My aquarium notes

Started 24Jul2018 – updated 18May2019 (HTTPS URLs)

This page is in group Hobby and will become a blog note about my little aquarium. It has been a year since I introduced fishes to it, so it’s about time.


Fig. 1

Figure 1. Aquarium with LED light and undertank heating, with controller box. Øyvind Teig, July 2018 (Press for fullpixel)

Camera: Panasonic Lumix DMC-TZ100. The main picture above is shot on a tripod 2.4 m away under rather dark conditions since the aquarium is behind a carpet to try to remove reflections. Camera zoomed: lens focal length 34.6, f 5.1, t 1/25, ISO 1600. Then both pictures have been set up in Apple Pages and exported as PDF. Then converted to JPG and scaled to 900 and 4096 (press picture). Most of the recent pictures in my blogs have been made like this, even the picture here.

Observe that I am an electrical and computer engineer plus hobby maker, not a biologist. But I love aquariums for the beauty and tranquility of them. I haven’t forgotten the fishes in all the tech-speak below.

  • The tank is custom built of 6 mm “optiwhite” glass, 50 x 27 cm and 32 cm height. Topping it with water it would contain 39.5 litres. I ordered it at TAM Zoo (at Lade) here in Trondheim, and they ordered it from Imazo in Sweden. I think it’s produced in Poland
  • Volume taken up by not topping the tank, stones, sand and filter I have defined such that I have “30 litres” of water in the tank
  • The top enclosure and bottom frame is made from PVC sheet plastic. Both are lose, for the top this is really necessary; for the bottom it’s to make drying up spilt water easier, and for inspection of the aluminium heating frame underneath
  • The aquarium has undertank heating with 24 V DC max 2*24 W (with resistance wire about 2*14 m length for even temperature layout), controlled with max 40 deg C in the heating compartment. There is an overtemperature melting fuse of 84 deg C
  • The water temperature is measured on the outside of the aquarium. Yes it is possible. I also did this with my previous aquarium. The thermal conductivity of water and glass are, as a pair, quite different from that of air. This means that the air won’t be able to stop the water’s temperature from appearing on the outside of the glass. Besides, the tiny temperature fall would be almost constant over the regulator’s range. The temperature I2C sensor is pressed onto the glass and connected to it with a thermally conducting paste, inside a glued-on POM plastic enclosure, also for insulation from the air. Temperature is regulated to 25.0 degC
  • The temperature regulator, in addition to the heating compartment’s temperature and the water temperature, also measures the ambient temperature. I will later fully describe the regulator algorithm
  • Controller. The controller box contains an XMOS startKIT, coded in XC with the xTIMEcomposer tool. It also contains a high accuracy clock with battery backup. There are three buttons and a tiny display for readout and setting of parameters, some of which are permanently stored in internal FRAM memory. All cables are pluggable. See picture of the box here, plus all my blogs about the technicalities here. I also have a USB watchdog and two relay outputs, see My USB watchdog (and relay output) box. I plan to use one of the outputs for an internal water/filter pump (more in May-June/2019)
  • Cooling the tank would have been nice for a few days. But the temp seems to have maxed on some 28.2 degC so far. We don’t have an air condition, as that isn’t much needed in private homes in Trondheim
  • Lighting, se chapter Lighting
  • The LED strips are all in a frame with air passing through for cooling, air slightly sucked in in the back and out in holes in the PVC top. This compartment is between the PVC top frame and the aquarium’s top glass, sealed from water or moisture. Holding my hand on the top I don’t feel any heat
  • There are three DC power sources: 12V for LEDs, 24V for heat and 5V mobile charger with micro USB for the controller box
  • I have indeed been able to squeeze in a radio board and software to the controller box to export (to the web?). The data is now picked up by a client running on another XMOS board (here) with radio board plugged in. First install by the aquarium proper was 6Oct2018
  • There is an under gravel filter, vacuumed from a tube of raising air
  • The air pump is in the basement below. No motor or pump noise
  • For the fishes I only have some twelve cardinal tetra plus two armored catfish and a nice little shrimp. They seem to be quite happy. (I hadn’t forgotten, I said so)
  • Plus, too much common bladder snails and a single clea helena snail (more than one does not work, sorry. They mate all the time)
  • This is my fourth aquarium. The sand and some of the stones have been reused from my previous aquarium that was put away some 25 years ago. We don’t waste things here
  • A year’s experience with nutrition of fish and plants, aquarium salt, aquarium GH salt, tap water is quite ok, filtering and air, plus LED colour temperature and photosynthesis. Plus life and death..

