So this is a little off topic, but I am building a DIY projector and am having a tiny bit of difficulty with the elctronics (I got the optics down pat, but I am still learning bout electronics lol).
I am planning on using this LED http://www.better-led.com/en/products/leds/LQ-LED-200W-02.html which can take a 36volt 6 amp power supply and needs a constant power supply.
I am wondering if I can use an SMPS to power the LED? Something like the 36volt meanwell that hifimediy sells. From what I can see on the meanwell product page you can adjust the amp output from an SMPS? Otherwise the 36volt one puts out 13amp not 6.
Just looking at my options and trying to learn a bit as well.
Thanks for any replies,
Mitch
I am planning on using this LED http://www.better-led.com/en/products/leds/LQ-LED-200W-02.html which can take a 36volt 6 amp power supply and needs a constant power supply.
I am wondering if I can use an SMPS to power the LED? Something like the 36volt meanwell that hifimediy sells. From what I can see on the meanwell product page you can adjust the amp output from an SMPS? Otherwise the 36volt one puts out 13amp not 6.
Just looking at my options and trying to learn a bit as well.
Thanks for any replies,
Mitch
Yes you can use smps, in fact I wouldn't use anything else... but I think you can set only the output voltage... you would need led driver or constant current regulator, that being said it could be linear or switching...
led driver is the most... well I don't know if I would say simple but... you have input voltage, that needs to be higher then what led will use, say 40-50v (if you have such led driver) and driver needs to be set to provide 6A of constant current.
next DIY thing would be linear constant regulator, that is ultra simple (can be). you need transistor and resistor and you are good to go.
switching regulator is the same as led driver, in fact this is one and the same thing, good coz it doesn't heat up no where as much as linear would
Anyway... for your led, MOST important thing is to LIMIT current to 6A (voltage doesn't mean anything here, will be as mush as it needs to be for 6A) and to keep is as cool as you can, don't run it near 100C or whatever it is rated for
led driver is the most... well I don't know if I would say simple but... you have input voltage, that needs to be higher then what led will use, say 40-50v (if you have such led driver) and driver needs to be set to provide 6A of constant current.
next DIY thing would be linear constant regulator, that is ultra simple (can be). you need transistor and resistor and you are good to go.
switching regulator is the same as led driver, in fact this is one and the same thing, good coz it doesn't heat up no where as much as linear would
Anyway... for your led, MOST important thing is to LIMIT current to 6A (voltage doesn't mean anything here, will be as mush as it needs to be for 6A) and to keep is as cool as you can, don't run it near 100C or whatever it is rated for
Thanks for the reply Luka.
So pretty much any voltage power source will work as long as it putting out 6A ? Eg Something like so: eBay Australia: Buy new & used fashion, electronics & home d?r would do the job?
If so now I just need to find a BIg *** heatsink and rig up some fans
So pretty much any voltage power source will work as long as it putting out 6A ? Eg Something like so: eBay Australia: Buy new & used fashion, electronics & home d?r would do the job?
If so now I just need to find a BIg *** heatsink and rig up some fans
well you need the power... so with switching led drivers, if you input higher voltage, less current will flow, since I x U = P and lets say that all that power goes to output.
So if you had driver that could use 48v, you wouldn't need this 48v supply to be 6A too. Yes, that 42v supply would work great... now if you can only find led driver at 6A you are golden
but I think you might need to make one... if you are up for this... if not, you will need to dissapate suppy voltage - led voltage @6A x 6A... not too much, but still something
So if you had driver that could use 48v, you wouldn't need this 48v supply to be 6A too. Yes, that 42v supply would work great... now if you can only find led driver at 6A you are golden
but I think you might need to make one... if you are up for this... if not, you will need to dissapate suppy voltage - led voltage @6A x 6A... not too much, but still something
Last edited:
That power supply isn't technically a switching LED driver, it isn't as suitable as the one mentioned previously. It regulates based on voltage, the current is just the peak which isn't how you want to operate one so you get good lifespan from the SMPS, instead you'd want a current margin between SMPS spec and average current consumption.
