Buying a Dimmable Power Supply

Good day to all,

Please, I need your guide and support before buying this dimmable power supply:

http://www.osram.com/osram_com/prod…ent-dimmable/optotronic-intelligent/index.jsp

I have a PCB that has a cluster of 120 LEDs on it with series-parallel connection. (A cluster is made of 5 LEDs all connected in parallel). 12 clusters (60 LEDs) are then connected in series, this 12 cluster in series is now connected in parallel to another 12 clusters which is also in series. According to my calculation the PCB requires 200 mA and 38.4 V - 43 V (this is base on this series - parallel arrangement and the datasheet of the LED). The problem I am having is that the dimmable power supply I finally saw after so many search on the internet has an output voltage range of 12 V - 54 V and constant output current of 250 mA. My question is that, can I still use this power supply to power my PCB considering the fact that the power supply output voltage range is up to 54 V (more than 43 V that I need) and at the same time is has a constant output current of 250 mA (although the absolute maximum current rating of the LED is 30 mA according to the datasheet. This will give me 300 mA based on my LED arrangement.

Your support and guide will be highly appreciated in this situation.

Best regards.
Note: I have only the LEDs on the PCB, there are no other components. So the 43 V and 200 mA will be consumed only by the LEDs. A copy of the LED datasheet is attached to this thread.



SE-SMD3528-UV5

There are a number of issues with your project. But the short answer to your question is, you almost certainly can use this power supply (a copy of its data sheet would have been helpful). The 250mA rating is almost certainly the MAXIMUM current it will put out. The ajustable voltage range covers the voltage you say you need. You'll just set the control for the voltage you want and connect the LEDs.

However, your LED connection scheme is going to cause you problems. LEDs in parallel are not all going to light up the same, since tiny differences between them will mean that some will "hog" the available current and cause them to be brighter than their neighbors. As they heat up in operation, this typically gets worse. The simple solution for this is to put a resistor in series with each LED to swamp out the effect. Of course this raises the voltage needed on the parallel cluster for a given current. That will probably mean changing your series-parallel configuration to keep within your voltage range, or finding a different power supply.

Your power requirements for this array of LEDs is pretty modest, so you should be able to find, or build, a power supply to run these LEDs, in almost any configuration.

A little more data about what you're trying to accomplish would be helpful.

Thank you very much Mr Mark.I really appreciate your sense of understanding.I equally understand that (though my knowledge of electronic is still limited)arrainging Led in parallel is not a good Option.Unfortunately the PCB was not made by me,it was made by somebody else.I was only asked to look for a constant current power supply that can be used to dim the Leds.So i do not have the opportunity of making any changes to the PCB.However,since you said i can still use the power supply,then i will use it to power the board.Actually the only fear i had before has to do with blowing up the PCB as a result of over voltage.
I have equally informed the owner of the PCB,that the series-parallel Arrangement is not the best Option and that with time the Led may start developing faults.But it appears he cares less.So my only Task now is to get the Led working.
Once again thank you for taking the time to reply my question.i really appreciate your Kind gesture.I attached to this message the datasheet of the power supply in case you might have any other Suggestion.Best regards.

Ah, the power supply spec was very useful. It wasn't clear to me in your original posting that it was an adjustable-current supply for driving LEDs. That makes it even better for your purposes. The voltage spec is the "accomodation range" for LEDs at the whatever the output current is set for.

If the owner isn't too concerned with the potential problems, and has a way of operating this power supply (I see it has a couple of contol methods), he just needs to monitor the current and make sure he doesn't exceed the LED maximums (which this supply is quite capable of doing). I'd bring the LED board current up slowly from minimum the first time, just to make sure all is going well before increasing to maximum (which I'd keep to about 180mA). Watch out for hot LEDs; in the series-parallel arrangement you've got, some individual LEDs may go over the maximum. Diode lifetimes may be adversely affected.

Thank you very much Sir for taking the time to reply my question.I am very grateful.I will take note of the Observation you made.Best regards.

DO NOT BUY THAT POWER SUPPLY!

