:15,19:19
3,6,7:15,19:18
3,7,7:15,19:18
3,8,7:14,19:18
3,9,7:14,19:18
3,10,7:14,19:18
3,11,7:13,19:18
3,12,7:13,19:18
3,13,7:13,19:17
3,14,7:12,19:17
3,15,7:12,19:17
3,16,7:11,19:17
3,17,7:11,19:17
3,18,7:11,19:17
3,19,7:10,19:16
3,20,7:10,19:16
3,21,7:09,19:16
3,22,7:09,19:16
3,23,7:08,19:16
3,24,7:08,19:15
3,25,7:08,19:15
3,26,7:07,19:15
3,27,7:07,19:15
3,28,7:06,19:15
3,29,7:06,19:15
3,30,7:06,19:14
3,31,7:05,19:14

Program just reads the entries, so for 4th March we're starting at 7:16 and will finish at 19:19. Times do vary depending on what date it is. By the end of the month we have sunrise at 7:05 and sunset at 19:14, already a significant shift.

 

[eatyourpeas]

I don't factor in rainy/cloudy/thundery days, they are random occurrences. I just take the sunrise and sunset times from the locality in question and program that into my lighting schedule, so for example here is my times for this month:

3,1,7:17,19:19
3,2,7:16,19:19
3,3,7:16,19:19
3,4,7:16,19:19
3,5,7:15,19:19
3,6,7:15,19:18
3,7,7:15,19:18
3,8,7:14,19:18
3,9,7:14,19:18
3,10,7:14,19:18
3,11,7:13,19:18
3,12,7:13,19:18
3,13,7:13,19:17
3,14,7:12,19:17
3,15,7:12,19:17
3,16,7:11,19:17
3,17,7:11,19:17
3,18,7:11,19:17
3,19,7:10,19:16
3,20,7:10,19:16
3,21,7:09,19:16
3,22,7:09,19:16
3,23,7:08,19:16
3,24,7:08,19:15
3,25,7:08,19:15
3,26,7:07,19:15
3,27,7:07,19:15
3,28,7:06,19:15
3,29,7:06,19:15
3,30,7:06,19:14
3,31,7:05,19:14

Program just reads the entries, so for 4th March we're starting at 7:16 and will finish at 19:19. Times do vary depending on what date it is. By the end of the month we have sunrise at 7:05 and sunset at 19:14, already a significant shift.

I see. Are you trying to mimic the Rio Nanay? Those times are close enough to Iquitos but not quite, still off by a few minutes.

 

[eatyourpeas]

They suspected that the Weld On 40 application was bad at the crack area. This was a very expensive system, I believe the crack took 4 years to develop, use of strong flow pumps for a reef system didn't help either.

And that was a very sad and sobering experience! Ouch! I have not gone through the whole thread but plan on doing so while I am away.

How big is this tank you want to make?

I am staying nano. Unfortunately I still do not know the final dimensions of the tank. Ideally it will house a couple of Rhamphocottus richardsonii or Eumicrotremus orbis. The grunt sculping will benefit from a taller tank with rock shelves since it is not really a swimming fish. If I go with the spiny lumpsucker the tank will take more of a horizontal format. It could be a big cube as well.

Even at the nano scale, the 25mm acrylic thickness would be desirable due to the condensation issue. I am not so worried about the weight since I am probably not going more than 60 liters. Also, my partner in crime (husband) would prefer to go through all this trouble and have the option of being able to locate the tank anywhere, even affording an all around view, so that throws the 3-sided insulation idea out the window. There are still a lot of components that need resolution in terms of location and proximity to the tank.

Following your advice, I should use Weld-On 4 for the first bonding, and then follow up with Weld-On 40 for extra strength. I could do a 90º corner L, or cut the sheet edges at 45º with 20º bevel to channel the Weld-On 40. More complicated and labor intensive on the cutting part, but doable. Then the pieces can be supported at 45º so the pour fills the channel. Lots of playing around when I get the material. This approach involves less clean-up fabrication to finish the pieces, which I do not have the patience needed to do a good job. Realistically, I am about a month away from assessing this part once all the travelling is out of the way.

