LED driver limits, power, everything

[flimflam]

Hello everyone,


I am trying to build a grow light for my orchids, but am getting stuck on a few points on the design. The goal of the project is split about 50/50 between helping my plants and learning/fun for my own self, hopefully without burning the house down. I have found some guides to picking a driver and basic info on them, but am utterly new to circuits and have some basic questions that I haven't been able to answer yet. So:


Do I understand right that a constant-current LED driver will vary voltage and resistance to keep current constant through however many LEDs I chose to power? Suppose I have this guy which is a 700mA CC driver, 16.8W and 16-24Vout 90-135Vin. Is it right to think that the driver would be able to run as many LEDs at 700mA as I would like, as long as their combined voltage is less than 24V? Where does the power of the LEDs come in? If the 16.8W spec for the driver refers to P=VI case where voltage is maximized, does this mean that the 16.8W would be distributed over all of the LEDs? Would 10 2.4V LEDs get 1.68W each? Some LEDs are advertised by wattage - would 1W LEDs in this case burn out/explode? What do you do in a case where wattage isn't advertised? For example the Philips Luxeon Z datasheet doesn't seem to have anything about power limits. How would I go about connecting the driver to US mains if I wanted to be able to plug the light in to a two- or three-prong outlet?


Can someone explain why the person in this video has an LED driver in addition to an LED power supply?


Sorry for the flood of questions. It seems like everything is tangled up with everything else, so it's hard for me to narrow the confusion down to a single point. I'll be happy to accept any little help I can get, and thanks very very much in advance.

 

[Murphy625]

Do I understand right that a constant-current LED driver will vary voltage and resistance to keep current constant through however many LEDs I chose to power? Suppose I have this guy which is a 700mA CC driver, 16.8W and 16-24Vout 90-135Vin. Is it right to think that the driver would be able to run as many LEDs at 700mA as I would like, as long as their combined voltage is less than 24V?

YES

Where does the power of the LEDs come in? If the 16.8W spec for the driver refers to P=VI case where voltage is maximized, does this mean that the 16.8W would be distributed over all of the LEDs? Would 10 2.4V LEDs get 1.68W each?

Led power (P) = The voltage drop across the LED (V) * The current passing through the LED (I)

Lets use some round numbers for an example.. Lets say you have 4 LED's.. each is rated for 10 volts and 1 amp.. so each led is theoretically 10 watts..
You hook up all four in series with a constant current power supply that delivers up to 48 volts at 1 amp..
The power supply will adjust the voltage to 40 volts because that is what it will theoretically take to "push" the 1 amp of current through.. Current will remain the same at 1 amp through all four LED's. In any series circuit, current is always the same.

1st LED: 40 volts @ 1amp goes in. >> 30 volts out
2nd LED: 30 volts @ 1amp goes in >> 20 volts out
3rd LED: 20 volts @ 1amp goes in >> 10 volts out
4th LED: 10 volts @ 1amp goes in >> 0 volts out and the current returns to the power supply negative.

Hope that clears some of your questions up....

 

[Murphy625]

Can someone explain why the person in this video has an LED driver in addition to an LED power supply?

Looks to me like he's using the large supply as the driver for the LED's.. I think that one is Constant Current..
The smaller white one says constant voltage so its probably being used to power the fan(s)..

 

[RetroTechie]

Welcome here, flimflam! :wave:

For example the Philips Luxeon Z datasheet doesn't seem to have anything about power limits.

Just read carefully... Datasheet says max. forward current for the white LED is 700 mA, and that it has a typical Vf of 2.95V at that current. That's 2.95V * 0.7A = approx. 2W maximum power. Which is pretty high for such a small part imho - probably better not to push it that far. Btw such limits always assume sufficient cooling. Without that, you'll just overheat & kill the LED. Keep in mind that cooling is an integral part of any design/build (if not, you're doing it wrong ;) ).

As a grow light, I'd do some Googling as to what's good colors to use. Most plants are green, which means their leaves absorb all colors except green. So (for example) a mix of red + royal blue LEDs might give better results than white LEDs. I'm sure there's more knowledge on that outside this forum.

Hehe... orchids, right? :devil: :naughty:

 

[Murphy625]

As a grow light, I'd do some Googling as to what's good colors to use. Most plants are green, which means their leaves absorb all colors except green. So (for example) a mix of red + royal blue LEDs might give better results than white LEDs. I'm sure there's more knowledge on that outside this forum.

Hehe... orchids, right? :devil: :naughty:

I pulled this info off a pot-head website.. Should be a good starting point.


