amplitude of the current. NOT by changing the duty cycle.
Like in AC voltage there is a peak to peak and an RMS value.
In a driver the peak amplitude remains the same and the average or effective current changes based on the width of the current pluses. In an HLG the pulses come at at rate of 100 KHz. The width of the pulse has a duty cycle of 0-100%. Or nearly. some drivers need at least a negative spike (100%) and or positive spike (0%).
All switching power supplies, voltage or current, use PWM to regulate the voltage or current. The switching signal's duty cycle is varied for regulation. The PWM switching signal is typically generated by an oscillator circuit with in the regulator chip. Most drivers fix the PWM switching frequency to a value that is compatible with the output inductor for maximum efficiency. Some will dynamically vary the frequency based on the load. Efficiency is a function of the switching frequency, the on resistance of the output FET, inductance of output inductor, and series resistance of the inductor. To maintain efficiency the switching frequency is fixed to balance with the inductance. The only means left is to modulate the pulse width for regulation.
Mean Well's 3 in 1 dimming is really only one circuit that uses a voltage reference to set the PWM duty cycle.
When the reference voltage is at 10v the duty cycle is 100%.
When the reference voltage is at 0v the duty cycle is (or near) 0%.
There is a weak pull-up resistor connected to an internal 10v source.
When a resistor is connected the resistor creates a voltage divider with the pull-up resistor.
When left open the reference voltage is pulled up to 10v by the pull-up resistor.
When a voltage source between 0 and 10v is applied, it over-rides the pull-up resistor's 10v.
When the applied voltage is alternated between 0 and 10v the duty is also alternated between 0 and 100%.
Their 3 in 1 dimming is patented but a quick patent search I conducted did not find it.
If someone can find it, I am sure it works as I described.
There is very little difference between one switching LED driver design and another. They all use PWM to regulate the current. A CC driver cannot regulate voltage as the voltage is determined by the forward voltage of the LED string. The current is typically sensed by a shunt resistor.
I think what he was saying is that the actual output current (in the load) does not consist of pulses that are reduced in width to dim - the output is filtered and smoothed out by that point.
Typically an LED driver's output has the pulses. How much they are "smoothed" is indicated by the ripple current spec in the datasheet. Mean Well does not spec current ripple but rather peak to peak voltage ripple which is a spec for a DC voltage supply.
Additionally there is conduction mode. A driver in continuous conduction mode will look like the ripple is elevated to the middle of the signal as shown in the green waveform previously posted. When in discontinuous conduction mode the output reaches ground. Do a Google image search for discontinuous conduction mode.
