Stealthy LED Project - help wanted

[Weezard]

"Hey, Mac...Weez...
Double check me on this rant, willya?
Don't wanna make a typo or procedural error that might cause somebody problems... "


Spot on!
Thorough.
And quite clear!
Can't add or correct a thing, darn it!
I'm gettin to be a third wheel.
Actually, that's a good thing.
(Did I mention dat I'm lazy)

Weeze

 

[Weezard]

Howdy Gib?

Howdy Gib?

Peekaboo bump

 

[Oldmac]

First, Giblets, meant to chime in the other day that you made perfect sence (cents?), I even understood it all....what happened?

Weezard made me think, when I asked about 15w vs 10w 660nm diodes, that 15 watters would be the way to go.....so I was looking at Mouser Dist site yesterday and guess what, only 48 of them left and, they are now obsolete! Looked at LedEngin's site and sure enough they thave dropped them. My next light will be 10w based I guess.

Yo Jopedijoo, if you haven't bought your 15w 660 LEDs you better order them soon, while they have them. Already not enough of them for what I had planned, so I'll go 10w but it may work out better for me in the long run.

 

[giblets]

Weez, it's yer move.

Mac, if you need that many 15W., ledengin would probably do a run for you. The 15W. deeps were never part of their regular production, they'd whip up batches of 100 at a time for vendors. That's the advantage of dealing with a relatively small company.

Worst case scenario, the 5W deeps are still around. (The 10W are too expensive.) It's not like they're d/c'ing the emitter itself, or that flex-pcb. They're interested in capturing a share of the horticultural market, that's why they recently released the 5W far reds.

I'll check on it. I got people.

 

[Jopedijoo]

It's Christmas already..?

It's Christmas already..?

Hi guys! Sorry for my slow moves here. Hope you didn't think they already had me ... Seriously, I've been real busy with things for the past week but now I'm off duty from the real world and can hopefully get this project finally rollin'!

I've kept the mail man busy lately... I got the whisper fan that I mentioned about. Didn't get it powered yet, but I'm on it.. Sure hope it whispers softly! Today I got the two part Arctic Silver epoxy and the Mastech replica is in the mail... I feel it's Christmas already

Now that I've had a few days to think what's been said here I suppose it's best to follow your advice and try to make a light work first with just the reds and the blues before adding anything. And do that properly by using the Mastech.. I've got 6 reds and 2 blues. The Mastech can hopefully do the reds. Could drive the two blues with a Meanwell PLN-30 for example, but if I were to drop the extra blue I'm still not quite sure how to drive just one blue with the Meanwells that do exist, or with something else..? The PLN-20 is coming but it's not available yet, hmm... One trick would be to add two more reds and use both blues making it 4:1 R/B. (Not sure if driving 8 reds with the Mastech is wise though..) As both you Weezard and Giblets have mentioned that you have bleached your plants with too intense light I'm not sure if sticking that much watts (150) into my 3,2 square feet cab will do any good..?

Now I just have to wait for the power supply to arrive and deside what to epoxy to the heat sink... Got any handy epoxing tips?

Replys coming right up in a sec...

ps. It's real nice to come back here with great posts waiting from you three Weezard, Giblets and Old Mac! Thanks a bunch again for helping out here guys!

 

[Jopedijoo]

....I had a few more cents to throw into the color discussion.

As Weezard and Giblets point out, and rightfully so, it only takes red and blue to make a good LED grow light. Best example of that is the Procyon100, which works remarkedly well considering it's use of 635nm Cree diodes. I'm no expert when it comes to white LEDs, but have been interested in using them for awhile. I actually believe with the "right" white LEDs you could use them IN PLACE of blues. If the white's blue spike was right at 435nm you could accomplish the blue you need and MAYBE pick up some lessor needed wavelengths. I'm not sure whites along with blues are going to add anything and Weez's point that the wattage given over to them might be better used on the blue is probably spot on.

That's very interesting. I haven't really been looking for white emitters peaking at 435nm so I don't know how that is but the closest to that from Ledengin would be the warm white peaking at around 445nm. It would be great to see how it would work if one replaced the blues with whites!

