Connect with us

LED Aquarium hood, power but no light.

Discussion in 'Troubleshooting and Repair' started by H3LLRA1Z3R3, Jan 9, 2018.

Scroll to continue with content
  1. H3LLRA1Z3R3

    H3LLRA1Z3R3

    4
    0
    Jan 9, 2018
    Was given an LED lighthood to use if it could be fixed, trying to understand why it isn't working.
    I'm a complete novice so hopefully that's okay here?

    Lighhthood has two LED drivers in the power supply and runs two channels on the actual hood. i.e. 1 channel per driver.

    I've tested the channel outputs at the power supply and at the led hood. They both showed voltage on both tests.
    Only one channel is working, all lights work for that channel and voltage is the same across all tested LEDs for that channel.

    The Broken channel gives a voltage as the wires connect to the hood but the first light shows no voltage whatsoever.
    I swapped the channels out and same problem.

    There are 4 leds that don't light up when using the multimeter so I thought they were dead and this was the cause like an old xmas tree lights set up but I went over each led for that channel and they all seem to be connected to the previous one although the wiring seems odd and mapped to two leds then there seems to be a positive to negative from the second to third LED and then the next two are positive to positive and negative to negative. Even when the broken leds were skipped and the previous and next led were tested they gave a reading back using the continuity settings so there appears to be no break in the circuit .

    The LEDS are built into the board and then the board is covered so it looks like there is no easy way to remove the diodes, the entire board of LEDS seems to be the PCB with the LEDs built into it.

    Reading at the PS was CH1: 132.4 V and CH2: 133 V, reading at the board was Ch1: 112.6 V and CH2 133 V using the AC volt setting /200 And the working LEDs are showing 6.5 V.

    PS says total 60WX3W.

    Each LED has a little +/- space that I could wire up to bypass the broken LEDs but I am unsure of that will do anything considering there is a circuit there already.

    Attached a terrible wiring diagram and photos of the board can be shown if need be, at this stage I am keen to get some advice on how to proceed.

    Anything would be helpful.

    Thanks in advance

    O/

    [​IMG]
     
  2. Harald Kapp

    Harald Kapp Moderator Moderator

    11,316
    2,593
    Nov 17, 2011
    I agree, this is really terrible ;)
    At a first glance it looks like + and - are short circuited - which is obviously nonsense. So I assume by the black dots you mean to designate the LEDs and by + and - the anode and cathode, right? This means that there are pairs of tweo LEDs in parallel (red circles) and the pairs are in series (blue arcs):
    upload_2018-1-9_12-24-39.png
    Ideally you'd replace the faulty LEDs by similar ones. Alternatively you could try and bypass the LEDs with two standard diodes (e.g. !N4001) in series (green). Two diodes to approximate the pass voltage of the LEDs.
     

    Attached Files:

    Tha fios agaibh likes this.
  3. H3LLRA1Z3R3

    H3LLRA1Z3R3

    4
    0
    Jan 9, 2018
    I did say complete Novice :D

    Yes that's what I am trying to represent. the red LEDs are the broken ones, although I have tested them with a 9V battery and the first and last lit up but the two in the middle did not so I will start with them.


    Thank you for the advice will check it out and see if I can get them and get that done. Is it bad to just bypass with a wire or does it need to be the diodes?

    Cheers
     
  4. Harald Kapp

    Harald Kapp Moderator Moderator

    11,316
    2,593
    Nov 17, 2011
    It depends on the driver that delivers the current to the LED string. If it is a curret controlled driver (constant current) then a strand of wire would be fine as the driver will compensate for the lack of voltage drop across teh wire.
    If the driver is a voltage driver has not current regulation but only a series resistor to limit current, the lack of voltage drop across the strand of wire (compared with taht acros an LED) will lead to a higher current through the string which in turn may damage the other LEDs.

    There is one thing that irks me: Your diagram shows LEDs pairwise in parallel. It also shows only one LED of each affected pair to be defect. The remaining LED should work, although at a higher current, therefore the whole string should work, too.
    Or is the power supply not DC (as expected from the +/- markings) but AC? In that case the LEDs in each pair would be connected in alternating orientation and only one LED of each pair would light up during each half wave of the sinusoidal input. Since 3 of the "upper" LEDs and one of the "lower" LEDs are broken, no current will flow in either half wave and the lamp will be dark. I liek to assume that this is the case here:
    upload_2018-1-10_7-16-9.png
    The the markings +/- are misleading as + for D1 is - for D2 and vice versa.
    You now need to be very careful when inserting the replacement diodes: Observe correct polarity!
     
