Primary side controlled LED drivers are generally more inefficient than non triac dimmable circuits?

4 posts / 0 new

Is the above question true?

The Linkswitch PH uses the bias winding to regulate. However, the secondary (output) and bias winding are only  loosely coupled due to the isolation barrier, so surely the output regulation will be inaccurate?

If it was highly accurate, then why  would anyone use optocouplers in any offline switch mode power supply?

The question is a bit convoluted. I would say dimmable circuits are generally less efficient than non-dimmable circuits, but I don't see any correlation between a primary side driver to non-dimming circuit.

If what you meant to ask is whether LNK-PH, being a primary-side controller, is generally inefficient than other competitor solution for non-dimming design, the answer is absolutely not.   You can check some of our DERs to verify efficiency data and some designs are 90% efficient.


The line/load regulation will definitely suffer if the bias is loosely coupled but if it is tightly wound with the secondary, then regulation is better. LYT-4 family can regulate to within 5% with regular bias feedback.


However, sometimes, customers would want a very tight regulation (<3%). Some of our designs have opto-coupler feedback to address this requirement. The tradeoff is additional cost.


thanks but I cant see how the bias winding can be "tightly wound" with the secondary? The bias winding is on the other side of the isolation barrier to the secondary?

Adding series resistor to the bias winding reduces the effect of leakage inductance which may affect regulation. Please see AN-49 for more info about how the bias resistor values affects regulation.