Did they understand the issue Spinnaker was raising? Hitting a wave and then crashing down at 40mph can cause some very large dynamic pressures. Or bumping a log or an unplanned "beaching". Doo doo happens. Just look at the hull of my Sea-Doo.

Exactly, A bucket won't do. That is a static pressure test. To test properly in the lab, you would need a large tank with wave machine and a way of dragging the light through the tank from one end to the next.

And even after that as wayne points out there can be other issues lurking out there.

It is probably best to test in the actual environment under actual conditions before you go to far.

 

At 3.8vf and 3 leds in series that is 11.4 volts. That is right on the edge. Add wiring and the voltage drop from your current regulation, the light won't light, if it does not for very long. I would consider putting 2 leds in series.

Also have you done your math for the time this light will stay lite on a single sla?

BTW a car battery is a poor choice for this project. Consider a marine battery, one designed for house lighting.

 

Do you have a part number for your LEDs?

 

spinnaker, I would absolutely test in a lab under those conditions. However I don't have access to that type of equipment. The closest I would have is a marine mechanics shop, with a dyno machine they use to test outboard motors, under pressure to simulate being on the water.

Alas, I am limited to old fashioned trial and error. The connection will be to a Marine Deep Cycle battery vs an automotive battery. To test the equipment, I have a 12 volt 1.75 amp power supply with test leads. Same thing I use for all my 12 volt projects. The image was for demonstration purposes.

ronv...I hate to post a link to an ebay listing, that will be a dead hyperlink in a few months, but I purchased these exact LEDs

http://www.ebay.com/itm/14105123914...eName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

 

spinnaker, I would absolutely test in a lab under those conditions. However I don't have access to that type of equipment. The closest I would have is a marine mechanics shop, with a dyno machine they use to test outboard motors, under pressure to simulate being on the water.

http://www.ebay.com/itm/14105123914...eName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

That would work but why not take a prototype out on the boat and test under actual conditions? You could get your testing done and some fishing too!

 

Exactly spinnaker! A perfect excuse to go fishing.

I'm hoping to get a good prototype working, so I can sea-trial it on the water. That's why I came here for some input on the wiring aspect on getting the length of wire, components, to go with my (3w) 24 LED (72 watt) set up.

 

My suggestion is just test the case itself. You will still want to drill a hole for your wiring and seal it up just as you might in the final product. That way if you get a breach then nothing is lost except your pride then it is back to the drawing board.

Put the prototype in bump it around. Maybe have a pipe or something on board where you can give it a smack every now and then bump it up and down. Even lower it to a level much further than you expect in actual use. Basically abuse the heck out f it.

 

I will do just that, I'm about to seal it and mount on the transom. After trailering it the launch and running around for a while I will inspect for a breach. I'll check in after while.

 

Here's what I have so far...

I'd consider adding that DC-DC converter, and running the LEDs at 24V (or some other, higher value). That takes the wire resistance out of play, adds a degree of isolation in between the noisy charging system and the LEDs (although I have no idea whether a transient spike will be stopped by that device), and ensures a steady supply to the LEDs whether you are charging or not. These are all good things. You could even adjust the device to give some dimming. Be sure to add a fuse to your light circuit.

 

I made a 42 LED lamp using 42 LED's in series-parallel configuration: 14 strings of 3 LED's each with 330 ohm resistor in each leg. Total power is 3.5 watts @ 12 VDC. If you are worried about alternator transients, a 15-16 VDC zener diode reversed biased across the input will solve that. The waterproofing issue is more complicated. Trying to make things completely waterproof almost guarantees that water will get in and not be able to get out. Precision machined grooves in a metal housing and double O-ring seals may get you there...but I would prefer a good epoxy marine potting compound leaving only the LED lenses protruding. Of course, its a no maintenance design then.

Cheers, DPW [Everything has limitations...and I hate limitations.]

3 deg. F here (no boating today)

 

I ran the enclosure for a few days on the water. I was pretty rough with it >>> smacking it a few times, trailered. No leaks, sealed tight. Very pleased with the quality of the enclosure.

wayneh, I think I will opt for the DC converter. I hate the extra cost, as I am building 4 lights (2 green and 2 blue) running them simultaneously. At least it will Isolate the circuit from transient noise coming from the engine alternator. I am going to add the 15v Zener diode on the source line like everyone recommended, and a potentiometer to dial down the voltage dimming the lights.

I calculated 22' of 22 AWG Teflon coated, military grade shielded wire for each light.

Duane, what you did is exactly what I am doing. Nearly half of the LED you used. 42, qty, that's a lot of soldering! What watt LED's did you go with? 1 or 3 watt? The 3 watt, that I just bench tested are insanely bright. blinding, if you look too long. I am using high grade opticlear epoxy designed for potting in marine conditions. Hopefully it wont have a lot of bubbles on the finished product. Once done, it will be sealed for life.

I am going to order some more 3W GREEN and BLUE LED beads. Probably 100 qty of each color. These last ones were from China. I would like a higher quality LED, if anyone can recommend a manufacturer. Or if they are all the same it won't matter. These were $14.00 for a 50 qty. Maybe someone has had some experience with a specific one. Considering the effort on this project, I want something that has the lifespan and quality.

