1
\$\begingroup\$

I have been reading up on LED bulbs and it seems power factor correction (PFC) is quite needed in most designs.

My goal is to achieve a PF greater than 0.9. I am considering using an LED Driver with built-in PFC like Fairchild's FLS0116. From the datasheet (and attached photo), it's easy to see that the IC ensures that the LED current follows the shape of the input voltage waveform as closely as possible - including where the voltage crosses zero.

enter image description here

My question is: because the LED current is zero every 10ms, will there be any visible flicker?

\$\endgroup\$
  • \$\begingroup\$ That's 100Hz. Fluorescent flickers at these frequencies, so it shouldn't be a problem. Source \$\endgroup\$ – WalyKu Feb 20 '15 at 15:17
1
\$\begingroup\$

PFC on switch mode power supplies would not be very effective if the dc output dropped to zero every ten milli seconds. This of course is avoided by using capacitors to store energy so that PFC works as expected.

I'm not sure about the chip you are using but it would seem a trivial addition of a couple of components would all that is needed to avoid LED flicker.

\$\endgroup\$
  • \$\begingroup\$ Would the chip still maintain a high PF even if there is a capacitor at the LEDs? \$\endgroup\$ – Saad Feb 20 '15 at 16:05
  • \$\begingroup\$ If you control the charge into the cap correctly it is indistinguishable from current into a resistor. Once stored it can be used when the ac voltage passes thru zero. \$\endgroup\$ – Andy aka Feb 20 '15 at 19:28
1
\$\begingroup\$

As it stands there would be flicker at 100Hz. If that is visible or not strongly depends on the person involved. Even if not visible directly, there are reports about headaches and other health issues which are probably connected to the flicker (like in the source given by Kurtovic)

As LEDs mostly use a different phosphor, which has no persistence as it's based on scintillation, they really turn off when no current is flowing through them. So I'd suspect that the flickering would be more noticeable compared to the flicker of traditional CCFLs, where phosphors are used which have a persistence in the range of 10ms or more. So the flicker frequency would be the same but the amplitude would be higher.

\$\endgroup\$
1
\$\begingroup\$

At 100Hz, the flicker may not be perceptible to most people. The manufacturers of this type of LED drivers must think so to at least a certain degree. As a side note, a digital camera would pick up the 100Hz easily as horizontal bands across the captured image.

The flicker (the LED current) can be smoothed out to a certain extent by a capacitor. Refer to FLS1016 data sheet -- figure 11 -- C5.

This type of single stage design is not going to give a rock solid LED current.

As refer to in the other answers, the more typical constant voltage power supply with PFC usually has two stages. A PFC stage that transfer energy to a "large" capacitor to give a semi-regulated DC. Then a second regulator stage would do the conversion to the final required power outputs. This second stage can be a constant current regulator for LED. I have not looked into whether there is any single chip that would do both stages for LED.

\$\endgroup\$
0
\$\begingroup\$

Most of the LED power supplies with PFC that I have seen provide smooth DC output. The PFC stage is part of the switch-mode power supply (or in front of the SMPS) and is used to make the driver circuit appear as a resistive load to the AC Mains supply.

Most modern LED power supplies are configured to provide constant current output over a range of output voltage.

\$\endgroup\$
  • \$\begingroup\$ Could you suggest a chip? \$\endgroup\$ – Saad Feb 20 '15 at 17:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.