While with these they do tend to drop rail voltage as you approach or exceed the current spec for the PSU, it isn't straighforward that you can assume it would deliver 6A, not more than that, or fall out of regulation because the forward voltage vs current curve for the LED doesn't match it. If it has a voltage adjustment feature you might be able to dial in something compatible but ultimately the far better option is a method that directly limits current.
While the LED has a 33V forward voltage spec at 6A, this is at 25C junction temperature - a temperature you surely will not have unless you use active cooling (peltier, compressor, etc) and even then, a pretty good cooling subsystem for a consumer app with 200W LED.
Instead, we should probably assume the forward voltage will drop below 33V at the actual operating temp as implemented in your project. How much I do not know, this cannot be an inexpensive LED so you are justified in asking for full specs from the manufacturer, or to forward you on to the actual manufacturer if this is only a reseller for the LED.
There are probably some industrial lighting or automotive targeted ICs for current regulating high power LEDs, but this LED is unusual in having so many in series and parallel, normal LED config would be more in series, multiple channels for parallel, not 200W in a semi-symmetrical arrangement. You might ask the LED manufacturer what they suggest to drive it, surely they had some ideas before deciding on this logical die arrangement??
If all else fails, the not-so-power-conservative but generic option is you get a SMPS rated for a few volts over 33V, not much(!), rated for more than the 6A, perhaps 9A or more. It would be preferred if it had a voltage adjustment feature. Based on the highest (or steady rated if non-adjustable) voltage it supports at 9A or more, you do the standard LED resistor calculation. For example,
(PSU voltage - LED forward voltage) / current = resistance
(PSU voltage - LED forward voltage) * current = resistor wattage ... add some margin for good life and easier cooling from lower thermal density
For example, 36V PSU and 32V LED forward voltage (@ higher thermal junction temperature than 25C) and 6A drive current:
(36V - 32V) / 6A = 0.67 Ohms
4V * 6A = 24W (add as much margin above this as budget and space allow, 50W power resistor would not be unreasonable, otherwise you need a bit better, more expensive heatsink instead unless the massive one the LED is mounted to serves double duty, which it might as well do if the design allows for that)
Maybe your actual values input above (plus an adjustable PSU voltage) will allow picking a resistor that is a commonly available value, or maybe you'll need to parallel some higher ohm resistors to arrive at the right value.
This is where an adjustable PSU might really come in handy. For example, "IF" your LED's actual forward voltage at operating temperature were 32V as randomly guesstimated above for the example, you could try putting 3 x 10W, 1 ohm resistors in parallel to limit current. That gives 30W power handling, 0.33 Ohms resistance, and per the equation above:
(PSU V - 32V) / 6A = 0.33 Ohms
PSU V = 34V
2V * 6A = 12W resistor dissipation
So, "IF" your PSU can adjust down to 34V and LED Vf is 32V, this becomes reasonable, manageable, using common cheap resistor values. Obviously LED forward voltage won't be exactly, if even near 32.0V, it is just a generalization about one method it could be implemented. Certainly when you start actually powering the LED from what you have built, you'd want to set the PSU to the lowest voltage it supports, then measure the current consumption of the LED with a meter as you slowly increase the SMPS voltage... also keeping in mind there will be a little lag, as current increases it takes a while for temperature rise to normalize which will drop forward voltage a little more causing a little higher current and so on till you eventually dial in exactly what you need for PSU output.
PS - my math is not always flawless, I usually calculate things two or more times if a critical value is being chosen.
How well this could work out depends a lot on how adjustable the PSU voltage is within an appropriate range. If it were a fixed 42V you have massive amounts of heat on resistors, even 36V would cause a lot, while if you can dial down voltage till it's barely more than needed, efficiency comes pretty close to what a proper LED current regulator subcircuit would... as per the example above you "could" only have 12W resistor loss driving a 200W load while many switching LED current regulation circuits aren't even 94% efficient.
While with these they do tend to drop rail voltage as you approach or exceed the current spec for the PSU, it isn't straighforward that you can assume it would deliver 6A, not more than that, or fall out of regulation because the forward voltage vs current curve for the LED doesn't match it. If it has a voltage adjustment feature you might be able to dial in something compatible but ultimately the far better option is a method that directly limits current.
