LED FAQ) Building and DIY

[Arf]

http://pct.cree.com/dt/index.html

unfortunately not much info for volts in the 30 range, however because they way less work with voltage id assume that the current would be almost 0 before it got that low, they r strings in parallel within the cob so for a total voltage has to be divied up between them all,sorry i can't remember now how the dies r set up in a cxb3590, u will have to google it(i think about 40-60 leds in there)




nice work, so they do drop to 30v at 470mA?





why does it matter about the volts? what r u trying to achieve?

I am trying to set the wattage to 30W per COB on the dimmer. Watts = Amps x Volts

 

[Dion]

I am trying to set the wattage to 30W per COB on the dimmer. Watts = Amps x Volts

well why don't u just use ur multimeter and dim til u get there?

seems more accurate than guessing, especially seen as at dif temps the cobs will pull dif voltage.

id guess that at around 800mA you will have a 30w cob

good luck

 

[Arf]

well why don't u just use ur multimeter and dim til u get there?

seems more accurate than guessing, especially seen as at dif temps the cobs will pull dif voltage.

id guess that at around 800mA you will have a 30w cob

good luck

As I mentioned before, the signal coming out of the driver is giving a bogus reading on the multimeter. It's probably not DC, it will be a PWM signal I suspect. Someone is sending me their spare oscilloscope next week, so I will be able to confirm that. If it is a PWM signal, then it will need to be integrated with some sort of an RC filter to get a constant DC value out of it that the multimeter will understand. It all depends on the frequency and mark space ratio of the waveform from the driver.

 

[CannaBruh]

You're only going to find VA and not true Watts this way.

I'd suggest someone who can, and who has a kill-a-watt, dial in various voltages and currents and see what the true ratings are coming back.

*this might not be easy to setup actually thinking about it further

 

[Arf]

You're only going to find VA and not true Watts this way.

I'd suggest someone who can, and who has a kill-a-watt, dial in various voltages and currents and see what the true ratings are coming back.

*this might not be easy to setup actually thinking about it further

Already tried the Kill-A-Watt, there is no easy way of eliminating the driver losses, and powerfactor to get an exact COB measurement, eg. the Kill-A-Watt can say 30W each COB but what if they are only running at 24W, that's a 20% error I wouldn't notice bye eye, but plants would. That's why I am trying to do it properly.

 

[Dion]

Already tried the Kill-A-Watt, there is no easy way of eliminating the driver losses, and powerfactor to get an exact COB measurement, eg. the Kill-A-Watt can say 30W each COB but what if they are only running at 24W, that's a 20% error I wouldn't notice bye eye, but plants would. That's why I am trying to do it properly.

i dunno what type of equipment you use, I've always been able to get the volts and current off drivers

mine does have a DC option though, and its a really old but really expensive one

 

[Arf]

i dunno what type of equipment you use, I've always been able to get the volts and current off drivers

mine does have a DC option though, and its a really old but really expensive one

Do you understand that dimmer style drivers put out a pulse width modulated waveform at full current to maintain color?

So if you have a 1400mA driver dimmed half way it will still be putting out 1400mA but just not for as long each second. Your eye sees it as constant light but it's not, it's pulsing on and of many times a second. How are you supposed to measure the power with a multimeter then? You need to know the duty cycle of the pulses at any given dimmer setting you are trying to measure.

The Cree data sheet shows a pulsed measurement in the tables, and relative luminous flux curves, but then they use steady-state measurements for the electrical characteristics curve. So for some specs they use a pulsed driver, and for others a stable bench power supply!

 

[wildgrow]

Hey guys. Great thread D. Ive been lurking for awhile. I appreciate that you are talking different setups and #s - a lot of other DIY threads want to stay on track with the specific light being built.

I don't want to derail the 'current' topic, but Im getting ready to order parts and I have a few Qs.

Im going to assume most of you have watched mau5' tutorials because Im going to reference them.

This one, in particular: https://www.youtube.com/watch?v=8fxcgF1FLCM

In his vid showcasing different setups for 3x3 to 4x4 spaces, he says that hes getting 56w out of a 3590 @ 1400mA. Does this mean that 8 3590s on 2 185H-C1400s is pulling 448w from the wall even though the drivers are only 200w each?

He figures the PAR w to be 224 with the equation: 400 x .56 = 224. In this case does the '400' represent the combined drivers wattage?

My space is 36x42. Weird, I know, but its what Ive got. Im also an apt dweller who will eventually move. The future may someday be a 4x4 for me, so I want to build a fixture that can grow if need be.
Originally, I was going to go with 8 3590s on 2 185-1400s because I think I could get away with using that in a 4x4.

