Economical multi LED Chip Projects for Growing

A POSSIBLE FUTURE DESIGN AND TEST BED
Yesterday I had an idea for testing the strength of silicone seals that evolved into an artsy kind of idea for an attractive water cooled fixture that could be hung over an aquarium or a light hungry living room plant.

The lamp would be constructed out of a sheet of 1/4" sheet of aluminum, with either aluminum, plastic or plexiglass spacers around the outside edges to create an sandwhich of 1/4" aluminum plate, the spacing material and a 1/2" thick sheet of clear plexiglass. There would be between a 1/4" and 1/2" gap between the aluminum and plexiglass sheets that would be the water cooling chamber, baffles could be included here in a functional and attractive pattern to direct the flow of coolant. The whole thing would be glued together with clear silicone adhesive, have four eyebolts in the corners for suspension (tapped into the aluminum plate) and brass water fittings coming in through the top plexiglass sheet and clear hoses. Once assembled and covered in masking tape to prevent scratches, corners could be rounded, the edges of the assembly might be filed and belt sanded smooth and perhaps polished. Maybe some low power blue LEDs in the water chamber for effect and some food dye in the water.

Attach the COBs to the aluminum plate with CPU tape etc, hide the drivers wires and hoses creatively, get innovative with the reflector and I could have something nice. Anyway, a good possible basement project for this winter when the grow conversion is complete.
Here is a picture of some of the materials I might be working with this winter. The lamp would be the size of the sheet of aluminum and, if I wanted, have hundreds of watts of power on it, though it's only purpose then would be to grow bud. Hey ya never know, medical growing is legal where I am and recreational will probably be in the spring, so a living room christmas tree might not be out of the question!

DIY-HP-LED said:

A POSSIBLE FUTURE DESIGN AND TEST BED
Yesterday I had an idea for testing the strength of silicone seals that evolved into an artsy kind of idea for an attractive water cooled fixture that could be hung over an aquarium or a light hungry living room plant.

The lamp would be constructed out of a sheet of 1/4" sheet of aluminum, with either aluminum, plastic or plexiglass spacers around the outside edges to create an sandwhich of 1/4" aluminum plate, the spacing material and a 1/2" thick sheet of clear plexiglass. There would be between a 1/4" and 1/2 gap between the aluminum and plexiglass sheets that would be the water cooling chamber, baffles could be included here in a functional and attractive pattern to direct the flow of coolant. The whole thing would be glue together with clear silicone adhesive, have four eyebolts in the corners for suspension (tapped into the aluminum plate) and brass water fittings coming in through the top plexiglass sheet and clear hoses. Once assembled and covered in masking tape to prevent scratches, corners could be rounded, the edges of the assembly might be filed and belt sanded smooth and perhaps polished. Maybe some low power blue LEDs in the water chamber for effect and some food dye in the water.

Attach the COBs to the aluminum plate with CPU tape etc, hide the drivers wires and hoses creatively, get innovative with the reflector and I could have something nice. Anyway, a good possible basement project for this winter when the grow conversion is complete.
Here is a picture of some of the materials I might be working with this winter. The lamp would be the size of the sheet of aluminum and, if I wanted, have hundreds of watts of power on it, though it's only purpose then would be to grow bud. Hey ya never know, medical growing is legal where I am and recreational will probably be in the spring, so a living room christmas tree might not be out of the question!

View attachment 3779228

Click to expand...

So what would go around the spacers to hold the water in?

Let's say I cut some 1/2" wide strips off the 1/2" plexiglass sheet to put around the perimeter. After sanding and cleaning all contact surfaces, I'd apply a continuous 1/8" bead of clear silicone all the way around and attach the side pieces to the aluminum, sealing the gaps between strips with silicone. Next the baffles get the same treatment, then the plexiglass top (with fittings installed) goes on using the same method, sand the contact areas and clean. When the silicone fills in the space when it's pressed together the scratches in the plexiglass largely disappear. This job has to be planned and carefully executed, but it can be done.

Airwalker16 said:

So what would go around the spacers to hold the water in?

Click to expand...

You don't need a thick plate like mine, you could mount the COBs on tinfoil as long as you weren't over a plastic baffle. Might make a better thinner design, thinner aluminum and plexiglass, though the fittings would be an issue. A fellow could go ipad with this and get real thin after a few tries.