While the LED has a 33V forward voltage spec at 6A, this is at 25C junction temperature - a temperature you surely will not have unless you use active cooling (peltier, compressor, etc) and even then, a pretty good cooling subsystem for a consumer app with 200W LED.
Instead, we should probably assume the forward voltage will drop below 33V at the actual operating temp as implemented in your project. How much I do not know, this cannot be an inexpensive LED so you are justified in asking for full specs from the manufacturer, or to forward you on to the actual manufacturer if this is only a reseller for the LED.
There are probably some industrial lighting or automotive targeted ICs for current regulating high power LEDs, but this LED is unusual in having so many in series and parallel, normal LED config would be more in series, multiple channels for parallel, not 200W in a semi-symmetrical arrangement. You might ask the LED manufacturer what they suggest to drive it, surely they had some ideas before deciding on this logical die arrangement??
If all else fails, the not-so-power-conservative but generic option is you get a SMPS rated for a few volts over 33V, not much(!), rated for more than the 6A, perhaps 9A or more. It would be preferred if it had a voltage adjustment feature. Based on the highest (or steady rated if non-adjustable) voltage it supports at 9A or more, you do the standard LED resistor calculation. For example,
(PSU voltage - LED forward voltage) / current = resistance
(PSU voltage - LED forward voltage) * current = resistor wattage ... add some margin for good life and easier cooling from lower thermal density
For example, 36V PSU and 32V LED forward voltage (@ higher thermal junction temperature than 25C) and 6A drive current:
(36V - 32V) / 6A = 0.67 Ohms
4V * 6A = 24W (add as much margin above this as budget and space allow, 50W power resistor would not be unreasonable, otherwise you need a bit better, more expensive heatsink instead unless the massive one the LED is mounted to serves double duty, which it might as well do if the design allows for that)
Maybe your actual values input above (plus an adjustable PSU voltage) will allow picking a resistor that is a commonly available value, or maybe you'll need to parallel some higher ohm resistors to arrive at the right value.
This is where an adjustable PSU might really come in handy. For example, "IF" your LED's actual forward voltage at operating temperature were 32V as randomly guesstimated above for the example, you could try putting 3 x 10W, 1 ohm resistors in parallel to limit current. That gives 30W power handling, 0.33 Ohms resistance, and per the equation above:
(PSU V - 32V) / 6A = 0.33 Ohms
PSU V = 34V
2V * 6A = 12W resistor dissipation
So, "IF" your PSU can adjust down to 34V and LED Vf is 32V, this becomes reasonable, manageable, using common cheap resistor values. Obviously LED forward voltage won't be exactly, if even near 32.0V, it is just a generalization about one method it could be implemented. Certainly when you start actually powering the LED from what you have built, you'd want to set the PSU to the lowest voltage it supports, then measure the current consumption of the LED with a meter as you slowly increase the SMPS voltage... also keeping in mind there will be a little lag, as current increases it takes a while for temperature rise to normalize which will drop forward voltage a little more causing a little higher current and so on till you eventually dial in exactly what you need for PSU output.
PS - my math is not always flawless, I usually calculate things two or more times if a critical value is being chosen.
How well this could work out depends a lot on how adjustable the PSU voltage is within an appropriate range. If it were a fixed 42V you have massive amounts of heat on resistors, even 36V would cause a lot, while if you can dial down voltage till it's barely more than needed, efficiency comes pretty close to what a proper LED current regulator subcircuit would... as per the example above you "could" only have 12W resistor loss driving a 200W load while many switching LED current regulation circuits aren't even 94% efficient.
Last edited:
You can probably buy something off the shelf that's designed specifically to drive LEDs for cheaper than you can build something.. like a Mean Well HLG-240-36:
Meanwell Power Supply Products
$110USD at Future Electronics, <$100 at Online Components, stock but a bit spendy at Mouser.
Meanwell Power Supply Products
$110USD at Future Electronics, <$100 at Online Components, stock but a bit spendy at Mouser.
^ Often they instead put the rail through a low ohm resistor after the inductor, the resistor value varied to meet the reference voltage the regulator is set to (or is fixed at) use to adjust current output.
