Why LNK626PG no sense output voltage?
When I connected load, output voltage will drop such as 5V drop to 3.2V and IC that not increase duty cycle or time of switch. Some time it increase but not enough. Feedback current about 3mA and the circuit is correctly connected, Load approximate middle load, Input voltage is in operate range (187-276 VAC), Diode is fast recovery and undamage.
Transformer Specification
I design by use the PI software.
Sec 1 : 5V 0.2A
Sec 2 : 13.5V 0.7 A
bias : 15V 0.1A
Feedback : 5.25V 0.1A
Primary inductance : about 1.3 mH +-10%
Leakage inductance : about 300 uH Max
Turn ratio
pri:Sec 1 = 137:10
pri:Sec 2 = 137:14
pri:Feedback = 137: 9
pri:Bias = 137:18
The Sec 1 and Sec 2 are stack winding.
The Feedback and Bias are stack winding.
Thank you very much.
Comments
Hi,
This very high leakage inductance (300uH) will make your power supply has loose regulation problem because the leakage inductance and clamp designs are very important for the LNK-CV. Improper designs will be given you incorrect output voltage information then this will also result to loose regulation.
Try to re-check your transformer and clamp snubber designs. Split on the primary winding will help to decrease leakage inductance around two times.
This is a signal from my circuit.
There is a lot of ringing on the feedback pin waveform. Do you have a damping resistor on the clamp circuit. Can you also attach the schematic and PIXls design file?
Also indicate under what loading the regulation suffers.
Hi guys,
I have designed flyback smps with primary side regulation using lnk626 chip for mobile charging application 5V 2amp and it seems to be working fine in no load condition but when i put load of 1.5 to 2amps there is a ringing appearing on output which occur at regular intervals so that is the problem on which i am stuck, i have attached some pics of waveform and my circuit, need some help.
| Attachment | Size |
|---|---|
| V/div 0.1v and timebase 1 microsec (120.17 KB) | 120.17 KB |
Hi Vignya
Please share your schematic with values and trf construction so that we can help you.
The ringing might be due to the Vds ringing. You can check the Vds waveform.
The LInkSwitch-CV uses primary side sensing. Therefore it senses the output voltage through its reflection on the bias winding. The FB pin is sampled after a delay of 2.5 us.
Need to decrease the leakage inductance by doing split primary.
Add a damping resistor on the clamp.
Thanks
Regards
PI-Chloe
The LInkSwitch-CV uses primary side sensing. Therefore it senses the output voltage through its reflection on the bias winding. The FB pin is sampled after a delay of 2.5 us and in order to ensure that sampling is properly done throughout the operating range you have to ensure that the secondary diode conductcs for atleast 4.5 us.
If the output diode conduction time i s less than 4.5 us then output cannot be sensed properly and regulation will be lost.
Please attach the PIXls design file so I can take a further look.