A Cheap, Radical Experiment in Water Cooling COBs
Here's something someone might like to try, it'd make a nice youtube video, maybe I'll make a demo myself .
Cut out a square 3"x 3" section out of the bottom of a square plastic sandwich container that has a water tight lid. Silicone on a piece of heavy kitchen aluminum foil a single layer thick on the inside of the container bottom. Sand and clean the inside bottom of the plastic for increased adhesion and say put the dull side downward facing, be careful not to have any wrinkles in the foil, or get a heavier aluminum oven tray from the dollar store and cut the thicker aluminum from a flat area of it. After successfully water testing, mount a 100 watt COB on the underside in the middle of the aluminum square with double sided CPU tape. Now figure out a way to pump water into and out of the container with a cheap ebay pump and get rid of a little heat on the other end (say a 20 liter bucket or tote). You just made a water cooled jacket for testing silicone sealing, thin aluminum back plates and anything else you can think of that could be used in an innovative larger design. See what you can get away with for the cost of a cheap 100 watt chip, a 50 watt cheap driver and a $4 (CDN) ebay pump. Say $5 for the chip, $6. for the driver $4 for the pump, with silicone and other materials you would probably blow 20 bucks. Even if it's a test model, if it works well, it can still grow a small plant during further testing! You might say to yourself, " if I can do this, why not something bigger and better!"

CAUTION Don't run this rig unattended without some kind of thermal protection system to shut down the power! When you order your parts order some of these at least and maybe one of these, just in case things fuck up! Run them in series with the driver's AC line power supply and make sure the thermal sensing points and probes are stuck next to or on the COB edge with CPU tape. Maybe a gob of silicone or something better to make sure it stays there, perhaps aluminum tape. Safety first folks, don't burn down yer house, your wife would be PISSED!
I'll cover the topic of protecting your lamp and life with multiple layers later in the discussion.

Maybe a better name for this thread might have been, Redneck's Ghetto Grow...

Once folks get used to the idea that these things run at no pressure, ambient temperatures and no thermal cycles to stress the adhesive joints, I figure the idea will catch on. Modern adhesives are quite good if used correctly. When you accept the new paradigm and realize that the cooling medium is an order of magnitude better that what you're used to, it frees up yer imagination a little.

WARNING: If yer imaginative, logically inclined and practical, this can be addicting. YOU WILL BE ASSIMILATED!

PRESSURE TESTING TUBES AND OTHER FIXTURES
I think I'll head down town and pick up a schrader valve for a couple of bucks to pressure test the tube designs.

I've got a car 12 volt tire inflator compressor with a gauge, so why not. I'm curious to see what pressure they can hold, not that it's required for the system I'm planning, but it's a good indicator of leak resistance and test of adhesive bonding strength. Without a doubt, welding or brazing end caps on and pressure testing would be the best approach to building aluminum water cooling tubes, but other approaches might work well enough. Stay tuned for some hydrostatic pressure testing and how to safely hydrostatically test adhesive strength and leak resistance using aluminum tubes.

I'll fuck up, so you don't have too...

Got my schrader valve, sometimes known in the hardware and plumbing trades as a snifter valve for about $6 CDN. A bike inner tube or car tire has the same thing for free or cheaper, cut the large rubber end off, slather it with silicone, shove it into a length of hose and put a hose clamp on.

PRESSURE TESTING TUBES
Just finished supper a puff and I'm enjoying a coffee, ah the finer things in life... When I came up from the workshop hungry, this is where I left off. The hydrostatic test rig is completed and ready to go, just add water. First up, a tube with nylon fittings threaded into the end plates with silicone for a sealant.


Now only a Goddamned fool would hydrostatically test something without at least doing this first. You'll also notice that the 120/12 volt power supply is nowhere to be seen, keep it far away from water, don't be stupid.


I should be able to get pressure readings (dunno how accurate it is) from this. I have another one with a gauge that I could compare I suppose, to get some idea. I got it! I've got a tire pressure gauge somewhere... Next problem, finding it! Then a short hose with a (cheaper) schrader valve on each end to calibrate the compressor gauge, maybe cross check with the other compressor I have...


