Pwm on power led with more current than needed

Rimbaldo

Newly Enlightened
Joined
Jan 24, 2013
Messages
20
Hi!!

I have a 3w common anode RGB power led. Each color needs 300-350mA max to work.
I've got a few 600mA tiny pwm-able drivers (based on PT4115 chip). So I could use three of them with arduino to generate PWM for each color.

But as each color needs half the current these drivers provide, if I PWM each of them at 50% max, It would be the same effect as having 3 drivers of 300 mA each, and my leds would be at 100%?

If positive, would I have the same color (in 0-255 pwm values) that would be for instance (255,255,255) for a RGB color in the led, would be the same effect in my case that the pwm is cut by half as (128,128,128)? I mean instead of full pwm to 255, half pwm to 128 would provide the max current in my led (300mA for each color, in a 600 mA driver)?

Thanks!
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
With ideal diodes, that would be correct. With real diodes, it's close to correct, but not quite.

There are some differences between a diode at 300 mA 100% DC (duty cycle) and at 600 mA 50% DC.

1. The Vf at 600 mA is higher, so the power consumption is higher.

2. Efficacy/efficiency is (generally) lower at 600 mA, so the average output is less.

3. Both 1 and 2 mean higher heat generation.

4. Since efficacy changes are probably not the same in each color, combined color will likely not be the same.

Also, I don't know if colored LEDs have this issue, but pumped phosphor white LEDs show color changes with current changes and also with temperature changes.

How significant these differences are depends a lot on your LED and how it is cooled, not to mention how critical your color balances are.

That said, if you color balance at 600 mA/50%, then change the DC, the only variable affecting color is die temperature. This would yield much better results than balancing at 300mA/100% then changing to 600mA/PWM.
 

Rimbaldo

Newly Enlightened
Joined
Jan 24, 2013
Messages
20
With ideal diodes, that would be correct. With real diodes, it's close to correct, but not quite.

There are some differences between a diode at 300 mA 100% DC (duty cycle) and at 600 mA 50% DC.

1. The Vf at 600 mA is higher, so the power consumption is higher.

2. Efficacy/efficiency is (generally) lower at 600 mA, so the average output is less.

3. Both 1 and 2 mean higher heat generation.

4. Since efficacy changes are probably not the same in each color, combined color will likely not be the same.

Also, I don't know if colored LEDs have this issue, but pumped phosphor white LEDs show color changes with current changes and also with temperature changes.

How significant these differences are depends a lot on your LED and how it is cooled, not to mention how critical your color balances are.

That said, if you color balance at 600 mA/50%, then change the DC, the only variable affecting color is die temperature. This would yield much better results than balancing at 300mA/100% then changing to 600mA/PWM.


Hi! Thanks for your explanation!!

In this driver I've got, there were two smd resistors in paralel (R300 and R330) I removed the R330 one and the current lowered to 350 mA!

So now, pwm at 50% would give me the brightest of each color.
 
Top