Now Im considering 10 3590s on 2 185-1050s. That comes out to 37.8w/cob for a total of 378w at the wall and 230.58 PARw. 36v x 1.05A = 37.8 then 37.8w x 10 =378w then 378w x .61 (3590 @ 1050mA is 61% efficient) = 230.58 PARw.

That's a considerable leap in PAR over the 400w hps that Im currently using, while pulling fewer wall watts.

I know my upfront costs go up quite a bit using 10 cobs instead of 8, but the wall watts I save can be used to run a small room sized dehumidifier. I really cant afford to burn more watts than I am already and this seems like a good way to get better quality light and lower my humidity without changing my monthly electric bill.
And, as I said before, I could use the 10 cob setup in a larger space - esp. if I upgrade the drivers when my $ is less tight (read: when the kids have grown up and moved out).

What do you guys think? Are my #s sound? Do you think this light will work for me?

Thanks for the time.

 

[Arf]

If it were me trying to light a 36" x 42" I would make 9 squares 12" x 14" one COB in the each square, probably with 60degree optics to try and avoid a hot spot in the middle.

 

[Dion]

Do you understand that dimmer style drivers put out a pulse width modulated waveform at full current to maintain color?

So if you have a 1400mA driver dimmed half way it will still be putting out 1400mA but just not for as long each second. Your eye sees it as constant light but it's not, it's pulsing on and of many times a second. How are you supposed to measure the power with a multimeter then? You need to know the duty cycle of the pulses at any given dimmer setting you are trying to measure.

The Cree data sheet shows a pulsed measurement in the tables, and relative luminous flux curves, but then they use steady-state measurements for the electrical characteristics curve. So for some specs they use a pulsed driver, and for others a stable bench power supply!

do YOU understand that u never mentioned u were using PWM dimming, I assumed you were using the internal pot(type A) or an external pot(type B) which is either resistance or 0-10v not PWM


Also I believe u are mistaken, the drivers are CC and the current is lowered to achieve better efficiency- if driven on a PWM then they don't increase in efficiency, just decrease watts consumed

This has been discussed in this thread before as well

If I'm wrong I'm happy to learn, please do correct me

 

[Dion]

Hey guys. Great thread D. Ive been lurking for awhile. I appreciate that you are talking different setups and #s - a lot of other DIY threads want to stay on track with the specific light being built.

I don't want to derail the 'current' topic, but Im getting ready to order parts and I have a few Qs.

Im going to assume most of you have watched mau5' tutorials because Im going to reference them.

This one, in particular: https://www.youtube.com/watch?v=8fxcgF1FLCM

In his vid showcasing different setups for 3x3 to 4x4 spaces, he says that hes getting 56w out of a 3590 @ 1400mA. Does this mean that 8 3590s on 2 185H-C1400s is pulling 448w from the wall even though the drivers are only 200w each?

He figures the PAR w to be 224 with the equation: 400 x .56 = 224. In this case does the '400' represent the combined drivers wattage?

My space is 36x42. Weird, I know, but its what Ive got. Im also an apt dweller who will eventually move. The future may someday be a 4x4 for me, so I want to build a fixture that can grow if need be.
Originally, I was going to go with 8 3590s on 2 185-1400s because I think I could get away with using that in a 4x4.

Now Im considering 10 3590s on 2 185-1050s. That comes out to 37.8w/cob for a total of 378w at the wall and 230.58 PARw. 36v x 1.05A = 37.8 then 37.8w x 10 =378w then 378w x .61 (3590 @ 1050mA is 61% efficient) = 230.58 PARw.

That's a considerable leap in PAR over the 400w hps that Im currently using, while pulling fewer wall watts.

I know my upfront costs go up quite a bit using 10 cobs instead of 8, but the wall watts I save can be used to run a small room sized dehumidifier. I really cant afford to burn more watts than I am already and this seems like a good way to get better quality light and lower my humidity without changing my monthly electric bill.
And, as I said before, I could use the 10 cob setup in a larger space - esp. if I upgrade the drivers when my $ is less tight (read: when the kids have grown up and moved out).

What do you guys think? Are my #s sound? Do you think this light will work for me?

Thanks for the time.

the topic is always diy led. welcome.

As a nomad myself I can relate to your situation.


i think u/he is mistaken about the 56w @1400mA its more like 49 or smith, we just say 50w for simple speak.


