Saturday, November 24, 2018

Day 7, and then there was light, and heat.

So two things need to be programmed for the aquarium.  The heaters, and the moonlights.  We will start with the heaters, because that's the easier of the two.

I have 3 Finnex 300W heaters.  I located my temperature probe up in the top of the tank, in one of the overflows.  I wanted it pretty far from the heaters, so the readings would be stable, and unaffected by the heater, and anything else down in the sump. I also don't want all the heaters turning on and off constantly, trying to keep the temperature in check and vastly overdoing it.  I'd prefer to not run 900 watts of power all at once, as that gets expensive.

So there was a pretty good thread on the Neptune forums about a multiple heater setup that pretty much did what I wanted.  So I hacked it up, and we have the following:

 Heater1Left

Fallback OFF
If Tmp < RT+0.0 Then ON
Defer 001:15 Then ON
If Tmp > RT+0.1 Then OFF
If Time 16:00 to 03:59 Then OFF
If Tmp < RT+-0.1 Then ON
If Output Heater2On030 = ON Then ON
If Tmp > 80.0 Then OFF
If Tmp < 70.1 Then OFF
If Output LeftReturn = OFF Then OFF
If Output LeakDetected = ON Then OFF
When On > 060:00 Then OFF

Heater2Right

Fallback OFF
If Tmp < RT+0.0 Then ON
Defer 001:15 Then ON
If Tmp > RT+0.1 Then OFF
If Time 04:00 to 15:59 Then OFF
If Tmp < RT+-0.1 Then ON
If Output Heater1On030 = ON Then ON
If Tmp > 80.0 Then OFF
If Tmp < 70.1 Then OFF
If Output RightReturn = OFF Then OFF
If Output LeakDetected = ON Then OFF
When On > 060:00 Then OFF

BkupHeaterR

Fallback OFF
If Tmp < RT+-0.2 Then ON
Defer 003:00 Then ON
If Output RightReturn = OFF Then OFF
If Output LeakDetected = ON Then OFF
If Tmp > 80.0 Then OFF
If Tmp < 70.1 Then OFF
If Tmp > RT+0.1 Then OFF
When On > 045:00 Then OFF

Heater1On030 and Heater2On030

Set OFF
If Output Heater1Left = ON Then ON
Defer 030:00 Then ON

These three heaters have a pretty slick setup IMHO.  The right/left ones are split, so for 12 hours, one is primary, and then for the next 12 hours the other one is primary.  There is a virtual outlet for each, that triggers when the heater has been running for 30 minutes.  When this happens, the second heater kicks in, and helps out, even though it's outside of it's timeframe. When the temp is far too low, we also have the backup heater, so in theory all three can be running at once, but only if things are getting a bit chilly.  Finally the When statement at the bottom of each, is a failsafe.  If any of the main heaters run for an hour straight, they shut off.  But not just off, they flip off Auto in the Apex, so they have to be manually reset.  You can use this to generate an email alarm with the If Error statement.

So now I'm pretty happy with that, lets move on to the moonlight.  In my previous setup, because the Apex is still broken after all these years and won't let you access the moon intensity values, I had to kinda hack up a lunar phase.  In my case, it's simple, I follow the moonrise/set times, but the intensity is on a 1 week schedule.  Low intensity on Sunday, highest on Wed, coming back down to Saturday with no moonlight at all.  However, I don't just want the moon to flicker on into existence, I want it to very slowly fade in and out, like a rise.  This required some very overcomplex programming.  I don't actually recommend anyone do this.  It's slightly insane, and required so many virtual outlets it was absolutely absurd.  It might be possible to tighten this code up a bit too, but anyhow, here it is.  Quick note, I have my moonlight offset by 2 hours, so that is why you see the 120/120 statements.  If you don't offset, you would use slightly different code there.