This setup is for demo purposes, before I use it I'm going to shorten up the air supply hose and turn it up the other way before pouring water in to fill the tube and connect the supply hose. I'm shortening the hose so that there will be less compressed air in the apparatus during testing.
I'll explain after another puff and slurp or two of coffee. In the meantime you can bone up, or figure out why I'm doing this.

A Tale of Two Fools
Let's say two buddies, call'em Jimbob and Cletus decided to do something similar to the above and pressure test a 3"x 1" 4 foot long rectangular tubing made into a grow light water cooling tube, like the ones I'm gonna use in my grow room conversion build. After installing the hose barbs, and siliconing in the end plugs, then letting everything cure for a couple of days. Jimbob heads for the 1/2 ton to grab the compressor using a step ladder to get up into the high the truck cab (we're dealing with stereotypes here).

Now they seal one of the hose fittings with a bent back and kinked piece of vinyl hose, held tight with a nylon wire tie (a good plan for testing) and hose clamp the other end to the fitting. These boys don't use water and run a "dry" test to look for leak bubbles with soapy water and a brush. Jimbob fires up the compressor and the noise fills the barn, meanwhile Cetus is standing a few feet away from the end of the tube making a YouTube video of the whole thing on his phone. After the compressor was running for awhile, the pressure on the gauge climbs to, say 40 psi (the boys did a real good job on the silicone) all of a sudden KABOOM! A plug blows out of one end of the tube and dorks ole Cetus square in the nuts, down he goes like a sack of potatoes, writhing on the floor in exquisite agony. Jimbob who was to one side painting on soapy water looking for leaks, shits his pants. Now if this wasn't bad enough, the neighbors hearing the explosion rush to the scene, "I thought someone wuz shot", one of them says to the cop...

This tragic situation could have been avoided entirely with hydrostatic testing, all they had to do was fill the fucking tube with water before applying pressure! When the silicone let go there would have been a gentle "blurp" sound as the end plug failed and the water leaked out, the plug would probably still be in place. Cletus would have his nuts unscathed and Jimbob wouldn't need a change of underwear. If there was any pressure in the supply hose, it might cause a little spray, but I doubt it. The compressor would have only run for a minute and there would have been much less potential energy in the system because liquids are in-compressible. Avoid explosive decompression, USE HYDROSTATIC PRESSURE TESTING, keep your nuts, your wife might appreciate that.

In this case, Jimbob and Cletus fucked up so you don't have to.

Hydrostatic Pressure Testing Update
I decided to make some changes to the hydrostatic testing equipment and procedures. The air compressor I'm using isn't appropriate to the task at hand, since I've only got a couple of cubic inches of air in the test and compressor hoses, the pressure increase is too rapid and high for this job. Since the cooling system I'll be using is unpressurized, I want to concentrate on more controlled, timed, low pressure tests in the 10 psi range. I also want to test some other types of tubes and plugs to failure, just to see what they will take and try improved methods etc. I'd like to develop the idea a bit and see where it leads, hopefully not to disaster!

Now I'm going to use a hand bicycle pump that has a good gauge and it will pump up to the 90 psi range max, well in excess of my needs. This should give me more control and time to record data accurately, instead of turning on the compressor and having the needle almost immediately climb out of the desired test range. My plan to test the 1" x 1" grow light tubes is to pump them up to 10 psi and see if it will hold pressure for five minutes with no leaks, one or two might go a bit higher and longer, but not to failure. I'm gonna make up a special test tube for testing to failure and trying new sealing materials and methods and also multiple tests to get an average failure pressure. I might test some of the 3" x1" tubes to failure too, but mostly to 10 psi, which is 3 times the max possible pressure they would ever experience. Since they are the highest point in the system, they should experience the least pressure, in actual fact there is no pressure in the system since it's open. This testing is more for leak resistance than anything else, if it will hold water at 10 psi, it won't leak at 0 psi. As far as I know, I'm trying something new here, so I'll have to collect some data, so I can design these things better and speak with confidence and integrity.

I picked up my nylon barbed fittings that were on order today, so I'm ready to get my testing apparatus set up in a new configuration, finish the remaining 4 tubes and test em.

I'll fuck up, so you don't have to.

