Sneak Peak

Happy holidays to everyone. I finally had time to take some photos of new nano tank in progress. Here is the rock work I did few days ago. I have been using Real Reef Rock 4.0 (normal and shelf versions, about 15 kg/33lbs). The Real Reef Rock 4.0 is amazing to compared my other tank Fiji Live rock. I will use it a again and it was cheaper than real live rock.

I also wanted to aquascape some shadow areas for critters. I think now the corals have plenty of space to grow upwards. Of course this will take lots of time:). More photos to follow…




Building the new WIFI & IOS controlled led light

For the new nano reef tank I decided to build new led light. The led light idea is based what I build for Fluval Edge diy led but instead of using led chips on star I ordered few Nanobox reef V3 arrays. I ordered two since I similar led chips on stars as well.

First I had to find good heatsink for the led. After spending few hour searching from ebay and other sites I ordered one from The heatsink dimension were 600mm x 140mm x 20mm. Fluval Edge heatsink was only 160mm x 90mm so there is plenty more cooling capacity in this heatsink. After a while I got an idea how to make enclosure to heatsink. I visited in painting/photo framing company and ordered empty frame with exact heatsink dimensions. I also ordered few plexiglass (white and transparent) with frame dimensions.

After ordering the enclosure items I started drawing the actual leds. I had chosen to make new tank aquascape with “two island” approach. That approach helped me with Nanobox Arrays placement. I wanted some shadows in the tank as well so that is why I have only 3 light sources with 4 channels of controllability.


Initial design of actual leds and spare holes.

After a drilling and tapping the heatsink I started the building process.



Nanobox reef V3 arrays are very nice. 


Arctic silver Ceramique 2 for heat trasfer between leds and heatsink.


Stainless steel bolts with nylon washers is used to secure leds on the heatsink


Led chips on star screwed in.

After some hours every led was in place and the heatsink was ready for soldering and wiring. The problem with soldering leds already attached to heatsink is the amount of heat needed to solder wires. So I cranked up my soldering station near to its limit and started soldering. I double checked every wire to ensure good quality attachment with the leds.


Wiring 4 channels of leds on star.

I also used wire covering tube much as I could.


Wires are covered with nylon tubes.

The Plexiglasses needed some drilling so I made drilling template. Next time I will not try to make holes by myself and order plexiglass from laser cut house.


Top plexiglass drilling template

The build uses 120mm cooling fan. I bought  Coolink SWiF2-120 fan and screwed it on the white plexiglass with metallic grill.


Cooling swift 120mm fan with fan speed control.


Rubber spacers were used to minimize vibration noise.

After some hours of soldering everything was in place. I used a bit modified version of my own electric imp based controller for driving leds. The led system uses 48v meanwell dc converter and only 4 channels is used (controller pcb have option for 5 drivers, but I will use it to different things in the future). I also placed one thermistor in the heatsink to measure enclosure temperature.


My diy electric imp based led controller/driver board.

The electric imp based controller needed little bit code changes as well the IOS app I have written to control whole thing. After few hours of tweaking the code the system was ready for testing.


Diy led enclosure without wires and hanging kit.



Diy led with WIFI and IOS app, first tests.

The led is very bright. I have to measure PAR’s in the future, but I think it will be enough for growing some nice corals. Light is quite even and I did not notice disco effect at all. Nanobox V3 arrays are awesome. Led chips on 20mm stars working quite nicely between these arrays and there is no noticeable difference between the coloration of the light.

Led light is secured by light hanging kit I got from a friend and the leds position can be changed easily.

After a few days of running led I can say the overall build is very good and the temperatures keep around 40c / 104f. The enclosure fan is dead silent.

Only thing I might replace is that top plexiglass with laser cut one to make light even more prettier. Also some wire covering need to be added since I run out of it when I was routing DC power to led controller.

And maybe some day I will replace leds on star with another Nanobox reef array or two (they are awesome and easy to install).


Modding the Tunze nano ATO

For the new Nano build I have Tunze Osmolator Nano 3152. It have only one float switch (its bigger brother (Tunze Osmolator Universal 3155) have float switch and optical sensor). I wanted my ATO have little bit security and redundancy. So I ordered typical float switch from fleabay and tried to wire two float switch in series. Problem with Tunze nano is that float switch operates in reverse. So when the water level is down the float switch shuts down and when the water level is correct the float switch is on. This is very weird decision by Tunze. I do not get this design because if the float switch wire is broken the water pump starts (and can cause flood in bad case). The ATO has 3 minutes safe guard which helps to prevent this problem but in small nanos and picos even 3 minutes running can be catastrophic situation. 


If the float switch wires are broken the pump turns on. Now it has little bit more security (added second sensor wires on top connector)

First thing to do was found out how to reverse operation of normal float switch.


Normal float switch. It is on when the floating device is down and it is off when the floating device is up.

This can be easily done by just rotating the float device upside down (not the float switch but the magnet part of the switch). Then it works like Tunzes one.


Normal ebay float switch taken a part. The float device direction changes switch operation direction as well.

So the case of wiring two float switch in series is out of the question, because then it would not work as needed. It would then be on if one of the float switches get stuck which I tried to avoid in first place. Answer is to wire them in parallel. When float switches are wired in parallel the pump is only running when both switches are in down position. If one of the switches stuck down the pump will close down when other float switch get up.

I also noticed that it is easy add another (or more) sensors just adding little plastic stick to original float switch holder.


After some soldering and testing the system is working again but now with some redundancy. Time will tell if this any good modification. But it still have the problem if the wire brokes down. I have not yet solved that part of the problem, but now there has to be two wires to be broken down before the pump starts if the float switches are in up position.