Outlet: LED_Moon

Fallback OFF
Set OFF
If Moon 120/120 Then ON
If Tmp > 82.0 Then OFF

VDM: Var_LEDMoon

Set OFF
Fallback OFF
If Output VMoon1Rise = ON Then PMoon1_Rise
If Output VMoon1Set = ON Then PMoon1_Set
If Output VMoon1On = ON Then PMoon1_On
If Output VMoon2Rise = ON Then PMoon2_Rise
If Output VMoon2Set = ON Then PMoon2_Set
If Output VMoon2On = ON Then PMoon2_On
If Output VMoon3Rise = ON Then PMoon3_Rise
If Output VMoon3Set = ON Then PMoon3_Set
If Output VMoon3On = ON Then PMoon3_On
If Output VMoon4Rise = ON Then PMoon4_Rise
If Output VMoon4Set = ON Then PMoon4_Set
If Output VMoon4On = ON Then PMoon4_On
If Output LED_Moon = OFF Then OFF
Then you need to setup a bunch of profiles.  By a bunch, I mean 12.  They all follow a similar setup to this:

PMoon1_On
Ramp
1 minutes
7%
7%

PMoon1_Set
Ramp
60 minutes
7%
0%

PMoon1_Rise
Ramp
60 minutes
0%
7%
Then, if that's not enough programming for you, you need 15 virtual outlets..
First, three for the moon state.

VMoonOn

Set OFF
If Moon 180/060 Then ON

VMoonRise

Set OFF
If Moon 120/-360 Then ON
If Output VMoonOn = ON Then OFF

VMoonSet

Set OFF
If Moon 360/120 Then ON

Then 12 more, like this, numbered 1-4:

VMoon1Rise

Set OFF
Fallback OFF
If DoW S------ Then ON
If Output VMoonRise = OFF Then OFF

VMoon1Set

Set OFF
If Output VMoonSet = ON Then ON
If Output VMoon2On = ON Then OFF
If Output VMoon3On = ON Then OFF
If Output VMoon4On = ON Then OFF
If Output VMoon2Set = ON Then OFF
If Output VMoon3Set = ON Then OFF
If Output VMoon4Set = ON Then OFF

VMoon1On

Set OFF
If Output VMoonOn = ON Then ON
If Output VMoon2Rise = ON Then OFF
If Output VMoon3Rise = ON Then OFF
If Output VMoon4Rise = ON Then OFF
If Output VMoon2On = ON Then OFF
If Output VMoon3On = ON Then OFF
If Output VMoon4On = ON Then OFF

VMoon2On (as an example)

Set OFF
If Output VMoonOn = ON Then ON
If Output VMoon1Rise = ON Then OFF
If Output VMoon3Rise = ON Then OFF
If Output VMoon4Rise = ON Then OFF
If Output VMoon1On = ON Then OFF
If Output VMoon3On = ON Then OFF
If Output VMoon4On = ON Then OFF
So how does this nightmare of programming all work, and why do you need so many virtual outlets and complex statements?  Well.. its "easy".  First off, VMoonOn/Set/Rise tell us what state we are in, are we rising, setting, or full-on.  Then we need to figure out if we are on bright today, or dim, or whatever.  Which of the 4 phases of the moon we have.  Thats where the VMoon1On, VMoon2On, etc all come in.

The rise outlet picks the day.  This means if moonrise is before 11:59pm, it will base the day of week on the day the moon rises.  We then have to chain the On to that particular rise.  If we use the day of week statement again, in the case where moonrise was at 11:59pm, we would rise to say 7%, but then it would be Monday, and Monday has a 14% moon, so that would then cause a sudden jump when we get to the On state.  Therefore the VMoon1On statement looks to see which state was on, and then selects which one to chain based on that.

Finally, the Set does the same thing as the On.  It chains off the previous On, so we get the same starting point for the Ramp profile as we had on the On state.

