LED-WATER-HEATSINK

[Weezard]

Major malfunctions in elktown!

I just returned from a 5-day trip and the "pipe-light" is out... I checked the drivers, but they're ok. The LEDs glow a tiny amount when powered.

I also checked the thermal switch/pump circuit and they work fine as well...

Next thing I guess is to test the LED-string and exclude one led after another until I find the ones that are burnt out or whatever.

My girlfriend has been watching the place and she casually mentioned something about the fire-alarm (on top of the cab) going off this morning... WHAT!?

I hope the circuit stops when one LED fries and the rest are good.

Need some help on this one!

Plants are fine, btw. Pulled a male, so down to 4.

I doubt if it's the leds.
There would be no glow at all if one is open.
If one shorts though, it will drop resistance in that string and overwork the driver.

About the only way you can have a low glow is for one led to pop and leak current, essentially, become a resistor.
That would be very rare.

The smoke alarm beep suggests that a driver is hosed.
When a driver dies, it's not unusual to show some leakage current.
So, if all leds glow with the same intensity, or lack thereof, it's a fried driver.

All decent drivers have a thermal shutdown as well.
Have you let it cool down and tried to restart?

If it is the driver.
Thermal glue the new driver to a heatsink to prevent recurrence.
That and overspec it to run it cooler.

An innovator's path is strewn with hazards.
Dragons to be slain.
Tilting windmills is great fun.
But does involve some pain.

Aloha,
Weezad

[LedElk]

Good call, weez. I ran off to try the drivers again and now it's back to normal!

I have no idea why the drivers seemed to work when I wrote the last post.. Might have tried the same driver twice for all I know:stoned:

I've still got thermal adhesive and aluminium. I hope it'll do the trick without a fan. It's been running for days now with no issues until this..

anyway, the red blaze is back

Cheers, mate!

[LedElk]

Problems are back...

I waited for the lights to turn on at 12:00 but only the purple light lit.

So I checked the driver for the red light on the purple light and vice versa. Both drivers work on the purple light, but no drivers work on the red light.

How can this be?

It still glows dim when I switch it on.


The Leds on the red lamp:

3x10w 625nm @ 9 V
2x10w 660nm @ 10,5 V

The driver is meanwell LPC-35-700

700mA and 9-48V

When I bought the driver I thought it was 9-50V, so with a total voltage of 48 on the LEDs I thought I was within range.

Could this be the problem?

Why would it stop working NOW? And with both drivers...?

[LedElk]

I think I found the problem. The driver-to-LED cord was simply too long to carry enough power!

I shortened it from about 5 feet to 2 and things are back to normal.

[Weezard]

That sounds dicey, brah.
Unless you were using smaller than 20ga. wire, it's not the whole problem.

Grab you meter and measure across each emitter.
It sounds like the regulator is dropping out for lack of headroom.
That 48V. rating on the regulator is the maximum voltage, not the operating voltage.
It needs a couple volts of headroom to do it's thing.

You are asking it to regulate current without giving it enough "wiggle-room" do so.
And, the voltage drop across the junctions will creep as your emitters "burn in" for the first few months.
It's only a tiny amount but it is multiplied by the number of emitters.

LEDs are non-linear devices.
A very small change in juction voltage creates a relatively large change in current, yah?
So, this will get worse with time.
If you operate this puppy on the ragged edge of functionality, you are "standing into danger".:microwave:

Since these are current regulators, there's a simple test as well.
With the original supply cable, and the unit mis-functioning, just turn it off and clip a jumper across one emitter.
If she comes right back on line through several on-off cycles you have identified the problem.
You will need to get a third regulator and add that 5th led to a new string of reds.
If it were mine I'd do 4 X 625nm. for 36V.
And a new string for 4 X 660nm. for 42 V.
The regulators will then be re liable and will last a heck of a lot longer too.

Bottom line for good design?
Run cool and leave plenty headroom.

Aloha,
Weeze

[LedElk]

So you're saying I should get more light?

Hm, guess I just bought myself some time with the shortening of the cord then... There is an LED store in town actually. Probably alot more expensive than mouser, but I'll pop in and see if they've got drivers and leds or just commercial led bulbs.

As I stated previously I thought the max V was 50. But I guess I was pushing it, anyways. well, lesson learned. I've got room for one more LED on the pipe-rig and making another heatsink for a couple of LEDs should not be a problem. I'll knock down the veg-chamber and expand sideways.

The plants are loving the glow btw! It doesn't seem to be a problem with the lights being too close, at all. I've got room to raise them, too. But here I go pushing the limits again...

Well, thanks for the info, weezard!

I'll be back with an update

[Weezard]

Raise the bridge?
Or lower the water?

Jumping out one red led would work too, without added expense.
And would still be enough light for that small vegging cabinet.
I just figured that you will be flowering with this light and could make use of the extra photons.

Something to keep in mind.
It's very easy to overwater with leds.
Less heat translates to lower transpiration.
They sweat less, they need less.

Just thought I'd mention that because I messed up my first 'lings under leds.
Tried to treat them like HPS'ers.
Drowned them, I did.
It was not pretty.

Aloha,
Weeze

[LedElk]

I am actually flowering almost from the get go. This is really more of a test run as I'm new to growing and there's just enough time before going on vacation to run a quick scramble.

So what you're saying about removing an led seems a little tempting, actually. And go for more light later.. I'll be doing lots of reading on growing styles and mediums for next time as well.

As for drivers, I was looking around and mouser seems to have only two 700mA drivers.. Could I use a 1000mA?

Attachment 291260

[LedElk]

I tried editing this into my last post but the time limit beat me, once again:thumbsup:

About the water, I don't have any experience with HPS, but I did notice a drop in evaporation going from CFL to LED. Probably not as much, tho. I've adjusted to a relaxed watering schedule

[Weezard]

Depends on the leds.
If 1 amp is I-Max, then no!

Hmm, wait a monkey!
You have water cooling.
That's a game changer.
Might actually get away with it, but efficiency will suffer.

As you increase the current to near maximum, efficacy drops off rapidly.
It translates to less actual photons per watt.

That, and it's a calculated risk.
LEDs can become slag in microseconds.
Once that magic smoke gets out, C'est finis, yah?

Your call.
But I'd go with more photons every time.

This is almost as much fun as building another array. :D

Mahalo,
Weeze