PMW and capacitors.

[65535]

Is it possible to wire a capacitor in parallel with the led to produce a flicker free output from a PWM driver? It's just a thought I had, it makes sense logically, PWM varies the power on and off very quickly, a cap would average the power and keep it relatively constant power then pwm

 

[EngrPaul]

Is it possible to wire a capacitor in parallel with the led to produce a flicker free output from a PWM driver? It's just a thought I had, it makes sense logically, PWM varies the power on and off very quickly, a cap would average the power and keep it relatively constant power then pwm

Yes, but some of the power will be consumed by the cap, leaveing the LED effectively less bright.

A cap acts like a resistor to AC voltage. It's effective resistance is easily calculated if you know the PWM frequency and the capacitance value.

 

[TranceAddict]

reactance equation:

Xc = 1/(2*Pi*f*C)

if freq is high enough then the reactance is small enough to be ignore, but during low freq pulsating, the reactance will be high enough to open the parallel connection.

but in either situation it doesn't make too much power loss, the main problem is when you parallel a cap to the pwm output, it will significantly reduce the pulse voltage due to charging and discharging, causing the pulse voltage lower than the Vf needed to fire up LED.

 

[Mr Happy]

Yes, but some of the power will be consumed by the cap, leaveing the LED effectively less bright.

A cap acts like a resistor to AC voltage. It's effective resistance is easily calculated if you know the PWM frequency and the capacitance value.

Not quite true. An ideal capacitor will consume no power at all on AC. It has reactance, not resistance. Real capacitors are not ideal and they do tend to have some equivalent series resistance (ESR) in addition to their capacitance, and so they will consume a little power and get warm in the process.

This is not the real problem, however. The real problem is that a discharged capacitor acts like a perfect short circuit across the PWM output. It will lead to extremely high surge currents through the PWM switch with a danger of making the output transistor go .

Therefore although using a capacitor across the output to smooth the output pulses seems like a really good idea, in reality it does not work out.

 

[65535]

So basically any cap. small enough not to fry the driver wouldn't be big enough to stop any flickering.

 

[Mr Happy]

So basically any cap. small enough not to fry the driver wouldn't be big enough to stop any flickering.

In a nutshell, yes.

 

[JWP_EE]

Just to be accurate, the output of a PWM driver is pulsating DC. The current is always in one direction. Alternating Current (AC) flows in one direction and then reverses and flows in the other.

The capacitor will look like a short to the driver only for the time it takes for it to charge to the Vf of the LED. This could destroy the driver if the capacitance was too large.

 

[65535]

Oh well, I like CC drivers anyways.

 

[Benson]

Just to be accurate, the output of a PWM driver is pulsating DC. The current is always in one direction. Alternating Current (AC) flows in one direction and then reverses and flows in the other.

The capacitor will look like a short to the driver only for the time it takes for it to charge to the Vf of the LED. This could destroy the driver if the capacitance was too large.

A series resistance could be added. This does cause some losses, obviously, but it's what I'd do if I wanted to smooth a PWMed LED.

 

[65535]

What do you think about this?

 

[TranceAddict]

what is the purpose having a diode there?

 

[Mr Happy]

What do you think about this?

I'm not sure. However, as soon as you introduce a resistor to reduce surge currents you also introduce power losses in the resistor. Since the goal of a PWM controller is to regulate power delivery without losses the introduction of a resistor undoes the purpose in having PWM regulation. At that point one may as well use a buck regulator.

 

[65535]

I was thinking with the resistor wired in the way it was, the LED would see the average power without flickering once the cap was charged to the proper power level. The resistor shouldn't effect the efficiency, maybe it would. More a curiosity than a need though. I do like CC powersupplies the best.