NASA tri light every other led failed.

[pcatterall]

My NASA trilight has worked well for 5 years but now every other led on all 3 colours has failed.
The circuit board componants show no signs of burning or have a burnt smell and that is the end of my electronics skills!! is there one componant that may have failed with the above result and can any one identify it and say how I could check that ( with a mutimeter)
I have fitted a new one but I guess it would be nice for a spare.
Many thanks.

 

[Lon nan Gruagach]

It does indicate a single point of failure, may be a component, may be a connection or wire or PCB track. A good magnifying glass may be as much help as a multi meter.
Method:
Asses what must be 2 identical circuits, test for voltage between similar points on both. At some stage the voltage difference will change from + to - (or the other way round), the place you just passed is the problem.

 

[rogerthebodger]

I have had that happen to both my NASA Tri colour and my anchor light. Both cases one or more LED's went open circuit and thus shut that bank down. You can check which ones by shorting the LED's in turn until you find which one and replace it with a high intensity led from your local supplier. This happened to me 3 times on my tri colour and once on my anchor. My anchor I sent back to NASA for then to tell be why it failed and really got no response at to why only the LED had failed which I already new. When NASA wanted me to pay for the repairs and the return postage that was more than the original cost I took the NASA lights down and replaced then with the aqua signal I has up in the first place and fitted warm white LED bulbs which work fine.

I think the NASA nav are not fit for purpose but not prepared to fight.

Mine failed within a month or 2 without being used in anger, when I refitted the repaired unit I had failed the next time I visited which is alternate month at home , month on boat.

Hope you have better lick than I did.

 

[RobbieW]

I'm not convinced by the sealed unit LED types either. I had one of the Lopo masthead lights fail a couple of years ago, replaced it with an AquaSignal standard fitting using LED lamps

 

[Alfie168]

After my NASA light had been on the seabed a day or so, it too...not surprisingly perhaps, did not work on all LEDs..every other one seemed OK. I too linked out each failed LED in turn with a short length of wire, and bingo! I soon identified the failed LED as all the others lit up (there was nothing visibly obvious). When I phoned NASA for the new dome and sticker I needed and described the fault they sent me an LED in the pack FOC. I duly unsoldered the old LED and fitted the new LED, and gave the whole circuitry a light spray of silicone grease to try and halt any incipient corrosion from its undersea adventures. Its been fine ever since, not that it gets used a lot, but I did 'soak' test the light on a car battery for three days and nights before considering it re-mountable.

With NASA, if you are anything like any good at DIY then the parts are not expensive to do self repair, and their service and supply is second to none on that score in my experience. Unless its under guarantee then sending things back without a firm quote is always risky as labour=time, and time=money. I'm not sure LED lights as a group have yet achieved the total reliability one might hope from them. Some with very pricey units have had problems no different to those experience by some with the NASA Supernova range.

Tim

 

[William_H]

LEDs fed with an electronic current control often use a kind of boost converter so that the LEDs run on a relatively high pulsed current and high voltage.
Such that while running lEDs on direct 12v DC the max number of LED is 3 in series on any one chain. But most of these commercial lights are set up for automatic 12 or 24v supply and the converter can then run 6 or more LED in series even from 12v using an inductor to increase voltage. Hence it is likely that the design runs 2 chains of many LEDs. The problem being a failure of any one LED will kill that whole chain. In your case half of the LEDs. As said try shorting out each LED and if you find the rest of the chain come good with one shorted that is your problem LED. However the problem could be caused by corroded printed circuit board track having a break cutting the chain. Bit more difficult. However if you try using a digital volt meter to track the current don't be surprised if you get crazy and varying readings. Good luck olewill

 

[pcatterall]

Thanks all, everything on the board looks good so will trythe ‘bridging’ idea to id the LED.
Then see if I can replace or could I just bypass that one?

 

[Alfie168]

Thanks all, everything on the board looks good so will trythe ‘bridging’ idea to id the LED.
Then see if I can replace or could I just bypass that one?

It's the same job really, and personally a missing LED would bug me. It's not a difficult job. The existing led unsolders easily enough and pulls out with fingers or needle nose pliers. I used a desolderer to get the blobs off. The new led comes with long legs which enables you to thread it through the holes and you solder it in carefully and once set you snip the excess length of leg off........then fingers crossed and test it!

Tim