400w is the combined watts of the two light systems yes
I'm too sleepy to check ur numbers but they look right

the top bin CD cxb3590 is 56% off at 1400mA, i don't think u need to go lower than that

I'm actually in the process of designing my next garden(as I am moving) and considering a 50x100cm bench/cubboard thing hence I've been crunching a lot of numbers figuring out the best lighting situation.

your space is 90x105cm so pretty big btw


we aim for 700+ppfd for flower

so I've got 998ppfd(accounting for a 10% loss to walls and optics) running 8 at 400w in ur space, that means u can dim them down another 50w or so

and 650 in a 4x4(just a hair low)

seen as the expansion will be some time away i suggest you take the 8 cob setup now and in a few years the tech will be better and cheaper, buying 1-2 better cobs then and a driver to run them will be simple enough

 

[GreenEngineer]

Heya fellas!

Would like some quick input.

3x3x6, Cree 3590 3000K top bin w reflector, MW HLG-240H-C1400A, passive fin cooling, would prefer to run at 700mA for efficiency (25W per cob?) and would also like to throw in a 4000K and possibly 5000K. 225-250W recommended for 3x3? Also would be nice with possibility to ramp up for flowering/possible upgrade of space without investing more.

Thank you in advance! Thankful for any replies

Sincerely,
GE

 

[wildgrow]

Thanks

Thanks

the topic is always diy led. welcome.

i think u/he is mistaken about the 56w @1400mA its more like 49 or smith, we just say 50w for simple speak.

Ive seen a Kill-a-watt in his other vids, so I assume he measured the 56w with that. But that's not important. He goes with 50w too, for simplicities sake. My math says 50.4w

your space is 90x105cm so pretty big btw

Never big enough. Maybe I just smoke too much.

seen as the expansion will be some time away i suggest you take the 8 cob setup now and in a few years the tech will be better and cheaper, buying 1-2 better cobs then and a driver to run them will be simple enough

I thought of that too.


On dimmers - I know they decrease the current being delivered to the cobs, but I didn't know if they decreased the watt consumption of the drivers. Youre saying that they do?

 

[Arf]

On dimmers - I know they decrease the current being delivered to the cobs, but I didn't know if they decreased the watt consumption of the drivers. Youre saying that they do?

They don't decrease the current being delivered to the cobs, they switch the full current and voltage on and off many times a second, just as if someone was standing at the light switch flicking it on and of as fast as they can, except faster.

A benefit of the PWM method is that it enables dimming with minimal color shift in the LED output. According to the Lighting Research Center, dimming causes LEDs to experience a similar shift in spectral power distribution as an incandescent lamp. However, if colored LEDs in an array are used to produce white light, the amount of shift, particularly with red and yellow LEDs, may produce an undesirable effect on the white light that is produced by the system.
Dimming does not result in a loss of efficiency. During dimming, the LEDs are still operated at the same voltage and current as during full light output. In addition, lamp life is not affected by dimming, as is sometimes the case with frequently dimmed fluorescent lighting. Rather, dimming LEDs may lengthen the useful life of LEDs, because dimming can reduce operating temperatures inside the light source.

http://www.ledsmagazine.com/article...hting-systems-introducing-the-led-driver.html

 

[jikko77]

depends what dimming function you are using.
the pwm signal work as mentioned.
on meanwell driver you've a couple of more option.

it's called 3 in 1 diming for a reason.
you can use pwm, a 1-10vdc signal (0-10 vdc as well with some expedient) or even a couple of resistor on a switch.

"LED driver will change its output current based on 3 different input signal/setup
including DC 1-10V, PWM and 100K ohm resistor dimming."

"DC 0-10V:
DC 0-10V is considered as the second generation of 3 in one dimming. PWM and 100K
resistor dimming is still same as first generation. Difference between DC0-10V and DC1-
10V is addressed here.
The maximum level is still 100% in case the dimming signal is given at 10V or open the
dimming wires. However the minimum level for DC0-10V is 5.7% in case the dimmin signal is given at 0.57V.
In case the dimmer is giving lower than 0.57V or user just short
the dimming wires, the LED driver will cut off the output current resulting no light output
in the LED module."

http://www.meanwell.eu/ExclusivePDF/DIMMER-NOTE.pdf
http://www.meanwell.eu/ExclusivePDF/DIMMING-NOTE.pdf (are two different document, called in the same way ...)

On dimmers - I know they decrease the current being delivered to the cobs, but I didn't know if they decreased the watt consumption of the drivers. Youre saying that they do?

maybe i'm not the best one to explain things in english, but yeah the idea bihind the other two option works like that.

by modify the volt on the dim+/- output the cob output is modified as well.
but your not touching the volt delivered to the cob, you are going to limit the current delivered (or at least you are not "touching" it by much).
the theory is still it, W=v*a, so if we decrease (limit) the current we are decreasing the total watt consumption.