All of this madness gives us a 7 day lunar cycle, where even if the moon rise/set spans across midnight, we still get the proper ramp settings to follow the moon.  And all of this follows the seasonal Moon table, so it varies day to day automatically.  The whole reason you can't use something like the Wizard to do this, is you want the whole thing to shift by the seasonal lunar rise/set, and when the times are longer/shorter you want the right thing to automagically happen.  This does that, at the expense of say, 5 hours of programming, and some sanity you weren't using anyhow.

Do I recommend this?  No.  It's so super fiddly.  It does 100% work though, and maybe staring at it will solve a different problem you have.  I wouldn't even dream of this on a Classic Apex, as it would probably cause the CPU to give up and tell you it hates you.  But it works, and it is super slick to witness.  If only I could do a full 30 day cycle.  Sigh.  Neptune..

Day 7 comes to an end, and I drift off to sleep slightly less sane than on day 6...

Day 6/7, Rat taming by moonlight

The 3d printer worked a bit of overtime during the week, and churned out about 30 different wire management clips and brackets for me.  The new Apex was all wired in, but as is the case with such things, just running wires every which way ends up making a lovely nest for a rodent.

Brackets in hand, I went about tidying up the left side of the tank.  The right side needs alot more love actually, but I still have work to do over there, so I'll leave it for now.

For the curious, the brackets are off thingiverse, (not my designs):
https://www.thingiverse.com/thing:122600
https://www.thingiverse.com/thing:1543201

Having sorted that mess out, I moved on to the dreaded lights.  The lights suffered some serious setbacks between the flood, and the dust ruining the fans.  My original build is actually documented on another aquarium site, but as I rework it, I will continue to document it here. Taking inventory of what I have, it looks like 3 of the 4 main lights have blown LED's, though it's often hard to tell just by looking.  I have ordered some new chips, so those are on the way.  In the meantime, the moonlights looked more or less OK, so I thought, well, lets find out.


So I very carefully checked all the connections.  Turns out the VDM cable I made is full of corrosion.  Ok, well, we can make a new one.  Looking at the old Tunze stream cable, one side is completely corroded, but the other wire looks as good as new.  The RJ45 end is done for though.  Snip, snip, crimp, and I now have an RJ45 -> Tunze Stream and 2 fork connects.  Fork connects hook up to the ballast, Tunze plugs into the little box, and we are off to the races. I carefully verify which VDM outlet is attached to which wire, edit the programing for that VDM to be "Set 15".  Just enough to power it up, but not enough to run them at full blast.  Plug in the wire, flip the outlet on and...


Woo.  I can has moonlight!  This is good.  This saves me so much work not having to reinvent this particular wheel.  The moonlighting was by far the more complex of the two lighting setups, so to be able to just plug it in and go is such a win.   More on that later...

So now that the Tunze's have VDM outlets wired into them, they need some programming, and a bit of love on the wire management side as well.  First, I made a pair of these:


Then I did some simple programming in the Apex for the streams, and got a decent setup going:

A few minutes of testing, and I was very pleased with the result.  Neither side turns off fully, which is good for my size of tank.  Mind you, filling out the giant form on the Apex to do this, and get them right, well that took forever.  Basically I have them run for 45 minutes at full power, then a 45 minute slowdown, and a 45 minute low power (30%).  They are on alternating schedules, so they really move the water around in different ways.  I've probably fiddled alot with the placement of them, but I'm still not 100% pleased with it yet. Time will tell.

One last (HA!) bracket needed to be designed, so back into Tinkercad I went, and spat out this bracket:
Now my probe wires are all nice and tidy.  One thing to note about these.  The Trigger system Ruby sumps come with some very nice premade probe holders.  I was pretty excited about these, and was looking forward to ditching my magnet mounts.  No luck.  They are awful.  They are too low in the sump, so the tops of the probes were underwater.  They don't have little screws, so the probes just kinda dangle in the holes and move around, and finally, they are in the area reserved for the skimmer.  Between the drains and the skimmer, that area is full of bubbles, and this was ruining the readings on all the probes! Nope, back to the magnets for me.