However, either way it might not be as simple in practice as it is in theory, you'll be needing a 2nd circuit tacked on or a controller that can take the place of the original, if your voltage drop isn't sufficiently high enough for the existing regulator IC to use... though I suppose you might be able to use a voltage divider to pull the signal up towards the rail voltage some instead of down towards the ground.
... unless you had something specific in mind, parts suitable for that so the time investment isn't high enough to justify buying a ready made driver supply like gmarsh linked. The HLG-240-36 model seems a good option if Mitch doesn't already have a suitable SMPS.
However, either way it might not be as simple in practice as it is in theory, you'll be needing a 2nd circuit tacked on or a controller that can take the place of the original, if your voltage drop isn't sufficiently high enough for the existing regulator IC to use... though I suppose you might be able to use a voltage divider to pull the signal up towards the rail voltage some instead of down towards the ground.
... unless you had something specific in mind, parts suitable for that so the time investment isn't high enough to justify buying a ready made driver supply like gmarsh linked. The HLG-240-36 model seems a good option if Mitch doesn't already have a suitable SMPS.
If it has a separate voltage sense line you might be able to use it as a constant current supply. But it will not be very efficient and will get quite hot. The V out goes through the leds to a resistor to ground. The sense terminals go across the resistor. The resistor is picked to have 6 amps through it at the lowest voltage the supply will regulate to.
With a good DC opamp you might be able to get the voltage across the resistor to under a volt.
If you try building such a supply use a fuse instead of the leds until you get it right.
With a good DC opamp you might be able to get the voltage across the resistor to under a volt.
If you try building such a supply use a fuse instead of the leds until you get it right.
I suggest you just build it. A 36V/6A SMPS is quite simple to build using an IC controller, 1-2 FETS, 1-2 Shottky diodes, a transformer, some caps, etc..
I think a SMPS is not just an useful circuit but also a "fun" one.
There is a lot of info available on how to build such a circuit.
If you do it, please make a thread.
I think a SMPS is not just an useful circuit but also a "fun" one.
There is a lot of info available on how to build such a circuit.
If you do it, please make a thread.
With a typical current regulating controller it is a simple matter of providing a true constant current.
The controller will have a built in comparator to it's reference voltage, taking feedback from a point just after the inductor AND right after a series resistor. No matter how the LED forward voltage varies as the die temperature changes, the supply can regulate to (a specific current) 6A because it is measuring across the fixed value series resistor.
Here's one example, LM3421 and a place it's available (digikey) in the US.
Digi-Key - LM3421MH-ND (Manufacturer - LM3421MH/NOPB)
Datasheet http://www.national.com/ds/LM/LM3421.pdf
The datasheet considers the resistor to sense current as "Rsns". There are a few example circuits with BOMs to make parts selection easy including a BUCK circuit on page 41, but as per their prior calculations you'd have to adjust some values to get 6A instead of 1.25A output.
However, this is only the back end of the whole power supply, you'll still need to drop input voltage down from AC mains line to within the range any particular controller can tolerate. There may be single IC solutions to manage the entire power supply from AC input to current regulated output but I don't know any part numbers for them.
The controller will have a built in comparator to it's reference voltage, taking feedback from a point just after the inductor AND right after a series resistor. No matter how the LED forward voltage varies as the die temperature changes, the supply can regulate to (a specific current) 6A because it is measuring across the fixed value series resistor.
Here's one example, LM3421 and a place it's available (digikey) in the US.
Digi-Key - LM3421MH-ND (Manufacturer - LM3421MH/NOPB)
Datasheet http://www.national.com/ds/LM/LM3421.pdf
The datasheet considers the resistor to sense current as "Rsns". There are a few example circuits with BOMs to make parts selection easy including a BUCK circuit on page 41, but as per their prior calculations you'd have to adjust some values to get 6A instead of 1.25A output.
However, this is only the back end of the whole power supply, you'll still need to drop input voltage down from AC mains line to within the range any particular controller can tolerate. There may be single IC solutions to manage the entire power supply from AC input to current regulated output but I don't know any part numbers for them.
Last edited:
Ok, So... Long time no talk lol.
I have been really busy, but finally been working on this LED power source again recently. I was actually wondering if something simple like this: New 36V DC 10A 350W Regulated Switching Power Supply | eBay would work?
