The top end of the sink has some nice M5 holes for mounting a ballast, or other nonsense to, but I'm going to steal that, and instead mount a stainless steel plate 2.5" in diameter I got off Amazon. So I took the plate, drilled a bunch of holes in it, got some stainless steel mounting hardware, and some small ABS spacers, and made myself little hangers.
Once the hangers were complete, it was time to mount the LED's to the sink, and bolt it all together. I got some 50mm M3 screws, and used them to mount the light rings I 3d printed, and used the lens assembly and light right to basically mount the LED to the sink. Time to give it a quick test at about 12% power on the floor.
Now speaking of power, I overdid things when I got my ballasts. I didn't quite understand how the Chinese rate LED's, so I got ballasts that were a bit more powerful than I think I'm comfortable running these days. My ballasts are 185 watts. I think realistically these are 133W chips, maybe less, but they run well at 140w, so, lets say 133. So 12% power is really 22W, which is pretty significant. They lit right up, no noticable heat at all.
Now was the time to bolt the hardware (stainless steel again) to the hangers I put in the ceiling, and start hanging a light! I used 1 1/4" washers on the top, and one of those little nuts with the nylon inside that grips the thread better. This should hold a small anvil, so I think I'm good here.
I experimented with various hooks, D rings, and other fiddly bits, but eventually went back to the rope hangers I purchased specifically for this. They were approximately the right length, and the slight adjustability was just about right.
I carefully set the Apex to 15% power, and plugged the ballast in, and...
Woo. I can has light. Lets see how it looks from the front of the tank, and slowly ramp it up.
This is 15%, or 28 watts.
Lets try 30%, (55 watts)
Ok, ok, working so far, lets get crazy and go 50%, 92.5 Watts.
Ooh.. I'm liking it. It's starting to look like it used to. The lights aren't quite as blue as the photos seem to make them, but they are a nice bluish white, which makes me happy. I'm getting about 300 PAR here through the rather dirty acrylic top brace that is directly under this light, and about 1" under water. Heat sink is measuring about 95' F, and a temp probe put directly on the chip is about the same..
Lets try 65% (120 Watts!)
Wooo. Liking this. About 400 PAR 1" under water (dirty brace still). Now the sink is reading about 113 F, and about 117.5F at the chip. These chips are generally happy under 60' C, and 117 is about 47.5 C, so, I'm still good here.
The sinks are *really* good. Like some of the best I've ever worked with in my life. If I take a measurement inside the sink, right above the chip, I get about 113'. If I move all the way to the top, and take a measurement inside the sink (on the central heat tube core again), it's like 112.5. That means the sink is wicking the heat up the full length of the sink, and not just dissapating it all at the bottom. This is exactly what you want these things to do. These are well worth the $104 each I paid for them.
At this point, I get a little nervous, and decide to crank them back down to 60%. (111W). Very little output difference really, but the heat drops to a steady 116' F at the chip. Just out of curiosity, I decide I should measure the heat on my 20W passive cooled chips over my softy tank. Those chips have been running every day for like 6 years now. 147'F. Wow. Just wow. Didn't realize they were running that hot. Either way, I think I like 60% power for now, so I decide I'll let it run for about 3 hours just to be sure it's stable.
The lens holders I printed have some V shaped sides on them, which you can see in a few of the photos. These reflect the light downwards a bit more than just the lenses. In fact, they do such a good job, that the light doesn't quite reach the front glass. I am unsure how happy I am with this tight of a beam. I'll probably have to try it out with all 4 at full running power (which will likely be 60-65%, even though I used to run them at 75%) and see how I like the effect. I also need to get some PAR numbers with all 4 running, but that will be saved for another day.
In the meantime, back to work on the other 3 lights. I cut some leads, this time using 16 AWG wire, instead of 12 AWG, as I just don't need that kind of current handling for the short leads, and they interfere with the lenses a bit. I twist the wire up, and solder up the last 3 LEDs. Now to make the rest of the mounting brackets, and assemble everyone!
I won't go into the full details of the build right now, I'll probably save that for an article. However, because I'm re-using some components from the previous failball, I need to do some cleanup here. The lenses are covered in salt (2 of them from falling into the aquarium), dirt, and 2-3 layers of failed glue. Need to clean all that off, and get the lenses all sparkly clean on the inside before I mount them up.
Also, because I am not exactly Hackaday material with my soldering skills, I had to dremel out a bit of the bottom of each reflector, so the solder joint wouldn't interfere with the lens fitting.
Two of the lights I recently purchased are 50/50 lights instead of the 60/40 blue/white I ran before. Seems the 60/40's are harder to get now, though I did find a line on them in case I need more. But mostly I'm hoping these 50/50's are good enough. The wiring on them does seem different though, so I'm a little unsure as to how they are wired internally. Normally I would snip off the outer tabs on the LED's, but in this case, I'm a bit nervous that they are integral, so I'll be leaving them on the new ones. I'll have to try them out tomorrow or something, and see how the light compares..
I'll leave you with a fun picture I took right before shutting the system off after the 3 hour test. The lights at 1%, or 1.85 Watts.
Exhausted from hours of holding a temperature probe in a little hole, I shuffle off to bed..
No comments:
Post a Comment