DIY-HP-LED said:

Hydrostatic Pressure Testing Update
I decided to make some changes to the hydrostatic testing equipment and procedures. The air compressor I'm using isn't appropriate to the task at hand, since I've only got a couple of cubic inches of air in the test and compressor hoses, the pressure increase is too rapid and high for this job. Since the cooling system I'll be using is unpressurized, I want to concentrate on more controlled, timed, low pressure tests in the 10 psi range. I also want to test some other types of tubes and plugs to failure, just to see what they will take and try improved methods etc. I'd like to develop the idea a bit and see where it leads, hopefully not to disaster!

Now I'm going to use a hand bicycle pump that has a good gauge and it will pump up to the 90 psi range max, well in excess of my needs. This should give me more control and time to record data accurately, instead of turning on the compressor and having the needle almost immediately climb out of the desired test range. My plan to test the 1" x 1" grow light tubes is to pump them up to 10 psi and see if it will hold pressure for five minutes with no leaks, one or two might go a bit higher and longer, but not to failure. I'm gonna make up a special test tube for testing to failure and trying new sealing materials and methods and also multiple tests to get an average failure pressure. I might test some of the 3" x1" tubes to failure too, but mostly to 10 psi, which is 3 times the max possible pressure they would ever experience. Since they are the highest point in the system, they should experience the least pressure, in actual fact there is no pressure in the system since it's open. This testing is more for leak resistance than anything else, if it will hold water at 10 psi, it won't leak at 0 psi. As far as I know, I'm trying something new here, so I'll have to collect some data, so I can design these things better and speak with confidence and integrity.

I picked up my nylon barbed fittings that were on order today, so I'm ready to get my testing apparatus set up in a new configuration, finish the remaining 4 tubes and test em.

I'll fuck up, so you don't have to.

Click to expand...

What fittings were you waiting on that were out of stock?

Thanks for dropping by, nylon 3/8 hose barb fittings like the ones shown in the photos a little further up the thread.
I'm mostly focused on coming up with low cost designs for DIY LEDs for poor medical users and others with limited means. Not worried about running costs and efficiency, so much as low cost start up. I'm also solving my own grow problems and sharing what I learn and experimenting with LEDs

220 Volt Driverless COBs
I was thinking of writing a little cautionary tale called, Jimbob and Cletus die, or maybe fry.
I've been noticing dirt cheap driverless COBs online that run on 220 AC, hook em up to a clothes dryer cord stuff. Now these probably use bridge rectifiers and perhaps capacitive droppers, are not isolated from mains voltage and it looks lethal to touch them and perhaps the heatsink they are mounted to, scary stuff. I wonder if someone would be stupid enough to try a water cooled light bar, full of these, maybe Jimbob and Cletus, what could go wrong? Maybe I'll call it, Low Cost 'n Lethal

Alas, we should write from experience, and playing with one of these things might be the end of all experience. I like my life and think I'll hang on to it for awhile. One of these could suddenly end a thread, the last post would be, "Got one of those new driverless 220 volt chips today, think I'll try it out".

I am curious though, I wonder would they be driven at 50% by 110 volts AC? Also, what would the effect of the 60 cycle flicker be on PAR? They would in effect be at low or no power 60 times a second as line voltage dropped to zero then rose, there's no capacitive reserve to mitigate this that I can see.

Kinda looks like something for those with big balls and small brains, hope they use a lot of copper grounding everything near the thing. If your working with 220 AC you'll definitely want a spec sheet written in clear ENGLISH from a reliable source. UL labs would cry at the site of this thing and the EU might put you in jail for selling it. For 220 volts AC, the line and neutral contacts look real close and I don't see an isolation gap in the PCB between them either, no ground contact on the chip that I can see...

Hydrostatic Testing Update and simplification
I was going to start in on testing the tubes today, but a dental appointment and tooth extraction kinda took the steam outta me for a bit. Just as well, because I've been a moron, fucking with compressors, gauges and fittings, what an idiot!

Problem:
Hydrostatically test a rectangular water cooling tube with two fittings at each end on the top, to one atmosphere pressure at sea level (where I am) for 5 minutes with minimal equipment and expense. Preferably, do it in a way that someone living in an apartment with minimal resources could pull off too.

Solution:
Supplies required:
1 Regular hand bicycle pump (no gauge required)
1 Schrader valve (tire valve)
1 Foot of vinyl hose the size of the the fittings used (3/8 or 1/2")
3 Small hose clamps
1 Nylon wire tie
1 Ruler (or eyeball it)
1 Sharpie type black marker.