Next chapters

I should have time for this in the autumn of 2018. I will, however fill some odd chapters here, more like a scratchpad, before I encounter on the more structured note:

The Manual

Figure 2 – Control box manual (in Norwegian, press to read PDF (15 MB))

It might be of some value to download this even if you don’t understand a word Norwegian.. (Press picture to read PDF)

Aside: Page by page source is in Pages, then exported to PDF. Then from Preview I copy/paste the PDF of each page into another Pages document with four pages (in one 2*2 table per page), each small page of size 119*87 mm if you print on A4 paper


The regulator does 25 degC. This is probably OK for all. The living creatures I have in the tank, the regulator’s range (we do have a night temperature lowering of four degrees (only electrical heating, so that’s possible), and how hot it may be in our house.

No cooling

At the time of writing (late in July 2018) we have had 34 degC outside in Trondheim, and with all windows closed (we don’t have heat pump that could be reversed for air cooling) I have seen max 27.2 degC in the tank (worse, see Max temp read, below). When I was going to do the weekly partial water renewal of 30/3 = 10 litres of water I inserted 26.5 degC into the 27.2 degC and I could see that all the cardinal tetras were swimming like they were playing into the current from the incoming, cooler water. I had not seen that behaviour before, when I normally insert 25.0 degC into 25.0 degC. I infer from this that 27.2 degC is too hot for them, since they seemed to rather enjoy 26.5 degC – even if they didn’t look bad in the 27.2 degC. But is this correct? But I wonder if I correctly can read them like this, because:

Max temp read: 28.8 degC in the tank 31Jul2018! Terrible. But the cardinal tetras seem fine! According to Wikipedia’s Cardinal tetra article (below):

They prefer warmer water temperatures [above 24 °C (75 °F) or warmer]… The preferred temperature range of the fish is 26 to 28 °C (79 to 82 °F). However, if necessary they will live at 24 °C (75 °F)

Maybe I should raise the normal temperature?



The aquarium has lighting with a total of 15.25W from mixed types of LED strips with 171 LEDs and more than 1210 lm (coloured strips unknown light power (lumen, lm)).
– Since different photosynthesis regions basically are a matter wavelengths (below), I should probably be around blue (450 nm) and red (660 nm) for good “PAR photosynthesis” (Photosynthetically Active Radiation, see Wikipedia below) and gpn mag [1]). Or as often seen, mostly in the 650 nm to 680 nm red range.
– So I must only hope that in this jungle of mostly colour temperature in Kelvin (that’s what they give me (plus CRI, see below)) that some of the wavelengths that are required would appear. More theory in this note later on.
– So far I have been in little good luck (update 11Nov2018: after 3 weeks of this light regime and new plants and fertilising better plus reduction from 14 to 10 hours light days – it looks much more optimistic. Earlier, even 3 weeks have shown degradation easily). You can find all my trials documented here. But this is my present try (on the top in the document):

  1. Channel 1: 3.25W total (from three strips) coloured with
    1. 18 LEDs blue (465 nm). This shouldn’t be to bad a wavelength
    2. 15 LEDs green (565 nm). See [5] about green light and photosynthesis
    3. 18 LEDs red (625 nm). This amy be too low a wavelength. But according to [4] it doesn’t look too bad. Also, in the PAR Wikipedia article (below) “The YPF curve indicates that orange and red photons between 600 and 630 nm can result in 20 to 30% more photosynthesis”
  2. However depending on the Colour Rendering Index (CRI) and PAR amount of the white LEDs they should contain, more or less, light from the full spectrum, including blue, green and red (of some frequency) (I hope..):
    1. Channel 2: 3.4W (from one double-strip) white of 4200K with 48 LEDs 440 lm
    2. Channel 3: 8.6W from these two:
      1. 3.6W total (from two strips) of white of 6000K total of 48 LEDs 390 lm
      2. 5.0W (from one strip) of warm white 3000K with 24 LED 380 lm

LEDs 1, 2.1 and 2.2 are “high intensity” 100000 hours from Inspired LED (but 25000 hours to 90%? I don’t know). I have troubled with three other LED combinations, none with explicit red and green, and one with probably too white light. Therefore I added LEDs point 2.4, which are on a strip from North light (Clas Ohlson). The above LEDS with 2.3 instead being a 1.4 W 6000K probably was too white (even with the colours), nothing but algae grew! The present combination is new on 19Oct2018, so I’ll keep you posted. (Fig.1 above shows the aquarium before introduction of the coloured strips and with less white intensity.) I know some say that green is for colouring only, but my web research shows that green is in fact also needed (here). Besides, it needs to be there to make the sum white-ish when blue and red are present. I am now testing plant types more specifically, to see if I have any luck (combined with nutrition and light intensity). Here is some comfort as to why LEDs are difficult, see [2].