Ok, Tunzes taken care of.  Check.  Moonlights work, check. Wires managed.. mostly check.  Time to do some programming..

Sunday, November 18, 2018

Day 4/5, the fun must go on!

During the week, my replacement Apex and all the bits arrived.  The previous one was the subject of some unwanted corrosion and rust, from poor placement.  I actually disassembled it, and carefully cleaned it with Q-tips and isopropyl alcohol, but not much luck was had.  The unit works, and the ethernet works, but every other port on the device is dead.  I actually have enough extra bits that I could have just gotten it working, but I decided that if it was partially dead, it was on it's way to being fully dead, and nobody needs that at 3am.  I may actually install it on my 125, because that one doesn't need much in the way of control, just measurement and run the LED's.  I wonder if I went PWM or 0-10v on those....


So the Apex and the EB832's were mounted.  I mounted the Apex super high up, to avoid any more splashes of doom.  The Energy Bars were mounted as high as I could, without having to make everything an extension cord.  The modules just kind of went where they made sense more or less.



I also decided to re-route the water line for the RO, that feeds to the ATO reservoir.  While doing so, I decided to change out the DI resin.  Originally, I used some heavy duty drywall anchors to mount the DI canister on the wall.  Turns out, bad idea.  When I tried to turn the body to unscrew it, I could feel the anchors ripping the wall apart.  So down it came, put in a header board, back up it went. 


I also tore some of the return pump plumbing apart, and wired in the FM-100 flow monitors.  I don't actually care how much flow I have in those return pipes.  What I do care about however, is if it is flowing or not.  I want multiple levels of redundancy here, so the plan is to put an optical level sensor on each sump, and the flow meters, so I can tell if one pump stops, slows down significantly, explodes, whatever.  This would cause the level on one of the two sumps to rise a bit, and unbalance them, meaning I need to compensate by slowing down the other pump.


Speaking of unbalanced sumps, I was thinking up failure scenarios, and I realized I had made a boo-boo.  There are two sumps.  Each sump has a return, and a drain feeding into it.  Now in the scenario where one pump stops, and the other runs, the problem is, that the water will come down the drains equally on both sides, but on the running side, it pumps it back up, where half of it goes to the dead side.  This continues until the working pump overflows the non-working sump.

To fix this, I watched a few Youtube videos on drilling acrylic sumps until I had enough courage to take a drill to a $400 sump, and went at it.  Two holes later, near the bottom of the pump housing, I had some bulkheads installed, and ran a pipe between the two sumps.  I then filled the system up, and simulated the failure scenario, and while they didn't remain the same height, the down sump remained well below the top, meaning that my evil plan worked.

Finally, with everything wired up, I set upon the programming a bit.  I've only finished about half of it at this point, and will post it all once I've got it all debugged.  Mostly I put in a bunch of code to deal with pump failure scenarios, because paranoia.

And then, finally, I got to turn the returns on.  Woo, very slowly I ramped both of them up to 100%.  It's interesting how much less power I have than with the old Dolphins.  The drains have absolutely no problem keeping up, I don't even have to tune the siphons on them.  Which admittedly is kind of nice, it just works and I don't have to mess with it.  In the end I'm ok with this,.

While testing, I found another critical error.  I installed a check valve on each pump.  However on one pump, I placed a tee above the check valve that runs to the GFO/Carbon filters.  This meant that when the pump was stopped, it decided to siphon the tank down through the reactors, and then into the sump.  Sigh.  For now I adjusted one of the returns to be right at the top of the water level. This doesn't actually present a problem, as I didn't like where that line was anyhow.  But I don't fully trust it, so.. I'll probably have to special order a tee from FlexPVC and redo that side.

Next was the ATO and Neptune ATK kit.  Ok, easy, install the PMUP, install the little float valve thing, run the wires over to the FMM, plug them in, and, hey, the 1Link cable isn't long enough.  Argh.  Well, guess I need to order an extension cable.  Sigh, now I have to manually run the RO in there to keep the system stable until UPS blesses me again.