I noted you made reference to far red 730nm light. (I like a person who does research) LEDs in this wavelenght are available and you could use them in a grow light if you wanted to stay all LED. But you can get a metric ton of far red from an incandescent for a lot cheaper so why? UVb is also a lessor needed wavelength, but again can be had cheaply with CFL or flourescent supplemental lighting. The TI SmartLamp that I mentioned has a mix of 150 diodes that are all 2w nominal, the mix is 105 red, 18 blue 15 white and 12 "proprietary". I'm really not sure what the whites bring to the party but someone put some time and effort into this light so I guess they are doing something. BTW this is the only LED light I've seen that can turn trichomes amber. This is due to the "proprietary" LEDs; UVb and IR are the "secret".

For a home brewed light, where simple circuts are in order, red and blue are the ticket to success. And keep in mind that having adjustable red intensity is really not just a trick, it's a great idea! Any other wavelenghts of light needed can be added via supplemental lighting very simply.

You are right about being able to use an incandescent and CFL or flourescent for uv and far red, I'm sure. Maybe I was thinking that by using UV and IR LEDs you could be sure it's those that do the trick and not something else the incandescent and the CFL would be emitting.. But that might just be needlessly HiFi. Adjustable reds sure sound good! I suppose you could mimic day light changes with that and... I think I don't really know what I'm talking about here Have to try it out myself.. Cheers Mac!

 

[Jopedijoo]

lll

Give this man a trophy! That was a great chimp proof instuction manual for the Meanwells, thank you so much! I don't think the Meanwell guys could do a better job themselves..

Both. I added 2 more reds to that light, and with both blues and 8 reds the dang thing was so powerful I was bleaching leaves and stopping growth from 6 inches away. To accommodate a slow-finishing sativa I disconnected one blue and turned down the meanwell that was driving 4 of the 8 reds. I'll hook that blue back up next time I flip a new plant into flower, though--to keep stretch under control.

If I had it to do again, I'd only use one blue per light, or at least separate the two blues as far as possible on the lamp, like Weez did. Or use 5W stars for blues. For a flowering lamp, I think they'd be plenty.

How exactly did you disconnect the other blue? Were you driving both blues with PLN-60-15? ..and then disconnected the other blue and drove just the other one with..? Or were you driving one blue for two lights from the same driver? I'm sorry but I don't follow you here..

 

[Jopedijoo]

Weez, it's yer move.

You guys playing chess..?

 

[giblets]

That's very interesting. I haven't really been looking for white emitters peaking at 435nm

That's about what a Cree Royal Blue does, isn't it? I guess they work okay for veg (I never tried it), but I was unimpressed with them for flower.

You are right about being able to use an incandescent and CFL or flourescent for uv and far red, I'm sure.

Before you get too excited about far red, you should take a look at this:

http://74.125.113.132/search?q=cache:lMx5yihPG-MJlasticulture.cas.psu.edu/DSuccess-mulch.htm+far+red+visible+human&cd=10&hl=en&ct=clnk&gl=us


Maybe I was thinking that by using UV and IR LEDs you could be sure it's those that do the trick and not something else the incandescent and the CFL would be emitting..

That would be the way to do it for controlled experiments, varying a single parameter at a time. If you use 5W stars, you can easily swap out different wavelengths if you mount with screws and thermal grease.

lll

 

[Jopedijoo]

Almost ready for the real deal...

Almost ready for the real deal...

The Master Mastech has arrived! Well, it's not really Mastech, it's a replica of some sort but is anyways a single output 30VDC/0-10A adjustable DC power supply. It's sweet Thanks for making me get it

My hands are itching to get to work, but it might still be a couple of days. I went today to my local Babylon Department Store to get the missing ingredients but they didn't have everything I wanted and I don't think I have time to get to a real store before Christmas, so... I got test leads and some crocodile clips but couldn't get proper gauge wire. All they had was 24 gauge which I left on the shelf. I should've saved my time and gone elsewhere...