 

[Dion]

depends what dimming function you are using.
the pwm signal work as mentioned.
on meanwell driver you've a couple of more option.

it's called 3 in 1 diming for a reason.
you can use pwm, a 1-10vdc signal (0-10 vdc as well with some expedient) or even a couple of resistor on a switch.

"LED driver will change its output current based on 3 different input signal/setup
including DC 1-10V, PWM and 100K ohm resistor dimming."

"DC 0-10V:
DC 0-10V is considered as the second generation of 3 in one dimming. PWM and 100K
resistor dimming is still same as first generation. Difference between DC0-10V and DC1-
10V is addressed here.
The maximum level is still 100% in case the dimming signal is given at 10V or open the
dimming wires. However the minimum level for DC0-10V is 5.7% in case the dimmin signal is given at 0.57V.
In case the dimmer is giving lower than 0.57V or user just short
the dimming wires, the LED driver will cut off the output current resulting no light output
in the LED module."

http://www.meanwell.eu/ExclusivePDF/DIMMER-NOTE.pdf
http://www.meanwell.eu/ExclusivePDF/DIMMING-NOTE.pdf (are two different document, called in the same way ...)



maybe i'm not the best one to explain things in english, but yeah the idea bihind the other two option works like that.

by modify the volt on the dim+/- output the cob output is modified as well.
but your not touching the volt delivered to the cob, you are going to limit the current delivered (or at least you are not "touching" it by much).
the theory is still it, W=v*a, so if we decrease (limit) the current we are decreasing the total watt consumption.

i think that he thinks that they all use PWM and is refering to the CV and mono leds changing color when current is dimmed

as far as I know(and u woudl be one to know) the current is reduced not pulsed, as i siad before if the current was pulsed then we wouldnt increase efficiency

 

[jikko77]

i think that he thinks that they all use PWM and is refering to the CV and mono leds changing color when current is dimmed

as far as I know(and u woudl be one to know) the current is reduced not pulsed, as i siad before if the current was pulsed then we wouldnt increase efficiency

i'm not sure to get what you are asking.

regarding pwm it is pulsed, it is modulated.
it's a square waveform with a cycle on (usually 5v) and a cycle off (0v), what you do is switching between 5 and 0 v in a certain period of time, with a 100% output we have a steady line on the on cycle, while if we put it to 0 we have a steady line on the off cycle.

the example of someone flickering with the on/off switch get a clear picture of how it work.

the "changing" colour led is something we won't see with the cob, nor i've read the data Arf is talking about.
on the datasheet or the cree site, regarding the dimming function or similar stuff the is a suggestion "check with the driver producer).
the only note i've found is on the kind of pwm.




Pulse Width Modulation, or PWM, is a technique for getting analog results with digital means. Digital control is used to create a square wave, a signal switched between on and off. This on-off pattern can simulate voltages in between full on (5 Volts) and off (0 Volts) by changing the portion of the time the signal spends on versus the time that the signal spends off. The duration of "on time" is called the pulse width.

in the image above we can see the pwm wave, this is from 0 to 5 v (is an arduino sample from learning) while meanwell use 0-10v.

http://www.cree.com/~/media/Files/C...Lamp Application Notes/XLampPulsedCurrent.pdf

i think Avid can provide a better explanation and information than what i can, may i suggest to check his thread and project? ( https://www.icmag.com/ic/showthread.php?t=327991 )

 

[Arf]

depends what dimming function you are using.
the pwm signal work as mentioned.
on meanwell driver you've a couple of more option.

it's called 3 in 1 diming for a reason.
you can use pwm, a 1-10vdc signal (0-10 vdc as well with some expedient) or even a couple of resistor on a switch.

"LED driver will change its output current based on 3 different input signal/setup
including DC 1-10V, PWM and 100K ohm resistor dimming."

"DC 0-10V:
DC 0-10V is considered as the second generation of 3 in one dimming. PWM and 100K
resistor dimming is still same as first generation. Difference between DC0-10V and DC1-
10V is addressed here.
The maximum level is still 100% in case the dimming signal is given at 10V or open the
dimming wires. However the minimum level for DC0-10V is 5.7% in case the dimmin signal is given at 0.57V.
In case the dimmer is giving lower than 0.57V or user just short
the dimming wires, the LED driver will cut off the output current resulting no light output
in the LED module."

http://www.meanwell.eu/ExclusivePDF/DIMMER-NOTE.pdf
http://www.meanwell.eu/ExclusivePDF/DIMMING-NOTE.pdf (are two different document, called in the same way ...)



maybe i'm not the best one to explain things in english, but yeah the idea bihind the other two option works like that.

by modify the volt on the dim+/- output the cob output is modified as well.
but your not touching the volt delivered to the cob, you are going to limit the current delivered (or at least you are not "touching" it by much).
the theory is still it, W=v*a, so if we decrease (limit) the current we are decreasing the total watt consumption.