It does seem like a crazy adventure for such a tiny tank and tiny fish, but the learning process is fascinating!

 

[Advanced Aquarist]

60 litre tank using 25mm acrylic sheets would be easy - you have a lot of leeway for "errors".

Because of excessive wall thickness to tank size you can afford to make mistakes. I recommend doing what they did in that video. Bevel cut one edge to be glued, pour the Weld On 40 mixed glue in, and machine off the excess as they did in the video. Do this one side at a time. Then dry the thing for a month or so.

Capillary gluing using water thin Weld On 3 or 4 would easy too, as in this video:

A lot less messier than going the Weld On 40 route. You just have to make sure that your edges are DEAD FLAT. If there are leak bubbles showing you can fill them up with Weld On 40 later.

 

[eatyourpeas]

Oh dear, my apologies! I just realized I completely hijacked your thread!

 

[Advanced Aquarist]

We'll talk about syphon box arrangements now in two parts.

First a view of the sump some months ago. I run my tank at 25 degrees celsius even though most advise 23 for planted tanks.

The room is about 20 degrees so there is a fair amount of condensation building up. Because the sump is mainly sealed this condensation shows on the inside walls, preventing the cabinet getting damp so I can have electronics inside. I have yet to add in all the RO dosing lines adaptors, there is the cable for my Schego titanium heater and at the front is my temperature probe cable from the Elitech unit (which itself is controlled by a Dennerle Duomat on the other side of the cabinet).

On the opposite section we see the Duomat and Abyzz controller with all remaining cables, the CO2 cylinder looks beat, because it's a rental. Again the trickle tower is completely sealed so the all electrical equipment are safe. Rotation rate of the spray bar is about 60 rpm, I can adjust this, and there is no need for it to be so fast.

The two containers are magnesium sulphate and the other glass one potassium phosphate which I dose occasionally. This is a mess at the moment which I fixed later on. There are two timers. One for the lights and the other for the CO2 solenoid valve. CO2 goes on with the lights and off one hour before lights out. At the moment this is fixed times and no dimming. I eventually upgraded to seasonal time cycles with dimming which I will post about later on. It's a fair bit of clutter.

Now regarding syphon boxes, I'm running the reef octopus one which I got cheap:

I had to make a bracket for it to fit on my tank without tilting.

This is not a great unit, and most people who run these are advised to get an aqualifter pump in case you lose syphon. The question is: when and how do you lose syphon? That's never been really answered by anybody so I'll go into that next. What a lot of people don't understand is that syphon boxes are so much more complicated than a simple barrier arrangement that most reef ready tanks have. This is waterbox's implementation of their overflow system:

You basically have one overflow barrier (the grate in black at the top), and two drains. The left pipe is the return line with the bleed off manifold if you want to run reactors. The main drain is the in the middle and the water level in this chamber is set by it, it has a gate valve to adjust the resistance so that for a certain flow rate we don't get air and water mixed up on the way down (gurgling). The drain on the right is the emergency drain in case the middle one get blocked.

 

[Advanced Aquarist]

I'll discuss about water flow now in gradual stages.

There is hardly anything on this in the main fishkeeping literature, so here goes (I don't have a CAD program so you'll have to make do with my awful drawings).

As I mentioned before in the "sump vs canister filter" thread, syphon boxes and overflows are just examples of "soft forcing". It's basically water following in the path of least resistance.

In 1 we see the familiar overflow barrier that most reef aquariums have. Here we just concentrate on the overflow itself. I assigned the greek letter delta as the "water level difference". In this case the higher the delta the higher the flow. It's not linear though, flow is approximately proportional to delta raise to 1.5.

In 2 we see the standard grate that most systems use. We don't have straight bare barriers because we don't want everything to go over, hence the grate to prevent fish, plants and whatever large object overflowing into the sump. The point here is all that matter is the black area, the length of the slits itself that the water is overflowing in, the so called "barrier length" In 1 it's just the length of the barrier since it's the whole piece, in 2 it's the sum of the horizontal length of the slits since this is what the water is flowing over.