I'm going to borrow a soldering iron from a friend but that's not gonna happen before Christmas either. I was also going to borrow a volt meter from someone but that turned out to be for crude measuring only. If I understood correctly I'll need a meter that will show me voltage with a one decimal accuracy, is that correct? I have to get a better meter next week...

By the way, about the LED testing, how's that again? Do I need to stick a meter in between the Mastech and the LED(s) or does the Mastech provide all the info needed. So again, once the Mastech is connected to a single LED I turn the voltage up by using the C.V. so that I get a reading for 1200mA, correct? Does current have to be turned all the way up or how was that? Do the same for each LED. That's for the reds. And for the blues you can go to almost 15V and it'll say about 1000-1100mA, yes/no?

So I guess there's nothing else to be done but to chill for a couple of days and get the soldering equipment next week. I suppose it's not worth the trouble to try to connect the test leads to the soldering pads on the emitters for LED testing? It's probably much easier with wires attached...

Your help has been invaluable! Merry Christmas to y'all! And Weezard, no bombers for Santa before Christmas this year

 

[giblets]

How exactly did you disconnect the other blue? Were you driving both blues with PLN-60-15? ..and then disconnected the other blue and drove just the other one with..? Or were you driving one blue for two lights from the same driver? I'm sorry but I don't follow you here..

My fault, not clear.

I built 2 lamps, the 8+2 (which started as the 6+2, and I added 2 extra reds trying to balance out the 2 blues) and a 7+1.

I was powering all 3 blues with a PLN60-15. The 2 blues on the 8+2 were too strong compared to the single blue on the 7+1. Frequently I had different strains in different stages of growth, so no problem-- I just put the strong blues over the plants I wanted to stretch least.

But to even them out, I disconnected one of the blues on the 8+2, and ran just 2 blues, one on each lamp, from the PLN60-15. I have had no problems doing this, even though the load is so "easy" for the meanwell that the current limiter isn't involved. (If this issue concerns you, you could always put an appropriate value resistor in the circuit and dissipate the extra energy as heat. This would allow you to turn the meanwell UP enough to engage the current limiter, at the expense of wasted electricity and excess heat. I forfeited the 'safety' of full-time CC for efficiency.)

Once I get the leads soldered to the emitters I group them as desired and just connect with wire nuts. That way I can easily re-group them if I change power supplies, etc.

To disconnect a blue, I just loosen a wire nut and ease out the "hot" lead to the led. If I did this frequently I'd probably add a switch (which should only be operated with the power off and the power supply discharged so as not to "spike" the leds) into the lead. As rarely as I make this kind of adjustment I opted for the more idiot-proof option of manually removing a wire.

Whoops, I see another post...

 

[Jopedijoo]

That would be the way to do it for controlled experiments, varying a single parameter at a time. If you use 5W stars, you can easily swap out different wavelengths if you mount with screws and thermal grease.

There's in an idea! Once you have the necessary blues and reds tight you could swap the other colors so that it wouldn't get too crowded on the heatsink

Royal Blue Cree® XR-E Stars seem to be 450nm, here for example: http://www.ledsupply.com/creexre-rb.php

 

[Jopedijoo]

My fault, not clear.

Is clear now, thanks for the clarification! Don't know about resistors though, but I guess that's a possibility, have to think about it... I wish they had those PLN-20s already

Once I get the leads soldered to the emitters I group them as desired and just connect with wire nuts. That way I can easily re-group them if I change power supplies, etc.

To disconnect a blue, I just loosen a wire nut and ease out the "hot" lead to the led. If I did this frequently I'd probably add a switch (which should only be operated with the power off and the power supply discharged so as not to "spike" the leds) into the lead. As rarely as I make this kind of adjustment I opted for the more idiot-proof option of manually removing a wire.

I think this sorta answers some of the questions I posted before I saw this post... Thanks! I got some wire nuts also today, forgot to mention about those. Nice and simple.

Whoops, I see another post...

Yea, me too...

 

[giblets]

The Master Mastech has arrived! Well, it's not really Mastech, it's a replica of some sort but is anyways a single output 30VDC/0-10A adjustable DC power supply. It's sweet Thanks for making me get it
They are very handy.

