Previously, on the basis of discussions in our local Bonsai club, I set a tree growing site (for winter use) out of an old aquarium. Then I used fluorescent lighting I already had with plans to upgrade it to more powerful light version.

The first test version of powerful lighting is now complete. The aquarium is itself quite ugly and as such is not supposed to be of any aesthetic value but only a place to provide humidity and light amounts for winter tree growing.


The used material/parts are:
  1. 3x 50W LED chip
  2. 3x LED driver
  3. 3x cooler with lens and installation bracket


LED chip

According to video the cheap LED chips are errors in production which I am sure these are also of that kind. Nevertheless, the tests didn’t show so big problems like depicted in the video. Below are some photos from tests with it:


As visible in the photos most of the LED’s inside the chip were ignited. Some more, some less but still all of them. The voltage when they started lighting was just above 23V.
The price was just above 3€. Time will show if this was something endurable or not.

LED driver

The driver is rated at 50W, just like the LED chip. All 3 used in this project were the same, 30W, 1A on 30V. Considering this is the constant current driver the LED chip also consumes 30W so I didn’t test the LED’s under max 50W conditions. The driver is nice and small, priced at 11€, received 3€ refund after complaints that is is not rated 50W.
The project uses 3x LED chips so I considered 3x 30W will be more than enough for a small aquarium.


The cooler is rated at 100W cooling power. It does cool very efficiently. The 8cm 12V fan on the top of it is silent but not any sleeping-room silent. Considering I received the lens (60 degrees) and the installation bracket I am very satisfied with the price. I got it at 9€ a piece.

Energy consumption

3x 30W + 3x ~2W (fans) makes it just under 97W of electric energy 🙂

Light strength

15 cm under the LED lens surface makes it almost all over the space 80-90 klux. This means that most of the trees have sun strength around their middle to upper part. Light strength wise this is very good.

The System

Looks like this:

The assembly is not a beauty nor it was meant to be. I wanted to test first if there is sense in constructing anything like this.
The lamp sets are installed under angle so that the 60 degree angle is not throwing too much light out of the aquarium borders and that it shoots it more to the side of the tree. This way I avoid over lighting (if that is possible?!) of the top of the tree and provide more light to the lower branches. The next in rows light is set to the other side of the tree so that the light is not being present on only one side of the tree. The lamps are not divided from the tree air by any glass, they are inside the same air space. Considering the efficient cooling they don’t heat up the space at all. I had to though tape the parts of the cooler (didn’t bother cooling effect) being inside to stop sucking the moist air from inside thus drying the tree air space.
The top cover is made of acrylic plastic sheet, 4mm I had already. I wanted to have visibility fro above though to be honest, it is unnecessary. There is enough glass around for any kind of inspection.

First experience

Am I satisfied? Yes, not extremely though 🙂
Several mistakes are present:

  • The lamps should have been set more to the edges of the cover so that they are even more to the sides of the tree if I have already thought of doing that kind of setting
  • Air drying is still present to some extent as any kind of hole between the cover and aquarium sides produces draft due to fans above
  • Concentrating light into 1 (or several) strong light sources is not the best idea. It is easy to install, easy to handle it but the tree is not covered as good as it could have been with several lower strength spot lights I use in the living room shelf, for an example. Because of this I doubt I will continue thinking of building any other indoor lighting based on a few string light sources. The wider area, less powerful light source seems to me a better idea. 
All together, it is OK for the test and experiment. The light is strong and time will show how efficient this system is 🙂

I will write updates on the trees growing there to see if/how the trees develop under these conditions.