LNK420, what am I doing wrong?

2 posts / 0 new

Hi guys, I did a development with PIExpert 8 for an LED driver of 78V 800mA. I selected LNK420E and transformer with EER28. I made the printed circuit according to my needs and I carefully respected all the indications. In fact the prototype works very well except for two things:

A) Operation is DM instead of CC.

B) The snubber net heats too much.Specifically R5

The first one I suppose is corrected increasing the inductance but the second one requires more work.

Please see in attached schematic and oscilloscope images.

The yellow color corresponds to Vcc (rectified voltage input, transformer pin 1) and green to the Drain of the LNK420. With great courage removed R6, that is to say that I completely eliminated the mesh. I know, a very dangerous game. Incredibly everything works well and the transient opening, although reaches almost 800V. Congratulations to PI !.I can kill that transient easily with 1nF and I have enough time to download it with 200Kohm in parallel. They agree?

Then I do not understand why they keep circulating almost 10mA per R5 throughout the toff. That energy is lost uselessly and D1 makes it look directly as current over the LEDs.

Another question: what is the mathematical expression that links the voltage in pin V, the current in FB with the current or circuit activity cycle ?.

The last one: is it absolutely necessaryD3? 

Thank you very much. 

첨부 파일Size
Oscillogram 188.32 KB
Oscillogram 289.92 KB
Schematic180.88 KB


The snubber design needs further optimization to reduce R5. First, the VR1 value is too low. 200V is typical. Then the R-C component is too much. For C3, keep it below 2.2nF and R5 above 100k to manage the dissipation.

You probably need to use split-winding to reduce the leakage inductance and the reduce the ringing, which in turn reduces the demand for stronger clamp.

D3 is necessary to prevent reverse current going into the pin. It is not recommended to remove it to prevent potential issue.

For the mathematical relationship for the vpin and FB current, there is an internal control that relates the ratio between IV and IFB.

Increasing the IFB increase the output current while increasing the IV decreases the current.