I ended up getting alot accomplished this weekend, and got myself some little side projects that need doing now.  For example, I've re-created a rats nest of wiring with the electrics and the Apex.  I also need to install some surge suppressors in the main outlets, to avoid any damage to the expensive bits.  The heaters are in, and are slowing bringing the tank back up to a reasonable temperature.  I hacked together a spare Tunze->Apex cable with the remains of the one from the old system, so I need to wire that in and setup the Tunze's to run.  UPS also blessed me with my MightyMagnet F5, so now I have some incentive to hit the front with a scraper and start being able to view the tank.

Next on my list, is a whole bunch of 3D printed parts.  I need to make a ton of little brackets to hold float switches, manage cords coming out of the sumps, manage cords on the wall, etc etc.  That ought to keep me busy for awhile.  Of course anything I design I'll put up on thingiverse.

Day 5 over, exhausted, I go to bed.

Monday, November 12, 2018

Day 3, The Taming of the Rat's Nest

OK, now we get to the very slightly more exiting bit.  Cable management!  No, well, yes, but I won't actually lead with that.

I did most of the heavy lifting for the sump plumbing yesterday.  All it had to do was dry overnight. My goal today was to fire up some pumps, just to get water moving through a sump, try to tune out the issues I could.  With that in mind, I remembered that for some unknown reason, I disconnected one of the 3/4" return tubes inside the overflow.  Why on earth did I do this?  I think my plan was to stop back siphon by taking them off, but it turned out I couldn't even get the second one to budge, so I just abandoned it.

This, was super problematic today.  I stopped the flow from the durso by flipping it upside down when I shut the sumps down.  So now I was trying to work around that, and get this awful tube back on the fitting.  Vinyl tubing after 10+ years, not so pliable.  It's like trying to jam SCH40 onto a barb fitting, and with the durso in there upside down, taking up most of the weir, it was impossible to even get a grip on the thing.  I stared at it awhile, and then got desperate.  I couldn't replace it, because the connection on the other end is at the bottom of a 36" weir, that is about 12"x8", so I can't even touch the bottom.  (I honestly do not remember the arcane trick I did to get it on in the first place)  So I grabbed an extra-long set of needle-nose pliers.  I put the end inside the hose, and opened them repeatedly, in different directions, to kind of work the tube with the hopes it would soften up without just splitting in half.  Luckily it did open up very slightly, just enough that I could spend the next 30 minutes jiggling it onto the barb fitting.  I am so glad I only ever took one of these off...

Having done that, it was time to fire up the left side.  I took the hacked up tube off, and threw the durso back on.  It was now time to witness the power of this fully operational 2" Durso Standpipe.  Well.. maybe..  Turns out since I shut the RO/DI down in the electrical fire a week or so ago, the tank level dropped a bit.  In what was pretty much one tremendous gulp, it sucked the water level down to the finger cutouts, and was done.  It was a spectacular gulp though.

So I went down to the sump side, lets see what we dumped in there.  A few inches of water, enough to cover the pump, not much else. Nothing left to do put put it back where I found it. I plugged in the COR20, immediately turned the power down pretty much as low as it would go, and got the system into equilibrium.  There is basically a trickle of water moving up and down now, definitely not enough to even look at the other sump funny.

OK, well, lets try the skimmer then.  I'm sure this water needs some love.  Plug it in, turn the pump down, hey, what's that noise?  Well the pond liner underside of the aquarium came in handy instantly.  I had the skimmer set to min, which shoved all the water up the pipe, and immediately overflowed the cup, and started gushing around.  Bonus points for the pond liner catching all of it and keeping my floor dry!  Fixed the settings on the skimmer, set it super low, just to let it do it's thing, and things look safe now.