    H3LLRA1Z3R3 likes this.
  5. H3LLRA1Z3R3

    H3LLRA1Z3R3

    4
    0
    Jan 9, 2018
    There's a lot for me to process there mate, thank you.

    Just a slight update, when I looked over my amazing diagram again I saw that the two dead leds in the middle were on either side off each parallel set. I started there and bingo that was the cause.
    I used a wire to bypass them and all the lights went except the two in the middle.
    This is great new for me as it means I know where the problem is, but I need to remedy the problem properly so it doesn't set my house on fire.
    If it helps , Both working, wired LEDs measure 6.5v and the input power of the channel effected was 98.6V which is 13V less than the other channel as this channel has two 6.5V leds missing.
    I tried testing mA once and it didn't go so well so not keen to try it on this just yet.

    It doesn't say on each driver specifically but the model number has AC and the wires coming out of it have A+ and AC-. All my other LEDS are DC and it sparked when I used the mulimeter on DC but not AC so I think AC out to the LEDs If I understand that right it means it's based on mA. I've been wrong before so will check if needed.

    There is no wiring, it looks like it has flat strips of copper or something between them and they zig zag to each light but they are covered in a black film of some sort so you can only just make out the pathways.
    the wiring diagram was from me testing each connection with the multimeter and writing the result down. Plenty of room for error there.
    When the dead led was touched the working one in the same parallel set would light up, then when tested with a 9V battery both LEDs lit up in that set but no others and this was the same across the entire board. Again though the two dead ones in the middle didn't work at all

    Hope some of that helps, it must be like teaching a blind man to drive, but the help is much apprecitaed

    I was unawares that the +/- wasn't used for ac sorry as the board has +and - written on it and its about as far as my knowledge goes. I will have to look that up and try to understand it better.
    The thing is though, the two wires leading out aren't joined to each other they lead straight out so the circuit isn't connected until they hit the driver. ( if i understand your diagram right. )
    There is little icons on each LED that are the diode symbol, on each side of them is tiny metal squares where I put the ends of the multimeter. It only lights up when the black and red are in the right order so is it a safe assumption this is the current direction?

    I looked up the diodes and they just confused me even more but I will deal with that later.
     
  6. Harald Kapp

    Harald Kapp Moderator Moderator

    11,316
    2,593
    Nov 17, 2011
    In your diagram they are, cf. post #2: each pair of LEDs is joined at both ends (one pair = within one red circle), then the traces on the pcb lead on to the next pair.

    For one you should not test with a battery without havin some current limiting resistor in series. For a 9 V battery an d a 6.5 V drop across teh LEDs, use a resistor of approx. R = 2.5 V / 10 mA = 250 Ω (anywhere between 220 Ω ... 300 Ω will be fine).
    Secondly: if these LED strips are for AC, as I assume and as you confirm by
    the LEDs should be anti-parallel as shown in post #4. Only one LED should be alight for each polarity of the battery: the one where anode (+) is towards + of the battery. When you reverse the battery, the other LED should light up. If not, the LED strips are driven by DC, not AC and the little + and - symbols are correct This is what you measured.
    You should therefore be able to measure a DC voltage at the output of the LED driver (input to the LED strip). By now I think this is the way this is connected. Sorry for any inconvenience my misinterpretation of your diagram may have caused.
    This is what I nor think your circuit looks like:
    upload_2018-1-10_11-23-41.png
    The pairs of parallel LEDs have the same orientation which matches your tests.

    This doesn't directly match my idea of the LEDs being antiparallel and thus confirms above circuit.
    It sounds like there is an interruption in the traces leading to the diodes. You could try to "repair" this by soldering a strand of wire from one diode pair to the next to bridge the broken trace.
     
  7. H3LLRA1Z3R3

    H3LLRA1Z3R3

    4
    0
    Jan 9, 2018
    It's amazing how I haven't set fire to myself yet.

    No Need to apologise I have much more to learn it seems.
    It's confusing but I think you got me there in the end.

    I have attached the wire and bridged the broken gap. light is on the tank and has done well today.
    I had it yesterday on for a bit at a time and seems to have gone okay.
    I will be less stressed about it as the weeks pass.

    Thanks again...

    [​IMG]
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-