Quad channel control for LED color tuning

[purduephotog]

Evening-

I hate to drop this thread here, but fixed lighting doesn't get enough traffic and it was going to originally be designed with 4x flashlight drivers (by modding the sense resister to a pot).

I have some 4 die LEDs (R,G, B, A). I would like to set up a small demo unit to drive them from a wall wort- and be able to vary the intensity of each die to drive my color to where I want it (high CRI, etc).

I may end up going with a 7 up Phillips lumiled with various dies on it in order to get the spectrum coverage I want.

Any inexpensive single emitter drivers out there that you know of? Or am I stuck modding some individual drivers?

Eventually I will be building a large scale troffer of these for testing and deployment, and I will probably end up daisy chaining them or placing them back into a feedback circuit with an atmel controller.

Thanks in advance-

Jason

 

[CuriousOne]

You need to specify desired current, voltage, topology, control method and available soldering skills

 

[purduephotog]

You need to specify desired current, voltage, topology, control method and available soldering skills

ROFL.

Assume I am an EXPERT! MUHAHAHAHA.

Actually I'll be reflowing the OSRAM onto some stars as soon as the order comes in from Digi. Soldering is fine. I just had a tech here remove some 0803 resisters from another driver for flashlights, so even if my skillz aren't up to task I know his are.

I'd love a turn key solution, but I doubt one exists, so I'm willing to make what I need to. Was looking at some Atmel chip MSL2041 4-String LED Driver with 4 PWM Inputs

Right now the demo is a 1 star unit with multiple diffusing materials (4 dies, 1 chip). That puts maximum Vf around 3.8v; a buck converter would gladly handle 5v. They will be wired independent of each other so I can tune the color spectrum.

I will hopefully also be testing different stars as well with different quad diodes and, perhaps, even multi chip MCPCBs so as many knobs as I can get the happier I'll be. When I go to production it'll be strip everything out, of course.

Jason

 

[CuriousOne]

PWM is evil on color leds, believe me.
Color leds have no phosphor, thus, no on/off delay, so PWM will be very uncomfortable on eyes.
I'm doing custom decorative led lighting solutions for latest 5 years and I know what I'm talking about.

I've manufactured a specific device for a customer, who wanted BCGAR (blue-cyan-green-amber-red) lights to be precisely controlled in his aquarium. My device outputs 0.1A-1A per channel, each channel is individually adjustable and can drive 1-4 led strings. Has logical on/off inputs for each channel, which can be used for PWM, but I don't reccomend it.

For another one, I've did 96 channel, SPI driven board, which was using DAC to smooth PWM output. Each channel outputting 1-100mA current and can work with voltages up to 400VDC.

So, to get professional suggestion, you need to be more specific.

 

[WalkIntoTheLight]

I have some 4 die LEDs (R,G, B, A). I would like to set up a small demo unit to drive them from a wall wort- and be able to vary the intensity of each die to drive my color to where I want it (high CRI, etc).

I don't think you'll be able to get a high CRI light by combining a bunch of RGB LEDs. Those colour LEDs are pretty monochromatic. You'll still be missing a whole bunch of the spectrum in-between each colour band.

You might be able to create a white light with a tint (colour temperature) you want, but its spectrum will have dead spots in it that won't reflect true colours from certain objects.

You need something (like a phosphor) to spread out some of the spectrum, to get a high CRI.

 

[purduephotog]

PWM is evil on color leds, believe me.
Color leds have no phosphor, thus, no on/off delay, so PWM will be very uncomfortable on eyes.
I'm doing custom decorative led lighting solutions for latest 5 years and I know what I'm talking about.

I've manufactured a specific device for a customer, who wanted BCGAR (blue-cyan-green-amber-red) lights to be precisely controlled in his aquarium. My device outputs 0.1A-1A per channel, each channel is individually adjustable and can drive 1-4 led strings. Has logical on/off inputs for each channel, which can be used for PWM, but I don't reccomend it.

For another one, I've did 96 channel, SPI driven board, which was using DAC to smooth PWM output. Each channel outputting 1-100mA current and can work with voltages up to 400VDC.

So, to get professional suggestion, you need to be more specific.

Agreed on PWM. I hate it. Which is why my last set of mods used variable sense resisters (took it down to 47k from 0.1 ohm- and it was STILL too bright for the application).

Alright- the diode right now is an OSRAM OSTAR LE RTDUW S2W.and the ACWUWV. I'm soldering them down to a MCPCB with 4 channels. Maximum current appears to be 700ma on each channel (lit says 1A, but I don't need that much power). MCPCB is a Bergquist 803808.