Clones / Seedlings

Blue Range 400-499nm .Peaks pref. @ ~ 410 nm , ~430 nm & ~453 nm : 20-25%
Green Range 500-550 nm .Peaks pref. @ ~ 550 nm : 20-25 %
Yellow-Amber Range 551-599 nm .Peaks pref. @ ~ 599 nm : 5-10%
Red Range 600-639 nm .Peaks pref. @ ~ 625 nm : 15-20%
Deep Red Range 640-670 nm .Peaks pref. @ ~642 nm & 662 nm : 5-10 %
Far Red Range 671-750 nm .Peaks pref. @ ~680 nm ,700 nm & ~730 nmm : 5-10 %


Young Plants

Blue Range 400-499nm .Peaks pref. @ ~ 410 nm , ~430 nm & ~453 nm : 10-15%
Green Range 500-550 nm .Peaks pref. @ ~ 550 nm : 10-20 %
Yellow-Amber Range 551-599 nm .Peaks pref. @ ~ 599 nm : 25-30%
Red Range 600-639 nm .Peaks pref. @ ~ 625 nm : 30-40%
Deep Red Range 640-670 nm .Peaks pref. @ ~642 nm & 662 nm : 10-15 %
Far Red Range 671-750 nm .Peaks pref. @ ~680 nm ,700 nm & ~730 nmm : 3-5%



Early Flowering

Blue Range 400-499nm .Peaks pref. @ ~ 410 nm , ~430 nm & ~453 nm : 8-10 %
Green Range 500-550 nm .Peaks pref. @ ~ 550 nm : 10-20 %
Yellow-Amber Range 551-599 nm .Peaks pref. @ ~ 599 nm : 10-20%
Red Range 600-639 nm .Peaks pref. @ ~ 625 nm : 25-30 %
Deep Red Range 640-670 nm .Peaks pref. @ ~642 nm & 662 nm : 25-35 %
Far Red Range 671-750 nm .Peaks pref. @ ~680 nm ,700 nm & ~730 nmm : 3-5%


Middle Flowering

Blue Range 400-499nm .Peaks pref. @ ~ 410 nm , ~430 nm & ~453 nm : 8-10 %
Green Range 500-550 nm .Peaks pref. @ ~ 550 nm : 1-15 %
Yellow-Amber Range 551-599 nm .Peaks pref. @ ~ 599 nm : 5-10%
Red Range 600-639 nm .Peaks pref. @ ~ 625 nm : 25-30 %
Deep Red Range 640-670 nm .Peaks pref. @ ~642 nm & 662 nm : 30-55 %
Far Red Range 671-750 nm .Peaks pref. @ ~680 nm ,700 nm & ~730 nmm : 3-5%


Late Flowering

Blue Range 400-499nm .Peaks pref. @ ~ 410 nm , ~430 nm & ~453 nm : 1-8 %
Green Range 500-550 nm .Peaks pref. @ ~ 550 nm : 1-10 %
Yellow-Amber Range 551-599 nm .Peaks pref. @ ~ 599 nm : 1-10%
Red Range 600-639 nm .Peaks pref. @ ~ 625 nm : 25-40 %
Deep Red Range 640-670 nm .Peaks pref. @ ~642 nm & 662 nm : 35-70 %
Far Red Range 671-750 nm .Peaks pref. @ ~680 nm ,700 nm & ~730 nmm : 3-5%

 

[Genes]

1st LED: 40 volts @ 1amp goes in. >> 30 volts out
2nd LED: 30 volts @ 1amp goes in >> 20 volts out
3rd LED: 20 volts @ 1amp goes in >> 10 volts out
4th LED: 10 volts @ 1amp goes in >> 0 volts out and the current returns to the power supply negative.

At best this is misleading and really is incorrect. There is no "volts out" from a LED. In this example, each LED has 10 volts across it with 1 amp of current. The sum of the voltage drops is 40 volts (4 X 10). The output voltage from the power supply is approximately 40 volts so they all sum to zero.

 

[Murphy625]

This makes no sense at all and is incorrect. There is no "volts out" from a LED. In your example, the output voltage from the power supply will be 40 volts (+ or - to keep the current fixed). As long as the LEDs are identical, there will be 10 volts across each when there is 1 amp through each. 4 X 10 volts = 40 volts, which is equal and opposite of the power supply voltage and the voltages sum to zero.

You must be misinterpreting something or reading it wrong because you just said basically the exact same thing.

 

[Murphy625]

1st LED: 40 volts @ 1amp goes in. >> 30 volts out
2nd LED: 30 volts @ 1amp goes in >> 20 volts out
3rd LED: 20 volts @ 1amp goes in >> 10 volts out
4th LED: 10 volts @ 1amp goes in >> 0 volts out and the current returns to the power supply negative.