So one problem with my DIY LED's, is that I tried to do a good job of managing the mess of cables, but it wasn't really that great.  I relied on those little stick-on cable tie mounts, which held for probably all of 15 minutes.  So I had some gnarly clumps of wire in what looked like a rat's nest in a hanging garden.  I knew one of the things on my long term plan was to go passive cooling on the LED's, so I don't need those fan wires anymore.  So very carefully, I measured and coiled the wires, removed the fan wires, and installed some 3D-printed screw-in ziptie holders on the wall.

That picture makes it look alot worse than it really is.  Now it's all up high, secured tightly, and out of the way.  From the back it looks a thousand times better.

In the great flood, one of the only things that survived was the LED setup, which I carefully bolted to a chunk of MDF on the wall next to the tank. I was very careful about setting this up, and I think it came out pretty clean.  I decided this would be my map for the rest of the electrical.  I got some 2'x2' plywood sheets, painted them (just to kinda look nice-ish, with whatever random samples I had lying about), and secured them to the wall.  I then went around and started bolting up everything that was ready to be connected, routing some of the wires, etc.  I had printed some nice mounts for my power bricks.  I did not want them sitting on the floor, or some shelf where they would inevitably be in a pool of water.  Bolted it all up, and huzzah!


Still much to add here, waiting for parts to come in the mail.  Everything is plugged into the wall socket while I wait for the new Apex to arrive and solve all of my problems via magic.

Finally, after hours of scrubbing from my loving wife, the TLF reactors were all clean and ready to be installed.  In last night's episode, we left you with the cliffhanger that we would try stretching out the vinyl tubing to see if we could make it workable overnight.  Turns out, yes, we did.  Unclamped it, and it was pretty much straight, almost no coil to it at all.  This made it super simple to cut to length, and re-plumb the media reactors.  A few minutes later and they were done.

Day 3 comes to a close, and I have some RO/DI dripping into a bucket to make saltwater.  I'm not really exhausted tonight, so that's a win.  I go to sleep, dreaming of a UPS van filled with more parts.

Sunday, November 11, 2018

Day 2, PVC glue is created from joy.

Time to plumb up the sumps and returns.  The tank has massive 1.5" drains.  In the old plastic tote arrangement, these were run full on.  I have to say, the sound the durso pipes made was kind of thrilling..  Kind of like a wind storm blowing through the trees and whistling slightly.

I couldn't find any sumps premade that had drain inputs that large, and honestly, I don't care so much.  When I originally designed this, it was prior to the existence of impeller style powerheads.  That meant I was relying upon the return pumps for alot of my water movement.  However, not too long before the day of doom, I had purchased two Tunze 6255's.  Wow did those move water.  The long and short of this is, I don't care so much about the speed of the drains and the returns.  It just needs to pass some water down there so the filters get a crack at it once in awhile.  Both of the Ruby sumps have dual 1" inlets.  So I went online and found some nice Y adapters, and went to work.

On one of them, I branched off a third output.  I'm not sure how well this will work in practice, but the Ruby Elite has a third input for the refugium.  I'm guessing that the raw downforce of 2" piping will force some water that way.

I also got the Neptune COR20 piped up to the existing returns on the left side (Elite), so that is just about ready for a test fire.  As I write this, the glue is drying on the right side, so firing up the sumps will wait until morning.




Now for the lessons learned portion of our show:
  1. If you ever buy PVC unions for your tank, buy twice as many as you need.  Why do you buy unions?  It's because maybe in the future you want to disconnect that thing, and replace it with something else.  However, 5-10 years later, what are the chances that the company that made your union is still in business?  Now you have a union stuck on something, that you cannot re-use, and you end up having to saw the whole joint off anyhow.
  2. A Japanese saw, is quite possibly the best tool there is for sawing off random PVC bits.  Seriously, made things so much easier than a hacksaw or whatever else.
  3. PVC pipe nipples (threaded on both ends, the SCH80 stuff) are great to saw in half and convert slip unions into male thread ends.
  4. Never ever use the type of ball valve shown above.  The kind you get from hardware stores.  The saltwater gets into the ball, and makes it impossible to turn.  I have a few of these with shattered handles.  The SCH80 gate valves are great though.  Love those.
Finally, I had left my pair of Tunze's in a bucket of vinegar overnight, and went to work cleaning them off.  All the ick came right off, and everything is happy.  Thank goodness these survived all the electrical stupid.  My two TLF reactors are now sitting in vinegar, so I should be able to get them plumbed back in the morning as well.