The last point is important, because sometimes for reason we'll go into later, you DON'T want such a large delta (water level difference), because the sump HAS to make this delta. For this reason you see weirs with big slits, or syphon boxes with LONG grated overflow weirs, they want to reduce the need for a large delta - or support higher flow rates. Also because delta determines that area overflowing on (black filled in, in 2) having it low means that there is less force pushing weak fish/animals into the suction.

In the above diagram we see the other side, where the water flows into the chamber, and this is an open hole, as in water is not plugged in, the flow is low so water just seeps on the inside wall of the pipe with air in the middle. This is not generally ideal since it can create noise, but it happens when you have an oversized pipe for the flow rate that you're running.

Note also in the water overflowing into the pipe there is a water level difference created. This delta will play a role in how a syphon box works.

In this diagram we've narrowed the down flow pipe so that air gets expelled and the water level rises. Again a water level difference is create and it's dependent on the flow rate AND the size of the pinched hole.

Why would you want to do this? The main reason is to eliminated water flow noise, when water and air mix on the way down you get a gurgling noise. there is another separate reason to do this which will be mentioned later.

So here we'll talk about the famous water buckets experiment that you can do at home. It basically demonstrates how U tubes work, or as a way to test your own U tubes.

You need two buckets filled with water and the U tube to be tested. Alternatively just connect them using flexible tubing used for canister filters. Make sure that there is no air in the connection like in the diagram. Now get a small pump/powerhead, attach flexible hose to it, and set it up as in the diagram.

Turn the little pump on. You will notice after a while the water level difference (delta) settles for the given flow rate. You can attach a tap onto the output of the pump and vary the flow, for a different flow a different water level difference. You can also try different U tube hose sizes and see how that effects delta.

You maybe asking why you would want to do this. Answer is you wouldn't. But what it models is what happens in a syphon box which I will talk about later.

 

[Advanced Aquarist]

Now is when things get a little complicated . . .

So in 6 we see a more realistic arrangement to illustrate how water level differences work. U tube is plumbed linking the aquarium and the syphon box. We've removed the skimmer cup here to simplify things a little - it's an added complication that we'll get to soon.

Now assume water is flowing at rate F litres per hours (say). Then, like the water buckets experiment, if we had to take our U tube and a water pump of flow rate F we will need water level difference h1 in the diagram. That's the water level difference we need.

BUT since the water had to flow over a barrier INSIDE the syphon box another water level difference is required to get it to overflow at flow rate F also, we call this water level difference h2. (we leave the discussion of the water going through the bulkhead to the sump out of this).

So in all we require not just h1, but h1 AND h2. This is what I call the dual form. And this sum (h1 + h2) means we have to use MORE water from the sump to create it in the tank. Making a sump only big enough to make h1 in the tank won't work, it has to make (h1 + h2) level. Because when the power goes off that's where the water that makes up (h1 + h2) will end - in the sump!!!

This is the complication you have if you use a syphon box like this. This is arrangement is rarely used because of this, but it illustrates a point.

The next diagram is what is generally being used:

So basically we have the skimmer cup here and the tank water overflows into it. Then BETWEEN the skimmer cup and syphon box we have the exact thing happening as explained above. BUT because it is happening INSIDE the skimmer cup we only need to make sure our skimmer cup is deep enough. The sump only has to make h1 water level, BUT the skimmer cup has to make the (h2 + h3) water level, but we've removed the need to make excess water level. But you still have to engineer those values right.

We put enough resistance on the down flow pipe that we get h3 to be high enough to block any air going into the pipe. If we want to reduce the height of h2 we can use a large U tube. As long as the water overflowing from the tank into the skimmer cup DOES NOT cause the water level inside it to reach the top of the cup's barrier we're fine.

But there's another added complication I'll talk about next.

 

[eatyourpeas]

I am anxiously awaiting the PID moment!