Insuring Short-term and Long-term LED Instensity Stability

Other Hardware LEDs and Multiplexing
1

I am wanting to alter a color measuring instrument currently utilizing fluorescent lamps to LEDs. Further, I want to incorporate a feedback from a lux meter to adjust the LED's power such that both short-term variations (e.g., temperature induced) and long-term (e.g., LED aging) are minimized. My thoughts were to set the LEDs intensity at time zero when the LEDs are new to about 90% of the optimum operating intensity and make power adjustments as necessary to insure the stability. The color measuring device I am using is based on a digital camera used as the detector, so any photographic limitations will also apply to this project (e.g., the induced fluctuation, i.e., noise, in the camera's RGB output as a function of a light source having an inherent pulsating intensity).

Having made a significant attempt to review the extensive literature on adjusting LED intensity, most of the literature is directed toward dimming of LEDs where the preferred method is PWM of a constant voltage to prevent failure of the LED as a result of low voltage. I did however run across the YouTube video below ("How To Make An LED Panel" by DIY Perks, see link below) written by a photographer. Disclosed was a panel comprising 5050 LED strips utilizing a variable voltage driver. In the video, he demonstrated the flicker often observed with constant voltage circuits could be eliminated with a variable voltage driver. He did however limit the voltages utilized to a range between the lower and upper limits of the LEDs utilized.

For my particular application, I too am wanting to create a panel comprising a bank of several 5050 LED strips and as in the video, I'm not intending to dim the LEDs in the conventional manner, but only wanting to make small adjustments in the power supplied to the LEDs to maintain a stable intensity light source. My question is: Is a constant voltage power source for the LED's a bad thing under this particular application? I'm currently taking digital images of the samples utilizing the fluorescent lamps at 1/60 of a second to get a complete AC cycle powering the lamps so as to minimize the pulsation of the fluorescent lamps; this length of shutter speed my make my concerns mute with PWM, but I'd rather ask others who know more than me before just starting a project.

Thanks to all who are able to share their input,

Jack

2

My question is: Is a constant voltage power source for the LED's a bad thing under this particular application?

Yes it is.
Constant voltage means the current is limited by a resistor and the actual current is critically dependent on the forward voltage drop of the LED. This changes with both temperature and aging.

I would have thought a precision constant current is what is needed, with the ability to make small changes to control the brightness. The brightness is a function of the current, although aging will affect the brightness as well. Try not to run it at too high a current to reduce this aging effect.

3

@ Grumpy Mike, thanks for your input.