At best this is misleading and really is incorrect. There is no "volts out" from a LED. In this example, each LED has 10 volts across it with 1 amp of current. The sum of the voltage drops is 40 volts (4 X 10). The output voltage from the power supply is approximately 40 volts so they all sum to zero.

What do you think the voltage would be if you measured it at the point between the 1st and 2nd diode and the supply neg?

 

[AnAppleSnail]

What do you think the voltage would be if you measured it at the point between the 1st and 2nd diode and the supply neg?

Welcome to circuit analysis, where the volts are made up until you define your ground and nodes! The only absolutes here are the voltage drop across each LED (At a given current). The rest is up to the electrician solving the problem.

If we analyze the LED loop (String of LEDs in series/parallel), the most common definition of 'ground' FOR THE LEDs is "The negative power supply wire." But component-by-component, LED 1 grounds to LED 2, grounds to LED 3...Grounds to the negative power supply wire.

However, even that 'ground' is definitely not GROUND, as in a sixty foot metal pole driven into the earth. It will have some potential from Ground, but I am not sure what the effects will be since I have little sense of what that voltage will be. I suppose we could test by hooking a 350mA driver circuit up to your water pipe and an LED on the negative side, and the LED on the positive side. Maybe it wouldn't make much difference. What are your thoughts? My EE is rusty.

 

[Murphy625]

What are your thoughts? My EE is rusty.

My thoughts are not to confuse the original poster and to get him the info he needs to complete his project.

 

[flimflam]

Thanks everyone for your responses! This is all really helpful. Murphy and Retrotechie in particular, thanks for all the specific answers.

As a grow light, I'd do some Googling as to what's good colors to use. Most plants are green, which means their leaves absorb all colors except green. So (for example) a mix of red + royal blue LEDs might give better results than white LEDs. I'm sure there's more knowledge on that outside this forum.

Hehe... orchids, right? :devil: :naughty:

Nya nyaaa. It actually is orchids that I'm working with, but this brings up another issue. It would be really useful to be able to adjust the output from a fixture so I could dial in the proper output for a particular species. There are lots of dimmable drivers available, but (naively, probably) looking at pictures of them I don't see a knob or anything that suggests dimming function. How is the dimming handled, for example in a driver that is sold as dimmable by PWM? Just to be clear, I understand what PWM is, but I'm confused on whether there needs to be a separate controller to apply the dimming function.

It's not hard to find suggested lumens and this brochure suggested photons/area/day for a couple of plants, but it's hard (for me at least) to figure out what a reasonable output for a red/blue fixture would need to have in order to be roughly equivalent.

The question of color of light is an interesting one. This site says that growers of some plants get best response from white light, and the information Murphy found on the pot-head site hint at complexity. My suspicion is that there are non-chlorophyll pigments responsible for the responses to different wavelengths. Chlorophyll looks like it can absorb a pretty broad spectrum, but AFAIK all the light that gets absorbed there goes in to photosynthesis so it's hard to think that the signal to start flowering/whatever is handled there.

The ultimate goal is to have a dimmable fixture with a series of red/white/blue LEDs for growth (and patriotism), and another with just whites for showing off the plants when I"m tired of looking at purple, with the entire thing cooled by a fan or two. On/off will probably be handled by an external store-bought timer. Does all that sound reasonable so far?

 

[RetroTechie]

Dimming is one way. But since you're likely to use a large # of LEDs, you could also arrange them in groups. Such that you can eg. switch off, 1/3rd red LEDs on, 2/3 red LEDs on, all red LEDs on. Same for blue and green (or whatever colors you choose) independently. Choose the ratio of maximum LED currents well, and a full-on mix will look white. That is: multi-colored when you look at the LEDs, but like white light when you look at objects under it.

If space is tight, you could use RGB LEDs. If not, any mixture of different colored LEDs.

 

[DIWdiver]

Yes, a driver dimmable with PWM almost certainly needs an external source of PWM (although there are a few driver chips around now that can use a resistor to set PWM duty cycle). The AM7135 chip uses external PWM, but a lot of drivers built using them also have the controller on board, giving you 'modes'. The other popular method is to use a potentiometer (supplied by you). If you use the right part, you can put a knob on it.

If you use red, green and blue, you can make white, and even adjust the tint of the white to suit the particular plant you're illuminating. If you do that, I'd use RGB leds, to eliminate the colored halos caused by the shadows being in different places.

 

[duyanhconnect]

However, even that 'ground' is definitely not GROUND, as in a sixty foot metal pole driven into the earth. It will have some potential from Ground, but I am not sure what the effects will be since I have little sense of what that voltage will be. I suppose we could test by hooking a 350mA driver circuit up to your water pipe and an LED on the negative side, and the LED on the positive side. Maybe it wouldn't make much difference. What are your thoughts? My EE is rusty.