  • LEDs are controlled with day/night with gentle morning rise and sun-dawn intensity control that each last for 30 minutes. Light changes at 0, 10, 20 and 30 minutes and the steps are soft
  • There also is a random cloud simulation effect with soft light level shifting.  It basically takes the light down in the average every second hour for 3-10 minutes to a random entry in an intensity list. This may be switched off
  • Also, the length of the day may be set to 14, 12, 10 or even 8 hours. I use 10 hours light on at the moment

Control of my LEDs

As seen above I have three light channels that I control with three SW-driven PWM (pulse-width modulation) outputs. They basically do 1/3, 2/3 and 3/3 individually for each channel, with soft light changes that lasts for about seven seconds for every step. When 3/3 a channel is all on 100% of the time, not even a short pulse left. Code is written directly in xC. PWM frequency is 222 Hz (1500 ms period) and FETs would pull the LEDs that have a common 12V. This flickering is invisible for me, by a large margin (I could see 55 Hz),  and should be ok for most biological life [6]. EMI noise is avoided by limiting the rise and fall times of each power FET transistor’s output pull.

Aside: I see that commercial LED lights typically are controlled by PWM, 0-10V signal or the DALI protocol. See Wikipedia (below). I had no criterion to do anything standard, but it’s nice to see PWM mentioned.

What nanometer to hit which photosynthesis?

Excite Lighting GmbH has quite some interesting notes about LED and photosynthesis. I think the company mainly is in the plant-above-water business, but I am not certain how much this matters. However I did read “light spectrum which applies to all plant species and in all growth cycles” on a product page, so I guess that might include aquarium plants as well.

I have tried to analyse a figure and the accompanying text present in one of the Excite notes, BLUE AND RED LIGHT IN PHOTOSYNTHESIS [3]. I think I found some discrepancies. (The source of the curve is not given, but the PAR Wikipedia article (below) naturally also shows this phenomenon.) I also added a light spectrum from Wikipedia. Press to read as PDF:

Figure 3 – Analysis by me of a figure and text by Excite (press to read PDF)

The original wavelength axis is quite confusing since it has a 30 nm/part and the numbers some times won’t match, so it is difficult to read and compare it to the text. Therefore I made each 10 nm also lined. I think that according to the figure and not the text:

  • Chlorophyll a (Wikipedia)
    • violet blue to blue peaks at 420 433 nm
    • orange to far-red peaks at 680 663 nm
  • Chlorophyll b (Wikipedia)
    • violet blue to green-blue peaks at 480 457 nm
    • yellow to red peaks at 630 650 nm
  • Beta carotene (vitamin A, don’t know if my aquarium plant needs this)
    • Peaks at 458 nm and 480 nm (figures not given in the text)

Disclaimer: this is all based on me hitting right with the positioning of the 700 nm point.

There is a figure also at Wikipedia (here) that is even less accurate than the above.

Where this takes me in the search for simple(r) LEDs for my aquarium is difficult to conclude.

However, just by coincidence I found a four year old paper (2012) [4] that shows that LEDs are indeed possible. Excite also shows this, I guess. But “good” LEDs come at some cost. And is this for my thin light fixture that I guess could allow at max 20W?

Water filtering

I use a bottom gravel filter, with air bubbling up through a tube to suck water from below the sand, ie.e through the sand and back into the aquarium. I guess this has become rather biological by now. I stir the sand a little when I change water so that there should not be built-up of gases in the sand.

  • At first I added a small particle filter in that vertical tube because I had once had this solution in my previous aquarium. Therefore I would get less under-pressure below the gravel for the effect of the little mechanical filter further up
  • I still was not satisfied with the clearness of the water, so I bought this small pump mentioned above, to use once a week. This helped
  • However, it was when the incoming air’s tube slipped off my design that I for some days added a bubble stone instead of the long tube. Within a day or two the water was clear as water!
  • I then decided to add a new mechanical filter that I put inside a small stainless (heavy) steel tube. Bottom and upwards: water being sucked into the swamp filter inside the steel tube, 4 cm. of swamp, then bubble stone inside and above the swamp and then the 12-15 cm long tube containing up-going, small bubbles. Now again the water is clear as water and the gravel filter should also work. I’ll make a drawing some day


Wikipedia: Cardinal tetra, DALI protocol (Digital Addressable Lighting Interface)PAR (Photosynthetically active radiation), PWM (Pulse-width modulation)

  1. Possibilities of LED Lighting by Johann Buck (2011), see
  2. Wikipedia: Grow light, LEDs: LEDs (Light Emitting Diodes), see here
  3. BLUE AND RED LIGHT IN PHOTOSYNTHESIS (15 Jul 2016) by excite Lighting GmbH, see
  4. Regulation of the growth and photosynthesis of cherry tomato seedlings by different light irradiations of light emitting diodes (LED) by Liu Xiaoying, Guo Shirong, Chang Taotao, Xu Zhigang and Takafumi Tezuka. See
  5. THE (IMPORTANT) ROLE OF GREEN LIGHT IN PHOTOSYNTHESIS (Dec2016) by excite Lightning GmbH, see
  6. Potential Biological and Ecological Effects of Flickering Artificial Light at PloS One by Inger, Bennie,Davies and Gaston, see

Leave a Reply

Your email address will not be published. Required fields are marked *


* Copy This Password *

* Type Or Paste Password Here *

27,060 Spam Comments Blocked so far by Spam Free Wordpress

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>