LNK625DG, Dual Output 12VDC @ 0.4A / 3.3VDC @ 0.4A, Output regulation on 3.3V

6 posts / 0 new

I used PI expert to design a Link Switch CV design. The results are in the attached file. I laid out a board and have a prototype of this design. The problem I am having after initial testing is that the 3.3V output is not regulating well at all. Under light load (5mA) the output is at 3.2V. The designed load should be 0.4A. As soon the load starts to increase the voltage begins to drop. At 20mA it is 2.93V, 70mA it is 2.54V, 200mA it is 2.20V, and 400mA it is a very wavey 2.23V. The voltages here are the mean, there are spikes of 3.2V when the control chip is pulsing the primary (every 180uS @ 200mA load there are 2 or 3 pulses).

The 12V output is lightly loaded during this test, 10mA. If I load the 12V only, it regulates better. It constantly keeps a mean voltage of basically 12V under all load conditions, with ripple up to 1V depending on load. During the 12V loading the 3.3V output increases to a maximum voltage of 4.15V mean (4.5V peak).

I cannot use this 3.3V supply for its intended purpose as it is. Should I just create a single supply of 12V and use a separate regulator to get 3.3V? I thought I could save some components by building a dual voltage supply. Any advice would be appreciated.

LinkSwitch_Reference.pdf595.68 KB


You can try tuning the feedback resistors to achieve tight regulation at 3.3V. May I see the wavey 2.23V waveform you are talking about?

You can also try adding pre-load resistors. Here is a DER that has three outputs for reference. (link: https://ac-dc.power.com/design-support/reference-designs/design-examples/der-198-7-w-10-w-peak-3-output-supply/)


Hi PI-Kobe,

Attached are screen shots. Why is 12V regulating okay, but not the 3.3V? Let me know if you want to see anything else.


3.3V_regulation_issues.pdf2.93 MB


It could be that the bias winding is coupled to the secondary winding with the 12V output. That is a possible reason why it has better regulation for the 12V output.


That is not very clear advice. The design came directly from PI Expert. Do you suggest the 12V coil goes from pin 6 - 7 in the design instead of 5 - 7, so they only share a common return? Should I have two completely separate secondary coils? Or is a dual voltage supply with good regulation is not possible? What is your recommendation?


Sorry for getting back to you just now. Could you give me the voltages for both the 12V and 3v3 output for Full Load and no-load conditions?

Case 1:Output 1, full-load Output 2, full-load: 12V reading? 3v3 reading?
Case 2:Output 1, full-load Output 2, no-load: 12V reading? 3v3 reading?
Case 3:Output 1, no-load Output 2, full-load: 12V reading? 3v3 reading?
Case 4:Output 1, no-load Output 2, no-load: 12V reading? 3v3 reading?