Set it to 6amp 36v and off I go? I don't want to hash around with a $140 LED with my dodge electronics skills. The LED I plan to use has changed a bit too. Planning on this guy here: 200W High Power LEDs, Red, Green, Blue, White, Warm White, Yellow for Choice [LED-200WHPL] - US$139.80 : SatisLED Store, China LED wholesaler-China LED Manufacture-Wholesale LED lights at Factory Price Takes 36V 6Amp very well according to the manufacturers. I will be watercooling the LED, using PC CPU watercooling parts.
Possibly ok combination?
Thanks Again to everyone who replied, you lead me down some interest research avenues and showed me that I need to do electronics 101 very soon.
I have been really busy, but finally been working on this LED power source again recently. I was actually wondering if something simple like this: New 36V DC 10A 350W Regulated Switching Power Supply | eBay would work?
Set it to 6amp 36v and off I go? I don't want to hash around with a $140 LED with my dodge electronics skills. The LED I plan to use has changed a bit too. Planning on this guy here: 200W High Power LEDs, Red, Green, Blue, White, Warm White, Yellow for Choice [LED-200WHPL] - US$139.80 : SatisLED Store, China LED wholesaler-China LED Manufacture-Wholesale LED lights at Factory Price Takes 36V 6Amp very well according to the manufacturers. I will be watercooling the LED, using PC CPU watercooling parts.
Possibly ok combination?
Thanks Again to everyone who replied, you lead me down some interest research avenues and showed me that I need to do electronics 101 very soon.
Mitch311, this type of PSU doesn't have a current "setting". It regulates to the stated voltage of 36V... which might be a bit adjustable, outputting up to 10A at which point it might fold back in voltage a bit or not have reliable operation or good lifespan.
To use a voltage regulating PSU like that you need to either precisely adjust its voltage output to achieve the right current by default (a poor and risky method of getting the job done) or add some series resistance which wastes some power but gives a much broader safe operating area as the forward voltage of the LED array changes due to it & heatsink heating up during use.
Have you any reviews of that LED from a reliable source? I would want to read some before spending that kind of money as a lot of off-brand LEDs have fairly poor lifespans far below seller claims. Granted in a projector it might not matter so much as in general area lighting on more hours every day.
To use a voltage regulating PSU like that you need to either precisely adjust its voltage output to achieve the right current by default (a poor and risky method of getting the job done) or add some series resistance which wastes some power but gives a much broader safe operating area as the forward voltage of the LED array changes due to it & heatsink heating up during use.
Have you any reviews of that LED from a reliable source? I would want to read some before spending that kind of money as a lot of off-brand LEDs have fairly poor lifespans far below seller claims. Granted in a projector it might not matter so much as in general area lighting on more hours every day.
If your LED "happened" to be especially matched to the voltage and current, what you would find is the PSU had a short lifespan because it was outputting at max the whole time.
It "could" work for awhile to try to match like this, but I would not do it for more than a one-shot experiment, certainly not when $200 or more is involved.
As for lifespan degradation, I wish it were only 10K hrs vs 50K, but sometimes the reduction is far worse than that. I've bought (granted, they were different specs) LEDS rated at 100K before and they were below acceptable output before a year elapsed (< 9K hours).
Point is, maybe it is not the case with this LED, but it might be... which is why before spending this amount of money I'd seek confirmation from an impartial 3rd party and be very doubtful until receiving it... unfortunately in this day and age, generic specs mean nothing more than you can reasonably pursue during the warrant period... but do they even provide a warranty or just "if you hook it up wrong we won't take it back"?
Don't get me wrong, they could be legit and as honest as they tried to be, but on the other hand anyone can just rig a series of LED dies and make claims if there is only profit, no repercussions in doing so. When it comes to high powered lighting I tend to trust only companies with a high value reputation at stake. On the other hand, this is my first look at this product, if you have confidence in it then it is a more educated decision... but still your gamble to make.
It "could" work for awhile to try to match like this, but I would not do it for more than a one-shot experiment, certainly not when $200 or more is involved.
