Automatic Grow Light


We’ve covered automated plant growing before, but that project might be overkill for many situations. Many of us don’t need our plants to have facial expressions either. Sometimes, we just need a little bit of help. This automated grow light is a nice little project that supplies decent light when necessary. You can download the source files on the tutorial. It is currently set to supply an additional 4 hours of light, detecting the low light levels to turn it on.

22 thoughts on “Automatic Grow Light

  1. If you look at the spot created by a bluray laser it will seem blurry and not well defined because it triggers fluorescence of the vitreous humour inside the eye. Does anyone know of any ill effects this might have ? (similar to UV ?)

  2. Another silly post.

    Oh how has HaD fallen…

    How about some basic vetting of the projects viability? 3 measly LEDs aren’t going to give enough light to grow plants. (and as another poster pointed out, seedlings don’t need light)

    He didn’t do the resistor current limiting correctly, btw. There should be a current limiting resistor per LED.


  3. @8way
    Well, I guess it wouldn’t be a hackaday comment thread without someone complaining about the project and saying hackaday sucks now. Why do you even visit the site, let alone post?!

  4. Looking back at the past few days posts, most are interesting and involve some brains, are unique and actually work. Debian on a router, the wirewrap cpu, avr rss reader, paintball gun for example are all cool projects. Even the proxxon drill press post is a good one b/c drilling pcb boards without breaking bits is a common problem for hackers. These are the reasons why I read HackaDay..

    But stupid posts like this or the infiintely ridiculous one about the batteries wrapped in solar cells, are a waste and lower the credibility of this site.

    Ever since Hack-A-Day went to Hacks-A-Day, there has been the consistent stupid post every few days. I wonder if they hired more staff, and couldn’t find someone with the most basic knowledge of electronics or physics.

  5. @Paul AND 8Way
    Paul was right “seeds” do not need light but “seedlings” do. As soon as the seedling breaks through the ground with its little green stem they need adequate light otherwise they will keep growing rapidly until they do and eventually become weak leggy (dead) plants. By looking at the post the three LEDs look like 1 watt so definitely more powerful than your standard Radio Shack variety and would have some effect as a light supplement but would be lacking as a primary light source.

  6. I always thought plants needed white light , including an ultra-violet component, for growth. This project has red leds, which are probably doing little else but keep the plants warm at night. White leds would probably be a better choice.

  7. Sorry to keep beating this up, but there are a number of issues with this design. :()

    5V thru 50 ohms = .100 mA
    /3 leds = 33mA/LED

    Depending on the LED, that’s too hot and may prematurely burn them out.

    Red leds drop approx 2.1V

    .1/3*2.1=70mW each, or 210mW total. And that’s total power used by the LEDs. Assume a generous 50% efficiency for the LED, that’s about 100mW total radiated light.

    That poor Atmel proc is not going to be happy. Atmels can only source 45mA IIRC. So he’s running that poor drive transistor way hotter than it is designed for. Current is limited to about 45mA due to the small size of the drive transistors. So make that 50mW total radiated light.

    An external drive transistor (any old NPN switching transistor with a gate current limiting resistor will do) is necessary to source that much current. Or, alternatively, he could use multiple, (3 or more, IO pins to drive the leds.

    He should be sinking the current thru the micro, not sourcing — it’s more efficient to sink current thru a FET than to source due to easier electron vs hole mobility.

    A decoupling capacitor across the Vss/Vdd lines of the micro would help. Also a 10uF or so across the input power lines to reduce Vdd dropping when the leds turn on..

  8. @NeilJB
    Most plants only require red/blue light for growth. From NASA research a ratio of about 80% red light gave the best results. Colors between the red/blue spectrums provide very little energy to the plant, green for example is almost entirely reflected back.

    White LEDs are a decent choice but if you look at them under a spectrograph they actually contain a large intensity of blue light but not sufficient red light. For this reason I use white LEDs for “blue” light but have separate red LEDs for “red” light. Given this person is growing in a daylight lit room the plants are probably getting good blue light but lacking the intense red light which normally occurs naturally during sunrise/sunset. If anyone is interested in the more geeky explanation I have written a post on the effects of light spectrum on plants.

  9. @paul

    Many types of seeds do need light to sprout.

    Red/Blue is quite sufficient for many plants energy needs. However, plants also use quality of light to determine how to grow. The frequencies used for growth regulation are not necessarily red/blue.

  10. can someone give me a part number or something for these leds. I have looked on digikey and mouser and there are so many different red leds. I wanted to try this design out because i grow orchids and I wanted to try to learn how to solder.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.