Poor coupling okay?
I want to use a LinkZero IC with a poorly coupled transformer primary and secondary (like k = .4 to .8). This means a large leakage inductance. It seems to me that this won't work because of the way the IC monitors the primary waveform to detect the secondary voltage.
Is there a way around this? Would the proper snubber square up the waveform acceptably? Are there any specs on waveform levels and timing I can refer to?
Thanks!
Comments
I'm talking about REALLY bad coupling, not just a k of .95 instead of .99. The leakage will amount to an inductance close to half of the primary inductance.
If you can tightly couple the secondary and bias/feedback windings, your regulation will be better. However, it will not be great. You may have another issue with LNK Zero is that the power supply will be processing a sizable amount of power into the drain clamp. The IC will not recognize a unloaded state and not switch into the zero power mode.
Z. Cochrane
I don't have the option of good coupling for this design, unfortunately, but I'd like to make the LinkZero work if possible. Good point about the wasted power -- that sort of takes the shine off this approach.
Is the timing for waveform sampling available? If I knew when the IC expected the secondary voltage to be accurately present at the primary winding I might be able to design around that.
Linkswitch zero averages the feedback over many switch cycles. The feedback voltage must be representative of the output voltage to have good regulation. Improvements may be made by blanking the leakage spike on the feedback signal so that it tracks the output.
Z. Cochrane
As a general rule the better the coupling, the better the regulation. If you can maintain good coupling between the secondary and bias/feedback windings will help the regulation. Filtering the leakage spike on the bias/feedback winding can also help.
Z. Cochrane