Thank you everyone for your input.

 

ALSO

I thought I had some Zener diodes stocked. All I have are rectifier diodes I use for car alarm/remote start installs. I'm going to order some online now.

What are your thoughts are the values for the zener? 15v 1 watt should be sufficient I would think.

 

After reading your post above, this is a better solution. It has constant current adjustment. You can replace the little trim pot with a full sized one if you like. For a couple dollars more I think they make one with meters on it so you can see the current. Then you can use a little bigger wire if you like and not have to worry about it.
http://www.ebay.com/itm/5-30V-to-0-...054?pt=LH_DefaultDomain_2&hash=item3f2d0d9b26

As for the zener - bigger is better, but if you run the boat without the battery connected it will probably still pop. (1N5353 is a 5 watt one)

If you use this supply think 6 in series at 2.4 amps max for all 4 strings.

Summarize the whole thing again before you go for parts.

 

ALSO

I thought I had some Zener diodes stocked. All I have are rectifier diodes I use for car alarm/remote start installs. I'm going to order some online now.

What are your thoughts are the values for the zener? 15v 1 watt should be sufficient I would think.

What is the purpose of the zener?

With LEDs you do not need to regulate voltage. You need to regulate current and for that you should use a buck regulator.


I have used these:

http://www.ledsupply.com/buckpuck-constant-current-led-drivers

Very easy to use. What is nice is they have the bonus of a 5V output, in case you want to add your PIC. You of course would need to be sure yiu select a 5V PIC.

You will also need to check if they make a Buck Puck in the current range that you need. Luxdirve make all types. I would be surprised if they don't have what you need.

 

ronv...that is a good find. I may order a couple, whether I use them on this project or not. Very handy piece.

spinnaker, I keep coming back to these buckpucks as well! It seems to be such a versatile piece. A broad DC input Voltage range, and with the harness and POT add on, really a good choice!

The zener was going to serve as to protect a spike in voltage from alternator cycles. The buckpuck will do that for me.

So the LED's have a 600-700ma maximum. I am thinking ordering the 500ma will be a good choice. What do you think?

I am now considering doing two enclosures, instead of four. I have so much room in the enclosure, and not enough room on the transom of the boat to add four. Two large enclosures would be better. Half GREEN and half BLUE in each enclosure. 48 LEDs in each. I am not sure about the heat dissipation. I need to calculate how much heat these would put off. It may be too much for the enclosure, not sure yet. Or I get 4 smaller enclosures.

 

The buck pucks are nice but the problem is this:
They are rated for current output. So if you buy a 500 ma one and run it on 2 batteries in series you could run one string of 6 (6X3.8 volts) for each buck puck. So you would need a lot of them.
The wire idea is ok, but is not dimmable, and you have 35 feet of wire running around. They will be brighter when the motor is running and dimmer as the battery voltage goes down. The constant current regulators keep this from happening but maybe you don't care?
You need a way to get the aluminum plate in contact with the water to get rid of the heat. Hopefully you have thermally conductive epoxy? If not heat will be a big problem as at 500 ma. each led is a couple of watts X 24. Think trying to unscrew a 50 watt bulb while it's turned on. Really hot. Much hotter than the LEDs can stand.

 

Ordered some of the new parts in, Waiting for shipping from China and then will be back on the complete the project. Thank you so far everyone.

 

Hello everyone. My parts came in, but I had to put everything on hold do to a very painful herniated disc that required spine back surgery. I don't recommend a microdiscectomy, long recovery.

So as I started feeling better I began working on my project again. The first prototype failed. The LED's overheated and burned out. Melted the clear epoxy resin. So I went back to the drawing board and found a potting enclosure that was smaller in size. I assembled the parts, aluminum heatsink (nearly the size of the enclosure), thermal grease, and mounted the LED's. I finished off with MG Chemicals 8322 optical clear potting epoxy. I spoke with MG Chemicals technical support and they sent me a large batch of the resin and hardner. Said it was best for marine and salt water conditions. I poured the resin, oven cured for two hours as per the instructions. Took it out and let it set at room temperature for 6 hours. The caps of the LED's were slightly exposed through the resin for direct contact with the water and cooling purposes.

Filled up a bathtub, submerged in water and connected my power supple.

BANG! Beautiful high powered green LED light! After a five minutes I disconnected the power and pulled out the enclosure to inspect.

CRACKED CLEAR EPOXY! several large cracks, similar to a cracked piece of glass. Disappointed can't describe it.

I don't know where I am going wrong. Am I not getting enough heat dispersion? Did the LED's flex and cause the crack?

PLEASE HELP. Any input would be tremendously valued. I'm stumped!

 

I'm guessing that is thermally conductive epoxy holding down the LEDs?
The back side of the aluminum plate needs to be in contact with the water or everything will get to hot. Is the back in contact with the water?

 

Hi Ronv. Yes the epoxy is thermally conductive. I'm using thermal paste for the LED's to sit in.

This new enclosure I am using is just a regular potting box, made from black ABS plastic. Since the case is sealed shut so water doesn't corrode the aluminum, How would I get the metal on the bottom of the heatsink or metal in contact with water?