I wasn't talking about the PWM signal input used to control the dimmer, I was talking about the PWM signal that comes out of the dimmer that lights the LED COBS eg. a PWM of 36V at 1400mA which pulses on and off many times a second.

I think people here are failing to grasp that dimming LED drivers don't output DC to the LED COBs, they output a PWM at the full 1400mA current, or whatever current the driver is rated at. In other words, if you dim the driver to 50% power, it's still putting out 1400mA, that doesn't change.

 

[jikko77]

I wasn't talking about the PWM signal input used to control the dimmer, I was talking about the PWM signal that comes out of the dimmer that lights the LED COBS eg. a PWM of 36V at 1400mA which pulses on and off many times a second.

I think people here are failing to grasp that dimming LED drivers don't output DC to the LED COBs, they output a PWM at the full 1400mA current, or whatever current the driver is rated at. In other words, if you dim the driver to 50% power, it's still putting out 1400mA, that doesn't change.

are you sure on this?

the dim + and dim - present on the B model of meanwell driver have an output of 10v and very little amp.
that output can be "conditioned" with the stuff descrived above.
with either a pwm signal or just some resistence or even a vdc.

what you do modify isn't the friking volts, when dimming the led.
they do modify the volts, by very little, in function of the temperature and how much current they are drive at.

when you dim down a driver you do limit the current the cobs will receive, and by current i'm talking about ampere.

maybe this piece of pct will do as an example:



let me comment something else:

In addition, lamp life is not affected by dimming, as is sometimes the case with frequently dimmed fluorescent lighting. Rather, dimming LEDs may lengthen the useful life of LEDs, because dimming can reduce operating temperatures inside the light source.

http://www.ledsmagazine.com/article...hting-systems-introducing-the-led-driver.html

the lamp isn't affected by dimming isn't complitly true.
maybe a vdc signal is less invasive than the pwm one, but the pwm can shorten the cob life:

"The Applications Engineering team at Cree is often asked
whether it is safe to operate Cree XLamp® LEDs with pulsed
currents above the maximum data-sheet rating. This question is
usually asked in the context of legitimate product requirements
such as those posed by emergency-vehicle applications,
specialized stroboscopic illumination and even pulsed
modulation for general-illumination dimming applications.
The short answer is “it depends.” Multiple variables affect both
initial and long-term performance and reliability of an LED.
These include thermal resistance, pulse duration, as well as
current amplitude, frequency and duty cycle."

 

[Arf]

are you sure on this?

the dim + and dim - present on the B model of meanwell driver have an output of 10v and very little amp.
that output can be "conditioned" with the stuff descrived above.
with either a pwm signal or just some resistence or even a vdc.

what you do modify isn't the friking volts, when dimming the led.
they do modify the volts, by very little, in function of the temperature and how much current they are drive at.

when you dim down a driver you do limit the current the cobs will receive, and by current i'm talking about ampere.

maybe this piece of pct will do as an example:

https://www.icmag.com/ic/picture.php?albumid=65534&pictureid=1634363View Image

let me comment something else:



the lamp isn't affected by dimming isn't complitly true.
maybe a vdc signal is less invasive than the pwm one, but the pwm can shorten the cob life:

"The Applications Engineering team at Cree is often asked
whether it is safe to operate Cree XLamp® LEDs with pulsed
currents above the maximum data-sheet rating. This question is
usually asked in the context of legitimate product requirements
such as those posed by emergency-vehicle applications,
specialized stroboscopic illumination and even pulsed
modulation for general-illumination dimming applications.
The short answer is “it depends.” Multiple variables affect both
initial and long-term performance and reliability of an LED.
These include thermal resistance, pulse duration, as well as
current amplitude, frequency and duty cycle."

FFS, I feel as if I am going round in circles.

I expressed a problem in post #1059 that my multimeter couldn't be used to work out the power the dimmable driver was producing because the output to the LED COBs was PWM not steady DC. Without an oscilloscope I can't determine the frequency, voltage, or the mark/space ratio of the PWM current going to the LEDs, hence the power they are actually consuming.

At no point have I talked about the PWM input to control the driver, until you mentioned it. It is completely irrelevant to my problem.