As for lifespan degradation, I wish it were only 10K hrs vs 50K, but sometimes the reduction is far worse than that. I've bought (granted, they were different specs) LEDS rated at 100K before and they were below acceptable output before a year elapsed (< 9K hours).
Point is, maybe it is not the case with this LED, but it might be... which is why before spending this amount of money I'd seek confirmation from an impartial 3rd party and be very doubtful until receiving it... unfortunately in this day and age, generic specs mean nothing more than you can reasonably pursue during the warrant period... but do they even provide a warranty or just "if you hook it up wrong we won't take it back"?
Don't get me wrong, they could be legit and as honest as they tried to be, but on the other hand anyone can just rig a series of LED dies and make claims if there is only profit, no repercussions in doing so. When it comes to high powered lighting I tend to trust only companies with a high value reputation at stake. On the other hand, this is my first look at this product, if you have confidence in it then it is a more educated decision... but still your gamble to make.
Last edited:
^ Let me correct or clarify what I first stated in my last post. You could use a PSU that regulates based on voltage monitoring, like the one you linked, if all the following situations were present:
- Forward voltage drop of the LED at the voltage the PSU is designed to regulate to, results in less than or equal to the max current the LED is designed to accommodate.
- Resulting current is a bit below the max continuous current the PSU is rated for so it retains good lifespan, OR sometimes if you give a PSU more forced (fan) air cooling you can extend its lifespan at max current levels but there is no guarantee, it can depend on the particular PSU design and component quality/margin used to build it.
- The current flowing through the LED in this configuration causes near enough to the rated LED wattage that you are satisfied with the light output per dollar spent on the LED, since it is typical you end up with a little less than peak light output unless finely tuning the PSU output current another way, like adding a series resistance (resistor(s)).
- The LED thermal rise will cause the forward voltage to drop some increasing current through it at any particular voltage so the design must keep the LED in its safe operating current region at the peak temperature it would see as implemented... including heatsink, airflow, ambient room temp, enclosure used, etc... set up for actual use.
Tying all these things together, the "absolute max" rating for the LED is 33V, 6060mA. http://www.better-led.com/en/products/leds/LQ-LED-200W-02.html This would usually be interpreted as, you will not need more than 33V to achieve max safe output from the LED, but you may not need all 33V. Whatever voltage you do end up needing to achieve max output of the particular specimen of LED you receive, you need to limit current so it never exceeds 6060mA. If you just hooked this LED up to a 36V, 6A SMPS, you would exceed the max ratings and potentially damage the LED without extraordinary measures like phase change cooling.
Oddly the peak pulse current rating for it is lower at 5000mA. This is at odds with most LED specs because it appears that their absolute max current of 6060mA really is an absolute max but what they did not provide which most spec sheets would, is the nominal, expected steady /average current to reach full rated output. Either that or the listed specs are in error somehow. I "suspect" the error is that the true absolute max pulsed current is higher than 6060mA because they listed other parameters at 6060mA, but then where does the 5000mA come from? Did they get these two backwards and it can only tolerate 5000mA average and 6060mA pulsed? Something is wrong with these specs.
I would contact them and insist on the right specs, what they have listed seems technically impossible. Their operating life and environmental test data is also a bit ambiguous. "Room temperature Operating Life", 25C temp, IF - max DC, "Stress Duration 1000 hours". What does this mean, that operating it at its ratings in a 25C room results in only 1000 hours lifespan?? Then they list several notes below but none of them are numbered yet all "Failure Criteria" points to "Note 2" which is only "Failure Criteria" itself if they mean the 2nd item down on the Notes list.
Now a word about LED junction temperature. It is not rated very high at 70C. I hope I am remembering correctly that it was you that mentioned using water cooling. If you do then 70C is certainly achievable, but will require a fair amount of attention to detail... but should be much easier than using a huge fan-only cooled sink or gigantic passive 'sink.
Perhaps I have given too much information or things to think about. Yes you can use a standard 36V SMPS, rated for 6A or more (more being a significant benefit) and use a means between the PSU and LED to limit current like resistor(s). Otherwise you need a PSU that regulates current, it will be rated for a maximum voltage or voltage range but not primarily as "36V", not a single fixed voltage. It just so happens that PSU specifically designed to be LED drivers are the most common type of current regulating PSU... being purposefully designed to do it.