Procedure:
1 Cut two clear vinyl hoses 6" long, one will be the supply line from the bicycle pump, put a schrader valve in one end of this hose, slather it with silicone and shove it in then put a hose clamp on it.

2 Take the second hose and bend it back over to make a kink about an inch from the end of the hose. Loop a wire tie over the bent back end and pull it up tight to seal one end of the tube. Alternatively you could put a plug in the hose and hose clamp it on, but the nylon wire tie is quicker, easier and probably safer. You can test this hose by putting the sealed end in a cup of water and blowing on the other end to check for leak bubbles. This is going to be the pressure gauge, so it's accuracy depends on yours.

3 Lay the aluminum tube to be tested flat on a level surface with the fittings up and pour in water until it comes out the other fitting, lift one end and rotate slightly in both directions to purge any air from the tube, then use more water and fill it to the very top of both fittings. Try to remove as much air from the tube as you can.

4 Put your supply hose on one end and the gauge hose on the other using hose clamps.

5 Measure the gauge hose with a ruler from the top of the water in the fitting to the kinked end, the tube should be standing straight up as well. Now place a mark at the halfway point on the tube, this is your pressure gauge. This will represent 1 ATM or 14.7 psi (approximately).

6 Start pumping the bike pump and put some pressure in the supply line. As you pump watch the water level in the gauge tube rise to the mark and stop pumping, now the tube is at, or close to 14.7 psi or one atmosphere. Now, time the test for five minutes and see if the water in the gauge tube falls, if it does, find out why.

If your tubes don't spring a leak with this test, they will be plenty water tight. Simple, cheap, effective and accurate enough for our purposes, if the air volume in the gauge tube is cut in half, then the pressure is doubled.

Why didn't I think of this before! I'm still gonna test to failure, so a gauge would go on the tube for that, if a water tube gauge won't do the job. A little more air in the system, but not a great volume, a few strokes of the pump should get you in the operating range and in the event of a failure you'll be safe. We're only working with 15 psi max and testing hydrostatically.

Note: A longer gauge tube will be more accurate than a shorter one, but will introduce more air and potential energy into the system. A short 4 or 5 inch high gauge is plenty for what we're doing and it will stand up straight.

I'm a fine one to be talk'n bout Jimbob and Cletus, Doh!

DIY-HP-LED said:

Yep, tape like this
CPU tape or pads have a thermal conductivity of 1.5W/m.K equal to that of regular thermal paste,good quality paste is over 7W/m.K

Click to expand...

Wait, I think I'm confused, which is not necessarily indicative of your post, I must admit. Are you saying you're using that tape without anything between the chip and the metal surface, while comparing the tape to paste? Is it made to transfer the heat without paste/pad? I assumed the tape was just meant to hold the chip down, not to be the main means of transferring heat.

Thank's for the interest and question. Follow the link "like this", look down at the specs for thermal conductivity and adhesion. If you ever tried to pull a chip off that this stuff is used on, you'll have no concerns about it staying on! My test system is doing just fine with cheaper heat sink tape holding on 4 cheapo 100 watt LEDs. (look on page 8 of this thread). The tape shown in the ebay link is rated for CPU's and high power LEDs.

With the power of water cooling you might even be able to use regular double sided tape, but I wouldn't.

No heatsink paste required, though if you were using a high end COB with a holder, epoxy down the holder and use a good quality thermal paste on the back of the COB. LEDs that don't use a holder, just use tape. Remember the water cooled aluminum bar they will be attached to is going to stay at ambient temps and the thermal transfer rate will be tremendous.

Perhaps I should make a section on mounting LEDs and COBs to water cooled bars, testing and describing different methods. If get my "K" type thermocouple probe working on my digital multimeter, I can get some chip back plate readings and compare them to the adjacent aluminum tube. Then we can get some real world data, I'd try it with one of the cheapo 100 watt LEDs that's a real pig in terms of efficiency, lots of heat. When I put my finger on the edge of one of these chips on the test rig it feels cool, like the aluminum tube it's adhered to. The LED is stuck on with cheap double sided heatsink tape (not CPU or LED rated) that would fry yer COB on an air cooled fixture.

emepher said:

Wait, I think I'm confused, which is not necessarily indicative of your post, I must admit. Are you saying you're using that tape without anything between the chip and the metal surface, while comparing the tape to paste? Is it made to transfer the heat without paste/pad? I assumed the tape was just meant to hold the chip down, not to be the main means of transferring heat.