I ended the day with an experiment.  I got two rolls of braided vinyl tubing.  That stuff is miserable to work with.  It comes in these tightly wound coils, and just wants to stay that way no matter what.  I really didn't want to spend 2 hours fighting it to get it into the tank, so I clamped one end down in my vise, and clamped the other end to a tablesaw across the room.  I'll leave these overnight with the hope that in the morning, they are less awful.


Tomorrow the plan is to test out the drains and the sumps a bit.  Maybe leave them running, not sure.  Remains to be seen how well they work.  Might start plumbing up the ATO, at least mechanically, Probably won't actually use it yet, still waiting for the replacement Apex to arrive to do all the electrical work.

I've been at it since 4am.  Exhausted, I go to bed.

Day 1, build progress

OK, everything is out, time to start putting things back in.

On the right side, the spacing on the vertical supports was a little too close, so one of them had to be moved. This is relatively easy.
  1. Cut a new board of the exact length, hammer it into place right next to the other one.
  2. Bi-metal blade on a reciprocating saw cuts out the old one.
One thing I've learned over the years, is that saltwater is constructed of pure evil.  It will spray a little, drip a little, creep all over the place, find some tiny hole and be annoying, etc etc.  It is not possible to have a perfect floor around it.  Even if all my plumbing was perfect, I'm going to fumble a bit of algae, or slip a fitting around and dump some on the floor.  I just need to assume this upfront and plan for it.

In this vein, the insides of the stand need to be protected.  Every time a drop of water hit the wood of the stand (no, do not tell me it should have been metal, that time has passed), I had a mini heart-attack.  So first I cut some cheap carpet, and lined the floor of the stand.  This is to give a little insulation between the concrete and the bits, and even out any rough spots before I place an acrylic sump on it.
 Next, I bought some pond liner.  The idea here is not to survive a flood.  The idea is that if a gallon drops into the area, it stays there.  No more seeping under the bottom, creeping up the drywall and making me cry.  I cut a big section of the liner out, and made a small tray with it in each section, with walls about 3" tall.  Mostly the idea here is just catch the drips and sprays, keep it contained.

OK, time to start loading in the sumps!

Argh.  We already have a new lesson.  When you have a stand like this, don't measure the first section, and assume the other three are the same size.  Turns out the right section, was 35 3/4" wide, not 36 1/8" like the left, which I carefully measured.  So the Ruby 36 instead has to be turned sideways in the sump area.  That's ok, I'm rolling with problems today.  I turned it 90', cut a chunk of the side wall off the stand so I can get to it from the side, and oh look, now I have a nice area to place that ATO reservoir now.  Win?  eh..

Not pictured, but I also managed to jam the Ruby Elite into the other side.  That one was a bit hairy.  It's one thing to measure that a 36" wide object will fit into a 36" wide spot, through a 15" opening.  Sure, it fits, but one forgets it has to rotate in there somehow once you load it in perpendicular.  2 hours later, somehow.. I managed.

Day 1 over.  Exhausted, going to bed.

Saturday, November 10, 2018

A fresh start

So here we go.  Starting over.

First thing, if I can buy it, I'm not building it.  Two reasons for that.
  1.  I know half the reason it's sat there, is that I don't want to go through the massive effort of building two huge sumps by hand from plywood, and I know I'll never get around to actually doing it.
  2. For some of these things, if I cut a corner here and there, it has long term consequences.  I need to just suck it up and spend the $$$.
So I bought some new equipment. First, a pair of Trigger Systems Ruby sumps.  One 36, and one 36 Elite.  Why two different ones? Simple, one is left drained, one is right drained.  Perfect.