Now, current driven is the key. For demonstration I wanted to just have variable current, say, 1ma +/- up to 500ma. That's some pretty decent light at 500ma and I can mix it however I need to to drive up the CRI, color temp, and still put it on a small enough heatsink to put in a fixture for out-of-sight measurements. Pot control is fine for now. Eventually I'd like to do color sensors (WAY down the road) but I'm not that critical- I love how the Cree LR6-DR1000s responds to color changes.

I suppose I could do PWM at a high enough frequency but, as you've said, I hate PWM. I personally was desoldering sense resisters and modding them with pots to do the last controller.

Your last item, mutli controller for up to 7 channels sounds just about perfect (although I'd want to go higher on the amperage). I take it this wasn't COTS? Did you have to route a board?

 

[CuriousOne]

I run a small company, where we design and manufacture in small quantities if needed, various electronics equipment, so that product was profesionally made on PCB.

Getting larger current output is not a problem at all, just it affects price, and also current maintaining precision is important.

If you want to go DIY way, try this: http://www.pcbheaven.com/userpages/LED_driving_and_controlling_methods/?topic=worklog&p=3

 

[SemiMan]

High enough frequency and you would never know there was PWM. Anything over 1KHz even at full modulation is not going to cause visual issues. If you are photographing, that may be another ballgame.

There are any number of color control units on the market .... just do a Google search ... easy to find. Not free of course.

Or you could just buy a Philips Hue bulb.

 

[purduephotog]

I run a small company, where we design and manufacture in small quantities if needed, various electronics equipment, so that product was profesionally made on PCB.

Getting larger current output is not a problem at all, just it affects price, and also current maintaining precision is important.

If you want to go DIY way, try this: http://www.pcbheaven.com/userpages/LED_driving_and_controlling_methods/?topic=worklog&p=3

Well if you own the design and fab, then you can sell more ... ?

May I send you off a PM to discuss costs? I've had on order a bunch of buckpucks (lumidrive?) for a while but they still haven't arrived, which was why I went searching for larger multi-channel solutions.

 

[alpg88]

just buy rgb controllers off e bay, for led strips, they work with 9-12v, and can take enough current to drive 5m strip. they can adjust brightness of inividual color, have few preset colors, flashing patterns,.... these controllers can work with other leds, but they have to be either 12v leds, or wired in series with resistiors or other leds. controllers are cheap, like $10 or so. with no strip and power supply, they take any wall ac 12v adapter

 

[CuriousOne]

There is a catch - higher the PWM frequency, change between brightness levels is much more noticeable. So you will have either smooth change and noticeable flicker, or more discrete change and less noticeable flicker.

 

[SemiMan]

There is a catch - higher the PWM frequency, change between brightness levels is much more noticeable. So you will have either smooth change and noticeable flicker, or more discrete change and less noticeable flicker.

Implementation detail. Within practical limits you can have both.

 

[CuriousOne]

Yes, and these "practical limits" will be visible by naked eye.
Just like Plasma TV. Some people notice flicker, regardless of "600hz 800hz" and so on technologies, some - not.

 

[alpg88]

I never saw a hint of flicker with rgb controllers for led strip, only smooth gradual dimming as brightness changed. im may have pwm but i see nothing what people here say i should have seen.

i also agree, you wont get high cri by mixing color leds, you can adjust colors in wide range, but cri, i doubt you'd get what you want

 

[CuriousOne]

As said, some people see, some - not.

 

[SemiMan]

Yes, and these "practical limits" will be visible by naked eye.
Just like Plasma TV. Some people notice flicker, regardless of "600hz 800hz" and so on technologies, some - not.

Sorry, but you are talking out of your butt at this point.

If I PWM at 1.5 KHz, you would never notice flicker. I don't mean just not see it, but no physiological effects at all. If I truly PWMed at 600 or 800Hz, you would not either unless you worked to specifically create conditions for it, i.e. something that modulated into a frequency below your particular flicker fusion. It is questionable whether there would even be any physiological effects at those frequencies.

 

[SemiMan]

I never saw a hint of flicker with rgb controllers for led strip, only smooth gradual dimming as brightness changed. im may have pwm but i see nothing what people here say i should have seen.

i also agree, you wont get high cri by mixing color leds, you can adjust colors in wide range, but cri, i doubt you'd get what you want

Considering the "preferred" tint is not 100CRI, does it matter?

 

[dc38]

Sorry, but you are talking out of your butt at this point.

If I PWM at 1.5 KHz, you would never notice flicker. I don't mean just not see it, but no physiological effects at all. If I truly PWMed at 600 or 800Hz, you would not either unless you worked to specifically create conditions for it, i.e. something that modulated into a frequency below your particular flicker fusion. It is questionable whether there would even be any physiological effects at those frequencies.

Whatever the pwm frequency of the 4/7 preon 1, i noticed. Nowhere near nauseating or debilitating or even annoying, though...