I'm going to borrow a soldering iron from a friend
You will want to use a 20-25 Watt soldering "pencil," something like a Weller WP25. A big "soldering gun" can melt your solder pads right off the PCB. If this happens (I did it a few times before I resolved to use a smaller gauge of wire as a 'leader' to the 14ga supply wire) you can scrape the lacquer-like coating off the flex-pcb over the copper traces, and solder directly to the traces, bypassing the pad altogether. This is fairly delicate work and you will definitely need a small, low-wattage tip, and small wire, to make this repair.

My advice is to get the proper (small) soldering equipment to start out with. You'll need it to repair what a big iron can melt, so save yourself some trouble.

A thermostatically controlled "solder station" is best, something with a grounded plug that's "ESD safe" so you don't send a static electricity discharge spike through your leds (like a Weller WESD51, set at ~ 450F, depending on your solder).

And you cannot use acid-core solder. Gotta be flux-core (rosin-core is the same thing) for electrical work. The acid-core stuff's for sweating pipes, etc.

If I understood correctly I'll need a meter that will show me voltage with a one decimal accuracy, is that correct?
I'd recommend at least 2 decimal places just cause it's easier to tell what you're doing. Your bench supply probably reads out to a single decimal place. A "meter" that only reads out to tenths of a volt may not be appropriate for this type of work.

By the way, about the LED testing, how's that again? Do I need to stick a meter in between the Mastech and the LED(s) or does the Mastech provide all the info needed. The bench supply will tell all, but only to one decimal; e.g. 10.8V.

Putting a meter that reads out to an extra decimal in the circuit will be helpful. As you get near max Vf a tenth of a volt is "a lot" to an led. A meter in the circuit also provides a confirmation of correct reading.

So again, once the Mastech is connected to a single LED I turn the voltage up by using the C.V. so that I get a reading for 1200mA, correct? Does current have to be turned all the way up or how was that? Might as well. It's gotta be turned up high enough so the voltage can "push" the current without "interference" from the current limiter. I suppose once you know approximately where that setting is, you could limit the current at that point as kind of a safety net as you test subsequent leds.

Do the same for each LED. That's for the reds. And for the blues you can go to almost 15V and it'll say about 1000-1100mA, yes/no?

Well, it varies. Otherwise, this measuring wouldn't be necessary. I always make a table, recording how much current each led draws across a range of voltages. This way, you can conveniently see how your "blue channel" will react to a variety of voltage settings. Here's the readings from my 3 blues:

Volts.......... 14.7..... 14.8.......14.9.......15.0......... 15.1........ 15.2....... 15.3....... 15.4

Blue #1..... 1.0A........1.0.........1.0........1.1...........1.1.........1.2............1.3..........1.3
Blue #2 .......9A ..........9...........9...........9 ...........1.0..........1.1...........1.1...........1.1
Blue #3 .....1.1A........1.2.......1.2........ 1.3........... 1.3..........1.4.......didn't test higher on this one (that's over 21 watts!)


If you keep your meter in circuit while you test, you can get another decimal place of precision. Since sometimes you must kinda compromise, with one led running a little hard, and one sorta idling, the precision isn't so necessary. You can "even out" the the leds by adding resistors to the circuit, but I don't like the added waste heat. If you have lots of small drivers you could run similar leds on the same circuit for more uniform performance. (Just FYI, the 660 reds seem more consistent.)

If I run my blues 'normally,' I set the PLN60-15 to about 14.75 volts (with the leds in the circuit) which runs them at roughly 1000mA, 900mA, and just under 1200mA respectively. At that V x A, Blue #3 was burning 17-18 watts! (not recommended!)

I ran it at 19.5 watts for awhile (15.03V) and it survived. YMMV!!!! Caveat emptor!

So I guess there's nothing else to be done but to chill for a couple of days and get the soldering equipment next week. I suppose it's not worth the trouble to try to connect the test leads to the soldering pads on the emitters for LED testing? It's probably much easier with wires attached...

