I'm using an LP5520 in manual mode and would like to be able to create very dim illumination levels with the smallest non-zero current level. Using a photodiode, I measure the same amount of ripple in the LED output, independent of PWM on-time. When the PWM is always on, I get a 0.6% ripple (7.2 mV out of 1.19 V from the XZ amp on my photodiode circuit). The ripple voltage stays the same as I drive the PWM on duty cycle to near 0. Thus, with a 0.9% duty cycle, I get 7 mV peak output and 7.4 mV of ripple for a huge variation in light output. The variation in light output at 0.9% duty cycle is easily seen by both my eye (as kind of a non-steady flickering in the light) and my photodiode. At 100% duty cycle, the light by eye appears constant. In this experiment, the PWM period was 250 us, so there is no way I would perceive a flicker if each flash had the same brightness. I was monitoring the PWM signal into the LP5520, and it was rock solid.
My boost voltage seems very steady -- some ringing during the switching from current on to current off, but no change in steady state.
I'm thinking the issue is noise in the DAC output creating the current source.
Any suggestions?
Thanks!