Hi,

LNK-PL operation and output current control is digital.  As such, hysterisis is needed in order to regulate the output current in a desired level.  The ripple in the feedback pin should be kept small in order to minize the difference when there is a line input disturbance.

Increase the capacitance across the FB pin or increase the series resistor gradually to FB-pin to increase the time constant.  Stop increasing the time constant when the output current starts to jitter.  FB pin ripple should not be less than 30% as a guideline. 

Make sure that it is operating in discontinous mode.  The IC is designed to operate in this mode only.

Use LYT-4 for tight regulation application.  Refer link for reference:

http://www.powerint.com/sites/default/files/PDFFiles/der370.pdf

 and http://www.powerint.com/sites/default/files/PDFFiles/der365.pdf

Regards,

PI-Jono

I tried both trick for good current regulation.

Increasing capacitance not works but increasing FB series resister works good.

But there is another issue i am facing in every driver.

It takes 2seconds to start the LED after input applied. If you swtich off light and start again then no problem. But when you wait for an hour and try to start it takes longer time to start.

Another question ..

Is there any problem increasing FB series resister upto 15K ??

Hi,

Increasing series FB resistor will slightly reduce the efficiency.  Do not increase such that there will be no ripple on the FB pin.  Since this is high PF controller we need the ripple in the FB pin to regulate the output. 

This is the very first feedback to observe slow start-up for a tapped-buck converter after burn-in.  Kindly check the parts on the design if everything is rated above 100 degree C.  Check also if there is some bad soldered components on the board due to thermal expansion. 

 Check also if the power inductor is not not saturating.  During start-up the power inductor is experiencing a huge flux swing.  It might be hitting the SOA prtection that is why it is taking some time to start-up.

Great you had some improvement on your design.

Regards,

PI-Jono

Thanks you so much for your quick reply.

I have set FB series resister on 5.1K. And getting 86% efficiency an 0.91PF. So there is no issue with that.

I have made 15 driver to check current regulation and stability of circuit. All are working good but all that have same problem. It's take longer time to first time start up. I noted that when I decharge output capacitor completely it takes longer time.

I cant understand about  " huge flux swing" and "SOA prtection".

PI software shows to use 300uH but I am using 185uH to achive good powerfactor and PF.

I tried to use 300uH but geting poor efficiency(82%) and 0.82 PF. My circuit paramaters are following

Input : 195-265V 50Hz

Output : 53.5V, 280mA, 15Watt with Dimming

Reference : DER-322

Efficiency : 87%

PF : 0.91

I have made following changed into DER-322

Inductor : EE25, 39Turns,  185uH, 27AWG (Instead of RM6)

D2 diode : US2G (Instead of STTH3R06U)

R8 Bias series resister : 50K (Instead of 36K, but tried 22K,33K)

R6 FB series resister : 5.1K (instead of 3.3K)

C7 output cap : 200uF (instead of 100uf, but tried 100uf)

R7 output resister : 220K (instead of 100K)

VR3 & VR1 are removed because no OVP is required.

D5 & D6 are replaced with 1N4007 (instead of two BAV21WS-7-F)

I must have to use LNK406KG because ordered bulk quantity already. So please guide me to sorout this issue.

Forget to mention another problem.

When I tried to increase inductor more than 190uH than its starts with one blicks. Dont know but this very crazy thing. But problem solved after using 185uH inductor.

Hi,

Please make sure the inductor is not saturating.  Maybe by grinding the core to reduce the inductance  with same turns improves the drain current.

It is hard for us to conclude unless we see some waveforms. Kindly observe the drain current of 900 vs 850uH

Thanks,

PI-Jono