So, actual LED driver PSU is best. SMPS rated 36V >6A and series resistors is 2nd best. SMPS rated 36V, 6A with extra cooling and series resistors is 3rd best. 36V >=6A rated PSU with only a voltage adjustment pot that goes down to about 30V is a far distant 4th best (poor, and risks the costly LED).
It might be possible to modify the PSU to output even lower fixed voltage, or even regulate current somehow but these things seem beyond the scope of the project.
- Forward voltage drop of the LED at the voltage the PSU is designed to regulate to, results in less than or equal to the max current the LED is designed to accommodate.
- Resulting current is a bit below the max continuous current the PSU is rated for so it retains good lifespan, OR sometimes if you give a PSU more forced (fan) air cooling you can extend its lifespan at max current levels but there is no guarantee, it can depend on the particular PSU design and component quality/margin used to build it.
- The current flowing through the LED in this configuration causes near enough to the rated LED wattage that you are satisfied with the light output per dollar spent on the LED, since it is typical you end up with a little less than peak light output unless finely tuning the PSU output current another way, like adding a series resistance (resistor(s)).
- The LED thermal rise will cause the forward voltage to drop some increasing current through it at any particular voltage so the design must keep the LED in its safe operating current region at the peak temperature it would see as implemented... including heatsink, airflow, ambient room temp, enclosure used, etc... set up for actual use.
Tying all these things together, the "absolute max" rating for the LED is 33V, 6060mA. http://www.better-led.com/en/products/leds/LQ-LED-200W-02.html This would usually be interpreted as, you will not need more than 33V to achieve max safe output from the LED, but you may not need all 33V. Whatever voltage you do end up needing to achieve max output of the particular specimen of LED you receive, you need to limit current so it never exceeds 6060mA. If you just hooked this LED up to a 36V, 6A SMPS, you would exceed the max ratings and potentially damage the LED without extraordinary measures like phase change cooling.
Oddly the peak pulse current rating for it is lower at 5000mA. This is at odds with most LED specs because it appears that their absolute max current of 6060mA really is an absolute max but what they did not provide which most spec sheets would, is the nominal, expected steady /average current to reach full rated output. Either that or the listed specs are in error somehow. I "suspect" the error is that the true absolute max pulsed current is higher than 6060mA because they listed other parameters at 6060mA, but then where does the 5000mA come from? Did they get these two backwards and it can only tolerate 5000mA average and 6060mA pulsed? Something is wrong with these specs.
I would contact them and insist on the right specs, what they have listed seems technically impossible. Their operating life and environmental test data is also a bit ambiguous. "Room temperature Operating Life", 25C temp, IF - max DC, "Stress Duration 1000 hours". What does this mean, that operating it at its ratings in a 25C room results in only 1000 hours lifespan?? Then they list several notes below but none of them are numbered yet all "Failure Criteria" points to "Note 2" which is only "Failure Criteria" itself if they mean the 2nd item down on the Notes list.
Now a word about LED junction temperature. It is not rated very high at 70C. I hope I am remembering correctly that it was you that mentioned using water cooling. If you do then 70C is certainly achievable, but will require a fair amount of attention to detail... but should be much easier than using a huge fan-only cooled sink or gigantic passive 'sink.
Perhaps I have given too much information or things to think about. Yes you can use a standard 36V SMPS, rated for 6A or more (more being a significant benefit) and use a means between the PSU and LED to limit current like resistor(s). Otherwise you need a PSU that regulates current, it will be rated for a maximum voltage or voltage range but not primarily as "36V", not a single fixed voltage. It just so happens that PSU specifically designed to be LED drivers are the most common type of current regulating PSU... being purposefully designed to do it.
So, actual LED driver PSU is best. SMPS rated 36V >6A and series resistors is 2nd best. SMPS rated 36V, 6A with extra cooling and series resistors is 3rd best. 36V >=6A rated PSU with only a voltage adjustment pot that goes down to about 30V is a far distant 4th best (poor, and risks the costly LED).
It might be possible to modify the PSU to output even lower fixed voltage, or even regulate current somehow but these things seem beyond the scope of the project.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.