Be very careful of "spiking" the leds with a static charge. You shouldn't touch the soldering pads, or any wire connected to them, unneccessarily. Avoid connecting and disconnecting wires as much as possible. A wrist strap, and /or other anti-static measures is recommended.

And, don't run the leds without some kind of heatsink. I epoxy the leds in place on the heatsink first, and test them "in place." The downside is, if they're horribly dissimilar, you're stuck (because they're "stuck"). You could probably clamp them to a heatsink with some thermal grease for testing, but having the leds glued down makes it much easier to get the supply wires soldered on--otherwise that flex-pcb squirms all over. The "solder first, then mount" warnings are for leds on stars, because the stars act as enough of a heatsink to pull the heat away from your soldering attempts. The 15W ledengins on flex-pcbs are altogether different.

HTH, Giblets.

///

 

[Weezard]

Kinda, sorta

Kinda, sorta

You guys playing chess..?

See the,
"Calling out to dreaded hermie for a game of stoner chess" thread
over on Cannabis dot com.

In which, we get ripped and talk tech over a very odd chess game.

Old Mac says he learns more about what we're up to there than in our diy thread.
We encourage digressions and kibbitzers.
Come set a spell you might enjoy.

Aloha.
Weeze

 

[Weezard]

This just in

This just in

Food for thought.
http://www.icmag.com/ic/showthread.php?t=147095

Busy thinkin'.

l8r
W.

 

[Jopedijoo]

See the,
"Calling out to dreaded hermie for a game of stoner chess" thread
over on Cannabis dot com.

In which, we get ripped and talk tech over a very odd chess game.

Old Mac says he learns more about what we're up to there than in our diy thread.
We encourage digressions and kibbitzers.
Come set a spell you might enjoy.

Aloha.
Weeze

Well... I got to abmit, I somehow by accident found your chess club thread yesteday and I like it! I wasn't sure if all of it was for my eyes so I was kinda hesitant of mentioning it to you.. But anyhows, event though I just read the last five pages it's a good read, I appreciate it! I have to agree with Old Mac there.. and finding the right places isn't supposed to be too easy, right..? I got to pop by there some day...

It's been a good day altogether.. Just came back from meeting two of my good ol' friends that I haven't seen for five years, it was real nice! Giblets, your reply seems real good and informative, I have to read it mo' betta tomorrow, I might have had a beer too many to decipher it properly now... Same thing for the thread Weezard was pointing out to... Zzz...

 

[giblets]

Hey, folks, I edited my post above. I removed some rambling about working with dangerous-kine power circuits that was irrelevant and probably just confusing.

JJ, you can certainly look at that thread, please do. There's an album in my profile you should see, too. Send me a friend request over there so you can view it. Lights in action.

 

[Jopedijoo]

///

Excellent post about soldering, meters and Mastech control Giblets! You don't leave a lot of room for questions and that's great! Simple and clear Thanks again!

Hey, folks, I edited my post above. I removed some rambling about working with dangerous-kine power circuits that was irrelevant and probably just confusing.

Missed the first version.. Nothing wrong with the second. Often less is more

Is still have to decide what do with those blues. So far the R:B 6:1 sounds probably best but as I was too eggknocked I forgot to wish for Santa to bring me a couple of those PLN-20s and now I'm left with too much blue or resistors, eenie meenie miney moe... Eh, just thinking aloud here, no biggie.. I'll let this simmer for a while, something'll come up.

JJ, you can certainly look at that thread, please do. There's an album in my profile you should see, too. Send me a friend request over there so you can view it. Lights in action.

Well, see ya there...

 

[Jopedijoo]

Been lazy thinkin'...

Been lazy thinkin'...

I thought that I could wait for Meanwell to get that PLN-20 ready so I could power a blue 15 watter with it. But I took another look at the datasheet http://www.meanwell.com/search/pln-20/pln-20-spec.pdf and it seems that there are no screws to addjust either current of voltage. It first says that current adj. range is 75-100%, but then there is a note 6 which notes: "The output current of Constant Current models are adjustable at factory from 75%~100%." Might not be that handy after all...