Next, a Reef Octopus 200INT Elite skimmer.  Something to note on this.  The dimensions on the Ruby 36 indicate that you could fit this in there.  That is a bold faced lie.  It will not fit.  It does however fit in the Elite, so, phew.

Next on the list, new return pumps.  These will sit inside the sump, so when they inevitably leak.  They won't leak on the floor and ruin the whole planet.  I went with a pair of Neptune COR20's, because I love controllable, and I love Neptune.
Finally, a Neptune ATK, and level sensor.  I will not make the same mistake twice with the auto topoff and the RO/DI.  Don't ever do what I did.

So this is the start.. More as I progress.

What went wrong?

So what went wrong with the big aquarium? Well a number of things.

First, my attempts to save money with DIY ended up costing me a fortune.  DIY things are great sometimes, and sometimes there is no other option.  But if you rely on certain things to keep working, like computer fans, and they don't, then you have a failure.

Second, the size of the aquarium caused me to make some foolish decisions with the routing of electricity.  Extension cords, placing the Apex EB8's inside the frame of the stand, etc.  If water drips on the interface between plug and extension cord, you get a fire. If that interface is anywhere under or alongside the tank, water will drip on it.  Never ever do this.  The EB8 looked like it was in a good spot.  But funny thing about saltwater, it climbs cords. This meant, that little bits of salt and water went up the cord, into the aquabus, and rusted it all out.  Even better, I was dumb enough to put the apex brain unit down there too.  Guess what else is a brick?

So it was time to tear it all out.  All of it.  Every last bit, everything has to go.


I tore out all the plumbing.  All the tubs, all the bins, tore out all the electrical, all the apex bits, everything.

Now I have a nice bare room, plenty of space to start over.  I guess you could say the near electrical fire was the spark that caused me to get moving on this again..

So this is the starting point.  It all gets better from here, hopefully...








The disaster that is my in-wall aquarium

I have a large aquarium.  When I say large, I mean, mistakes were made large.  800 Gallons, saltwater, reef.  It took me a very long time to get this thing running, and it was looking fairly decent.

It wasn't perfect, but it started to look good, and I was relatively happy with it.

There were however, some demons lurking in there...


One such monster was the plumbing, and the sumps.  I went with a DIY attitude, and was trying to save money, because as it turns out, setting up an 800g aquarium is way more expensive than 6x a 120g aquarium...

The tubs would overflow occasionally, in spite of having 2 2" pipes connecting them.  The pump, a big dolphin 3000, started rusting and leaking.  Yay.




And then we have the lights.  Actually, I'm pretty proud of the lights, right up to the heatsink. A big vertical computer heatsink can absolutely hold down the temperature of a 100w COB/SMD LED run at 140 watts of power.  No problem.  Right until the fan dies. Then poof, instantly.

That's fixable though.  What happened next, was not.

A little tiny solenoid, with a murloc push fitting, shattered.  This push fitting, was what held back the mains water to my RO/DI unit.
This bad design resulted in a massive, and I mean massive flood of my house.  The resulting damage destroyed tons of equipment, drywall, electrical stuff, etc. Many things did not survive this, including my will to keep working on the whole aquarium at all.
So it sat idle. For a long time.  I kept the machinery running, but kind of ignored it.  Later this bit me, as alot of the random things I plumbed in to keep everything alive, were done with extension cords and bits of stupid.  Then one day, a few weeks ago, a big zap sound was heard in the room.  Luckily I was home, and found a bit of water dripping on an extension code, and a pump cord that was obviously burnt to a crisp.

Well that took out the last of my support equipment.  Now I had a real problem on my hands.  A giant aquarium, all pumps down, nothing works, fiasco everywhere, and it's a fire hazard to boot...