Click to expand...

Guess I should have hit the reply button, Doh!

I managed to finish the rest of the tubes and hydrostatically test them, other than putting the tubes on a frame and hooking up hoses, the water cooling part of the lamp is done. In terms of hours spent on the tubes, about 3 hours max on actual construction, fucking around with hydrostatic testing, another matter. It was time well spent though, because now I can easily test all my tubes to 1 ATM for the upcoming builds, testing tubes to failure (other than some of the 3"x1" tubes in the upcoming build) and building the failure test tube etc can wait until later, I need to start in on construction soon.

Anyway here are the completed tubes and the test rig used, all tubes passed the 1 ATM 5 minute test with flying colors. The workbench is a mess and I'm gonna have to clean up and organize before I can find anything!


A close up of the test "apparatus", you don't need a gauge on the bicycle pump, just a gauge tube (mine could have been a bit longer).


I've just started a harvest and at a 1/2 lb a plant, three will give me sticky fingers for a bit, but when I get tired of trimming bud, I can always stop by the work bench for a little bit at least.

Keeping DIY and Grow Costs Down
Everybody likes to save money and sometimes quality isn't required or even desired, in today's world many things are disposable. There are many ways to save money on a build and grow, cruising yard sales, flea markets and charity second hand stores are among the best. You'll find totes, buckets, pumps, fans, baskets, tools and many other things you can use in your grow or build. Dollar stores are another place where you can equip yourself at low cost, for things like weed barrier and round laundry baskets that make better air pruning pots (smart pots) than you can buy, for a couple of bucks each. Measuring cups and spoons are required for mixing nutrients and other gardening stuff (in season). I've seen drip irrigation systems there for a couple of bucks, in fact, I have one downstairs that I might use as is, or for parts one day.

Many hardware stores sell good quality coco coir blocks for $5 -$10 that can fill three weed barrier lined 1 cu ft plastic milk cartons when expanded with a few gallons of water and mixed with about 20-30% perlite.

Scrap yards are another source of useful metal and other things, I've scored over 30 free (I stripped them and returned the carcasses) LED street lights at the scrapyard, many brand new and about 70% fully working. For the complete story and pictures of lights and builds using them, go back a few pages in this thread. I'm still using the hundreds of high quality stainless steel screws of various sizes in my projects, saved a fortune. Of course I still have a ton of drivers, LED panels and other useful parts kicking around from this adventure, all of very high quality.

Here are a few items I picked up at the local dollar store, when I was a young kid they were called five and dimes, and I ain't that old, despite appearances.
Items include: Hose clamps, screw hooks, bolts, washers, machine screw and nut assortment, blue duct tape and a 1/2" spade bit drill. The drill bit can be used to drill holes for fittings in aluminum tubes, using a drill with a 3/8" chuck (most folks have one). All of these items are a fraction of the price they would be at the hardware store.


TIP: When you go into a hardware store not sure about fittings and hoses, etc, look for the most grizzled old fart you can find, a guy with a lot of miles on him. Often he can be found with a crowd of people around him and younger employees interrupt frequently with questions, you have arrived at the master's knee. Pay attention, ask questions and be polite, you'll learn a lot about what you need to know, respect experience.

I've kinda figured out what I'm going to do about this thread, water cooling and other ideas I have for some how to projects with detailed information, kinda like an instruction manual. I'm gonna develop the idea and write up the idea here, then cut and paste a few solid pages of info instructions and pictures, in a more focused and coherent way, than they are here. When the grow room conversion is complete and I've had a chance to get a flowering cycle out of the new configuration and add a few more things, then it will go into a new thread. I hope to have a few pages of solid info to cut, past and edit for screw ups from this thread, to start off a new DIY water cooling thread. Same for a low cost apartment stealth closet or cabinet liquid cooled grow project.

This thread is my home, when worthy ideas are fully developed into a how to, they will be posted on a new thread with a relevant title. Each new thread will have several pages of how to info at the top, before questions and discussions begin. Perhaps someone else will pick up on this and run with it and post better ideas and projects. The ideas are the important thing, if it's good it will spread, if not...