If you represented this correctly, that supply is a current regulator and will attempt to force 250ma to flow through your LED array! And, more than likely, it will run the voltage up to it's maximum output of 54V in it's attempt to do it!

It's the 54V rating that is a 'maximum', not the 250ma. Also, even though those LEDs are rated at 30ma Max, it has been my experience that running LEDs at their maximum rating shortens their life considerably. Much better to run them at 20ma, or pulse them at 70% of the maximum Peak Forward Current [following the recommended Duty Cycle and Pulse Width ratings in the data sheet].

And Colin is right! Connecting LEDs in parallel is a bad idea. Check out my blog on powering LEDs:

* http://www.sinistercircuits.com/Blog/LEDs_LightemWithoutSmokinem/index.php

To elaborate, when you see a voltage range on an LED driver/power supply, what it means is this:

When applied to a [single] string of LEDs it will apply a voltage across those LEDs that will either draw the rated current through the LEDs or one of the two will happen:

1. If the sum of the forward voltages in the LED string is less than the minimum voltage [12V for the supply you indicated], then it will apply 12V to the string (the lowest voltage is capable of applying). This will cause a large current to flow through the LEDs probably frying them.

2. If the sum of the forward voltages in the LED string is greater than the maximum voltage the LED driver can supply (54V in the supply indicated), then it will run the voltage that high and the LEDs will either not night up, or will light up dimmer than they would if driven at their nominal current. In other words, the LED driver will not be able to get the rated current [250ma for example] to flow through the LEDs.

I explain the reasons for this in my Blog.

OK, I noticed your other attachment, the spec sheet on the LED Driver. And, actually, for the arrangement of LED on that PCB, this supply might work BUT only if you make some modifications to that circuit board. you need to add balancing resistors in series with each of the five parallel LEDs. 10 ohms ought to do it. Unless that runs the maximum voltage up higher than 54V (don't have time to calculate it). But, be sure to not run the total current higher than around 200ma [2x5x20ma], unless that output of the LED Driver is pulsed and conforms to the LED spec regarding pulsing these LEDs, again no time to get into this any further.

Steve -- did you notice the poster's response earlier: " i do not have the opportunity of making any changes to the PCB," and "I have equally informed the owner of the PCB,that the series-parallel Arrangement is not the best Option and that with time the Led may start developing faults.But it appears he cares less." He's got to work with what he's got, unfortunately.

Yeah, I missed that -- thanks.

Yes, I agree with Mark. And if the LEDs are well matched and stay balanced, this may work just fine and have a good life span. But, if any of the LEDs have different enough forward voltages, then expect the whole array to gradually (or even quickly) go dark and even possibly go up in smoke.

One thing that might work is if you make a testing 'jig' using the LM317 LED driver shown in my blog [*www.sinistercircuits.com/Blog/LEDs_LightemWithoutSmokinem*]. Adjust it to whatever current you want to drive each LED at -- I suggest 18ma [and then run the LED driver at 180ma]. Then, using a pair of alligator clips, connect the LM317 driver's outputs to the probs on a DVM and then use it to test each LED chip's forward voltage. On an antistatic surface, sort them into piles, each pile being within a 100th of a volt from each other. So, one pile for 3.00, one for 3.01, one for 3.02, etc. Then, when they are all sorted, create groups of 5 LEDs that are as closely matched as possible, then two groups of 12x5 groups so that the 12 average voltages [i.e. the average forward voltage of the 5 LEDs] of one group adds up to close to the sum of the averages of the other group of 12.

In other words, try getting the thing as balanced as possible from the start, and hope the forward voltages don't drift with age ;)

BTW: The supply voltage that is shown on my blog [for the LM317 driver] is for driving multiple LEDs [how many is dependent on the sum of the Forward Voltages]. For your testing jig, the voltage can be much lower. 7.2V or more.

Also, in case it wasn't clear: you will probably need to run the LED driver at or near it's minimum current output [somewhere around 200ma], so it's doubtful you will get any [or much] dimming action.