adjust output from constant current driver to different LED's?

[mds82]

Hi All. This may sound like a crazy idea but i'm putting it out there anyways. So lets say you have a constant current driver such as the Buckpuck 1000ma driver from LED Dynamics hooked up to a 12v input source. It wil be able to drive 3 XP-G's all in series, or a different number if you do them in parallel. I am already aware with the potential problems with driving LED in parallel however.

What i would like to try and do is to run white and blue LED's, however be able to somehow adjust the amount of current going to both without adjusting the BuckPuck itself. My First thought would be to have 2 parallel strings of LED,s 3 white, 3 blue, and somehow be able to adjust something to make the whit or blue string brighter.

Let me know if this makes sense and if it would be at all possible.

 

[jtr1962]

You could put a low-dropout linear regulator in series with one string. The Buckpuck will try to keep the total current to both strings constant. If the linear regulator keeps one string at, say, 300 mA, then the current going to the other string will be 700 mA. Such a setup is actually a good way to get variable CCT. You might have a string of 2700K LEDs and 6500K LEDs. You keep total current to both strings constant via a switching regulator so total light output is more or less constant. You vary current in one string via a linear regulator to adjust CCT to any value between 2700K and 6500K.

 

[Steve K]

or wire the LEDs in series, and wire a resistor in parallel with each LED. This would let you reduce the current in each LED. Not necessarily the most efficient method, though.

 

[DIWdiver]

Or you could put a small resistor in series with each string. By varying the value(s), you can change the current in each string. Unfortunately variable resistors that can handle high currents are bulky and expensive. This would only give you limited control range.

If you wanted to get fancy, I could show you how to build a circuit with a FET or two, some resistors, and a low-power variable resistor or two. One FET would give you limited range, two would give you full range.

 

[mds82]

OK, so what about doing something like this diagram below. It would have 2 parallel strings, and with the 2nd string it would have a potentiometer on it to adjust the resistance. This would allow to manually adjust the resistance and therefor the current to it. Would this work?

If this would work, what values would i need for the pot? ohms to ___ ?

 

[uk_caver]

Issues with potentiometers for direct power control include finding ones with sufficient power handling, and knowing in advance what resistance range is required for a particular set of LEDs given that the forward voltage of LEDs varies.
Additionally, the LED forward voltage for a given current (the Vf) drops with increasing temperature (and hence with increasing power).
Just using variable resistance would lead to a situation which will have some drift after adjustment.

For example, initially the resistance is high, your bluer string current is low, and hence the LEDs in that string are cool, and therefore have a 'cool' (higher) Vf.
You want bluer light, so you adjust the pot to increase current in the blue string to get the colour you want, and stop adjusting.
The LEDs heat up and their Vf drops, which causes a more current to flow, which causes the LEDs to heat a little more...
Some stable point will be reached, but it won't be quite the same overall colour you had set when you adjusted the pot.
I'm not sure how much this would happen - possibly not enough to be an issue, but it would happen a little.

A variable constant-current circuit on one of the strings would not suffer the same issues, and one could be built reasonably easily with a few components.

(I'm assuming that in your circuit you intend to wire the moving contact of the pot to one or other end of it, and that that just didn't come through on the diagram.)

 

[mds82]

uk carver - I completely understand what you mean about the slight changes that will happen over time so thank you for that information. As well yes i jus did a quick drawing and forgot to include the contacts on the pot.

As for a variable constant-current circuit, do you think you would be able to assist in pointing me in the right direction on this. Thank you in advance for your assistance with this.